Subscribe to Our Newsletter

Marital choice and partner selection are rich social interactions for economists to study as they involve many economic theory-based behaviors. For example, there is asymmetric information, where each potential partner has information about themselves that the other person does not know. There is also signaling, in which an individual intends to convey unobservable information to a potential partner. In a marriage, there are many decisions that need to be made by spouses regarding both selfish (i.e. personal) interests and communal interests, such as chores. Some form of implicit or explicit negotiation occurs, even if it does not seem like it to us.

It appears that many of the implicit impacts of spousal preferences differently impacts the genders, especially worsening the career outcomes of women.

Partner Selection – Gender Differences

As we anecdotally know and feel, men and women approach partner selection differently. But exactly how? It is usually difficult to get truthful responses from individuals just by asking them, especially on a potentially sensitive topic such as dating preferences. Thus, it is important to use a research method such that individuals will truthfully report their preferences. One such way is through setting up a field experiment.1

Fisman et al. (2006) conducted such a field experiment experiment in order to infer people’s preferences by hosting real speed-dating2 events. During these events, participants would record after each round whether they would like to see each other again. If a match did occur, contact information would be exchanged, thus participants had an incentive to be truthful. What Fisman et al. found is that women put more emphasis on intelligence and race, while men put more emphasis on attractiveness. Moreover, the men preferred women who are less intelligent and, especially, less ambitious than them. These field experiment findings on ambition are consistent with what is referred to as the ‘social structure theory’. This theory posits that social arrangements and societal roles are more important in determining individuals’ particular decisions rather than biological reasons. 

So why are economists concerned about these partner preferences – after all, they may simply be individual choices that have little bearing on others? The old economic adage is that individuals respond to incentives. This implies that simply knowing about other people’s preferences might alter your own behavior. As it turns out, further economic studies have shown this to be the case.

Real Costs of Preferences

Bursztyn, Fujiwara and Pallais (2017) conducted a field experiment on newly admitted MBA students. The authors used a survey to ask students certain career and personality questions. As with all survey-type studies, there is an issue of whether respondents will give truthful answers. We covered an example of how survey design was important to elucidating honest responses when we discussed gift giving. However, in this instance, the MBA surveys were perceived by the surveyed, as something more than just a meaningless questionnaire. As this was a newly joining MBA cohort, this was the first time they would be offering their information and preferences to the career center. The career center does use this information to set up internships – individuals that aren’t willing to travel as much would be steered away from consulting, while people that prefer shorter hours, would be less likely to get internships with investment banks.

Bursztyn, Fujiwara and Pallais conducted the surveys and randomly told some of the students that their results will be shared publicly with other students, while others were told their answers would remain private or anonymous. The results of this study were illuminating. Single women who believed their answers may be made public in front of their classmates, desired $19,000 lower annual salary ($112,000 vs $131,000), desired less work travel per month (6.6 days vs 13.5 days) and were willing to work fewer hours per week (48.3 hours vs 52.2 hours) than if they believed their answers would be anonymous! The same pattern was found for specific personality traits – single women were less likely to want to lead or state they were ambitious if their information would be shared in a public setting. 

Confirming Causality

The above experiment might not be sufficient to determine whether men’s preferences might be causing single women to act differently. Theoretically, there could be an unobserved variable that could make it more likely for women to both be single and alter their response if it would be presented in a public setting.

To address that issue, Bursztyn, Fujiwara and Pallais performed additional experiments and analysis. The authors conducted an experiment during which students, in smaller groups, were asked to rank which type of jobs they would prefer (higher salary and longer hours vs lower salary and shorter hours) and told that their responses may be discussed at the end of the session. When placed in all female groups, women were far more likely to choose jobs with higher salaries and longer hours, as well as jobs with faster promotion schedules and included more travel. Women were more likely to prefer lower salary outcomes in groups in which there were more single men as opposed to married men. This difference in response leads more credence to the idea that it is social expectations influencing behavior of single women.

The authors also noted that single women tended to have lower class participation grades than married women. In a class setting, their participation would be seen by others. However, their performance on midterms, final exams and problems was similar to married women, suggesting that skills and ability are not an explanatory factor.

An additional survey asked students about their experiences. Bursztyn, Fujiwara and Pallais found that 64% of single women avoided asking for raises or promotions to avoid looking “too ambitious, assertive, or pushy”. Only 39% of women who were married or in a serious relationship and 27% of men said the same. Similarly, single women tended to avoid speaking up in meetings to avoid looking too ambitious. All of these results may also be somewhat downward biased,3 as people pursuing an MBA are more likely to be more focused on career enhancement and getting a higher salary. 

These results demonstrate that the preferences of men for less ambitious women has real impacts on women’s decisions and behaviors. Whether consciously or subconsciously, evidence suggests that women alter their behavior to appear more ‘attractive’ to potential partners. What are the ways we can deal with this situation? One solution that can work in the near-term proposed by Bursztyn et al. is to encourage schools and workplaces to make certain actions less observable. As the studies have shown, single women mainly alter their decisions when it would be observable to others. Thus, removing this channel would potentially prevent this behavioral adjustment.

Ambitious Women and Divorce

The Bursztyn, Fujiwara and Pallais study, given that there were lower stakes at play, might still be seen with a dose of skepticism. To determine whether women’s ambition could impact marital outcomes, a more direct study would be preferable. Folke and Rickne (2020), using an extremely rich Swedish data set, were able to study the impact of promotions on divorce rates. Folke and Rickne have 30 years of data and are able to track individuals that are applying to be promoted to become a mayor or a parliamentarian, which are both seen as very high positions (both from a reputational and financial perspective) in Swedish society.

Firstly, Folke and Rickne find that divorce rates for promoted women go up, while there is no change for promoted men. Secondly, women that were promoted are twice as likely to get divorced than women that attempted, but failed to gain promotion. Folke and Rickne are also able to track pre-promotion data and thus, are able to rule out any differences between the women who got promoted or failed to get promoted, prior to the promotion. This pattern did not only hold in political roles – undertaking the same analysis on CEO promotions, the study found very similar results. The figures below visualizes the impact for political roles. As can be seen, in the 4 years prior to the election, women that would be promoted and women that failed to get promoted remain married at the same rates. After promotion, the marriage rate of promoted women drops by more than for women that failed to get promoted. No such effect can be seen for men. 

The rich dataset allowed Folke and Rickne to analyze which types of couples were more likely to divorce. They found that gender-traditional couples, where there was a larger spousal age-gap and more gender-skewed division of parental leave,4 were more likely to divorce after the promotion. Women in more gender-equal marriages did not divorce more often after a promotion. Folke and Rickne highlight the issue between couple formation and the glass ceiling for women in the labor market. High ability men are often in relationships where the focus is on the men’s career, while high ability women tend to be in dual-earner relationships where the woman is primary care-giver. Women in traditional marriages find it tougher to get higher responsibility jobs due to the adjustment costs within the marriage, which includes task renegotiation within the couple.

Addressing the Issue – Long-Run

The above research describes the current social and economic situation. In economics, this is called ‘positive’ economics, which is describing the world as it is currently. The issue that arises is that given that men today prefer less ambitious women, women may find it rational to act in a less ambitious way in order to find a spouse, if that is their goal. This behavior, although logical currently, is clearly costly for women in the future. 

One could, therefore, argue that in order to improve outcomes for women and the economy in general, the perspective of men on women’s ambitions should be changed. This is referred to as ‘normative’ economics – i.e. what the economy ought to be. Resolving the perception by men of women in the workplace, which is a cultural phenomenon, is not an easy question – the most likely solution is a long-run one. Economists have also studied what can lead to cultural change in the long run.

Fernandez et al. (2004) assess one cultural component that may be important in forming men’s attitudes towards working women – working mothers. Throughout history, Fernandez et al. point out how several changes impacted women in the labor force – consumer durables (most famously, the washing machine) that made housework easier, the ‘pill’ that enabled women to control their fertility, leading to controlling their career, and the growth of service sector ‘white-collar’ jobs. All of these changes had a dramatic impact on women’s participation in the labor force, changing the entire cultural perspective around women in the workforce. Regarding more recent times, Fernandez et al. argue both theoretically and empirically, that growing up with a working mother has changed men’s attitudes toward working women. They found that having a working mother increases the probability of the man having a working wife by 24 to 32 percentage points.

Societal Perceptions and Outcomes for Women

Finding a spouse clearly impacts men and women very differently. The preferences of men and women also unequally affect the genders. This results in real costs for women as both their careers and earnings potentially suffer. There may be ways to mitigate some of these effects, such as by anonymizing participants (for example, salary negotiations could be done without personally identifiable data) or obscuring certain actions (such as reviews) in the workplace to reduce observability.

However, any such stopgaps will not address the key long-run issue. Career decisions will always be publicly visible, so in order to enable women to reach their career goals and not having to worry about the implications it may have on their marital life, the perception of womens’ ambitions and desirability must change. That is a process that takes generations.

Finally, the above research has shown several important ways to assess the validity of studies which use survey data. When using survey data, it is critical to create incentives that will elucidate truthful responses. Surveys and social experiments in laboratories can typically suffer from incorrect data. If accessible, however, studies that use officially recorded data, such as the Folke and Rickne Swedish study of political candidates, will typically be better in establishing causality. This is due to the fact that the data they used was unlikely to be erroneous or suffer from any significant biases.5

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Subscribe to Our Newsletter

In almost any introductory economics class, ‘free’ trade is presented as a welfare-improving policy. However, in graduate economics classes, you learn that this increase in welfare is only certain in the long run. But what happens in the short-run? Here the picture gets murkier. Recently, a study by Finkelstein, Notowidigdo and Shi (2026) (“FNS”) and by Noghanibehambari and Fletcher (2026) (“NF”) looked at how the North American Free Trade Agreement (NAFTA) between the US, Mexico and Canada may have shortened lives of US manufacturing workers.

Opening Up to Trade

NAFTA, which removed tariff and non-tariff trade barriers between the US, Mexico and Canada, became law on January 1, 1994. NAFTA de facto created a single goods market that covered, at the time, 400 million people, and accounted for one third of the world’s output. The main impact of NAFTA was the removal of trade barriers between the US and Mexico (as Canadian-US trade barriers were effectively removed in 1987). Over 15 years, NAFTA reduced US tariffs on goods from Mexico from 2% to 0%:

The seminal paper by Caliendo and Parro (2015) found that the impacts of NAFTA on welfare were relatively muted – Mexico benefitted the most with a 1.3% increase in welfare, the US gained about 0.08%, while Canada saw a minor welfare decline of 0.06%. However, all countries saw an increase in real wages – i.e. workers were able to purchase more goods/services after the passing of NAFTA.

But the Caliendo and Parro paper looked at the question of trade liberalization mainly from a total income perspective. Since trade liberalization shocks may also impact workers’ well being in other ways, FNS and NF wanted to understand what happens to mortality after such a shock.

Concentrated Manufacturing Shock

To determine how mortality changed due to the impact of NAFTA, both papers had to first determine which areas in the US were most ‘exposed’ to NAFTA. To measure this exposure (or vulnerability), economists looked at three key factors:

1. Whether the industry had a high share of imports from Mexico – this captures the fact that Mexico had a comparative advantage in this industry than in other industries;

2. The size of tariffs prior to NAFTA in the industry – a higher tariff would mean there would be greater competition from Mexican producers once the tariffs are removed;

3. The share of US workers in a given region working in the industries that are influenced by 1 and 2 – the more workers working in industries exposed to Mexican imports, the bigger the shock the region will face.

The map below from FNS shows which US regions were most vulnerable by quartiles (the darkest color being the most vulnerable):

Using the fact that regions had different exposures to NAFTA, FNS estimated how mortality responded to NAFTA. FNS found that average annual mortality increased by 0.68% over the period 1994 to 2008. To put into context, average annual mortality between 1965 and 2000 fell between 1% to 1.5% (Cutler and Meara, 2004). This means that NAFTA may have eroded half the improvements in mortality.

Mortality Increases Among Working Men and The Elderly

It is important to note that the 0.68% increase in mortality is an average. NF, using similar methodology in their paper, found that the mortality increase in the most NAFTA exposed areas was 4.2%. FNS broke out the increase in mortality and found two major groups accounted for a large share of the increase:

• One-third of the increase in mortality was due to an increase in deaths of working age men, between 25 to 64 years old;

• Half of the increase in mortality was driven by people aged 65 or more.

The mortality increase of working age men suggests that direct job loss/displacement may be the cause of overall increased mortality.

Regarding the increase in elderly mortality, a potential explanation is that adults over 65 may have depended on their children for financial assistance. Thus, when a worker lost a job during NAFTA, it may have also adversely impacted their parents/grandparents. FNS found evidence for this – over 60% of the increase in the mortality of the elderly was driven by increased mortality of elderly female widows, who often depend financially more on their children.

Did NAFTA Reduce Welfare?

FNS conducted a high level welfare analysis which, unlike previous research that focused predominantly on the income impacts of NAFTA, also included the mortality impacts they estimated. As noted previously, prior estimates of NAFTA showed a 0.08% welfare increase in the US. FNS estimated that including the mortality impacts, US welfare most likely fell – the reduction in welfare due to NAFTA was around 0.11% to 0.69% (depending on assumptions).

But, there are certain caveats around this conclusion that need to be taken into account.

Understanding the Econometrics of the Estimation Method

FNS point out that their estimation relies on looking at differences in mortality between regions – that is, how did mortality change in a region that was more exposed to NAFTA compared to a region that was less exposed. In effect, we are looking at a difference, and in simple terms we have the following problem: If previously, we had a difference of 3 (for example, one group was 10 and the other group was 7, giving a difference of 3), and now we have a difference of 4, we could have two causes for it:

• One group went up by 1 – so 11-7 = 4; or

• One group went down by 1 – so 10-6 = 4.

(A mix of both is also possible). In our context, this means that the observed increase in mortality could be driven in part by the fact that mortality for non-NAFTA exposed individuals decreased thanks to NAFTA.

However, looking at other studies, FNS argue that it is very unlikely that mortality for non-NAFTA exposed individuals reduced meaningfully. Most studies find that income increases have no impact or, surprisingly, a negative impact on mortality. Thus, in my opinion, it does appear that NAFTA did increase mortality of vulnerable workers.

Uniqueness of Trade Shocks

The adverse impacts of NAFTA on mortality appear to be driven by job loss/displacement. One could expect that job loss would always increase mortality. Surprisingly, economists have found that during recessions, when job losses may even be greater, mortality falls (Ariizumi and Schirle, 2012). FNS confirmed this by looking at the impact of the 2008 Great Recession on mortality. They found that mortality reduced in areas more exposed to the adverse impacts of the Great Recession. But, trade shocks like NAFTA or China joining the World Trade Organization in 2001, increase mortality.

The reason behind these different mortality outcomes to what appear to be similar shocks is a bit of a puzzle. One proposed answer is that trade shocks adversely impact jobs that are unionized, which typically have higher wages and other benefits. FNS found that regions with higher unionization rates did exhibit larger mortality increases due to NAFTA. This is suggestive evidence for that hypothesis.

Can We Better Redistribute the Gains of Trade?

The research by FNS and NF shows that trade shocks like NAFTA do have strong negative impacts on a concentrated group of people exposed to these shocks. Are there ways to mitigate this issue?

Providing Cash/In-Kind Transfers

One solution might be to offer income support to workers that are exposed to job loss from trade shocks. However, studies by Chorniy, Finkelstein and Notowidigdo (2025), as well as Evans and Moore (2012), have found that such cash transfers increase mortality, often caused by alcohol or drug abuse. In-kind transfers, like food vouchers, appear to not increase mortality. Just giving cash may not be sufficient to alleviate the negative consequences.

As mentioned earlier, there was a significant increase in mortality among the elderly that may have been driven by reduced financial support from their children laid off from their jobs. Although exacerbated by trade shocks, assisting the elderly should be addressable by policy, whether through better design of social insurance or more direct financial assistance.

Key Takeaways

The FNS and NF research has three key takeaways – one for economists/researchers and two for policy designers.

For economists:

1. The impacts of trade liberalization on individuals is very varied in the short run. Typically, as economists, we have looked at trade liberalization from a long run perspective which showed unambiguous long run benefits. However, a lot of the adverse impacts occur during the transition phase of opening up to trade, concentrated on a narrow group. The focus on the transition is crucial to understand the consequences of trade liberalization, as well as informing the design of better economic policy.

For Policy Makers:

1. Alleviating the negative impacts of the trade shock is important, as certain groups face the brunt of negative consequences. This may entail redistributing via taxes into in-kind transfers, as well as more broad outreach to communities that may see the negative consequences.

2. Tariffs and trade protection, especially after trade liberalization already occurred, may result in the amplification of the negative outcomes from trade liberalization. Attempting to protect an industry from foreign competition via tariffs may inadvertently incentivize workers to enter the industry. Once the tariffs disappear (or foreign competitors get even better), a similar trade shock to the NAFTA shock may occur again, resulting in a mortality rate jump. Thus, ‘closing’ to trade will not undo the negative impacts of opening to trade, but will likely repeat them.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. When People Cut Back on Instagram, Where Do They Go? – discusses a recent research experiment on how people use the time freed up by turning off Facebook and Instagram.

2. Reflections on Vibe Researching – talks about using AI for research. This paragraph summarizes it very well:

My point is that the experiment – can we do research at high speed without much human input – was a failure. And it wasn’t just a failure because LLMs aren’t yet good enough. I think that even if LLMs improve greatly, the human taste or judgment in research is still incredibly important, and I saw nothing over the course of the year to suggest that LLMs were able to encroach on that advantage. They could be of great help and certainly make research a ton more fun, but there is something in the judgment that comes from research experience, the judgment of my peers and the importance of letting research gestate that seems more immutable to me than ever.

3. Short Note from me:

Subscribe to Our Newsletter

Recently, The Georgia Department of Transport posted on their social media account the following statement with the below picture:

The widening of I-95 near the Savannah River will improve travel for thousands of drivers each day while supporting the continued growth of the Port of Savannah –one of the fastest-growing ports in the nation. A smarter, safer corridor is on the way. https://bit.ly/4ux3vjW

The picture suggests that widening the road by 1 lane will significantly reduce congestion and improve commute times. Economics research tells us that this extra lane will have no impact on congestion whatsoever.

The Fundamental Law of Road Congestion

Economist Anthony Downs, in 1962, posited the Law of Peak-Hour Traffic Congestion:

This Law states that on urban commuter expressways, peak-hour traffic congestion rises to meet maximum capacity.

Over time this law was rephrased as the Fundamental Law of Road Congestion. Economists have often studied this law to see if it actually holds, and research has shown this to be the case.

Evidence from US Cities

Duranton and Turner (2011) investigated whether the Fundamental Law of Road Congestion holds in the context of US cities. The main question Duranton and Turner were interested in answering is what impact would changing the number of lane kilometers (i.e. increasing the number of lanes on road) have on vehicle-kilometers traveled (VKT). VKT is a measure of the total number of kilometers traveled by vehicles on a given highway system. If one vehicle travels 1 kilometer, that would be considered to be 1 VKT.

Suppose a one lane highway is 100 kilometers and 10 cars travel it in its entirety. This would be a road with 100 lane kilometers (1 lane times 100 kilometers) and VKT would be 1000 (10 cars times 100 kilometers). If we add an extra lane to this highway, the overall distance does not change, so 10 cars would still travel 100 kilometers, resulting in 1000 VKT. Lane kilometers would double since we have two lanes (2 times 100 kilometers). If VKT remains at 1000, while lanes doubled, then we would consider this road to be less congested. On the other hand, if the extra lane results in 20 cars on the road (2000 VKT), then the additional lane will have no impact on congestion.

Data and Results

To estimate the impact of adding lanes, Duranton and Turner looked at 228 metropolitan areas in the US over a 30 year time period, between 1983 to 2003. The number of vehicles on roads in the US is collected quite accurately – for example, states must provide accurate daily interstate highway usage to the federal government.

Overall, VKT in the US doubled from 1983 to 2003, from 7,700 VKT to 15,900 VKT for interstate highways. For major urban roads, a similar pattern occurred – VKT went up from 15,000 to 30,000. In the same time span, the number of lane kilometers of interstate highways remained basically constant, while on major urban roads, it went up from 3,800km to 6,500km.

It is important to note that in order to determine the impact on VKT of adding additional roads (causality), we have to ensure we are not capturing other factors. For example, overall population growth will naturally result in higher VKT and it may seem like adding roads has no impact on congestion. Duranton and Turner do control for many factors and even use an innovative approach called “Instrumental Variables regression”.

So what is the impact of increasing the number of lane kilometers on VKT? Duranton and Turner estimate that increasing lane kilometers by 1% will increase VKT by 1.03%. Economists call this an elasticity1 of 1.03. Similarly, a 10% change in lane kilometers will increase VKT by, approximately, 1.03 times 10%, so 10.3%.

The Duranton and Turner result tells us that any feasible increase in roadways will have no impact on congestion – any increase in roads will be met by more vehicles on the road, more than offsetting the increase in roads.

Where Do The Extra Vehicles Come From

Duranton and Turner were also interested in figuring out what is the source of extra vehicles. They looked at four potential sources:

• Additional commercial traffic;

• People and economic activity migrating to places with additional roads;

• Diversion of traffic from smaller, local roads;

• Changes in individual behavior with people choosing to drive more.

Duranton and Turner studied how a hypothetical 10% increase in lane kilometers in interstate highways, that increases VKT by 10.3%, would be apportioned between the above 4 causes.

Duranton and Turner found that a 10% increase in interstate highways results in a 10 to 20% increase in truck road usage. As trucks account for approximately 13% of all vehicle traffic, increases in commercial truck traffic explains about 20% to 30% (thus, 2%-3% of the 10.3% increase) increase in VKT due to increased interstate highways.

Individual behavior is also a major contributor to increasing VKT. A 10% increase in interstate highways results in about a 1% increase in road usage by each individual driver annually. Although this may not sound like much, the largest source of VKT on the roads are individual drivers (87% of all VKT). Thus, a minor increase in car usage by each person, results in a significant increase of cars on the road. Individual behavior changes account for approximately 9% to 39% (1%-4% of the total VKT increase) of the increase in VKT due to additional interstate highways.

The other two reasons appear to have much less impact on congestion. Population growth from migration is minimal – a 10% increase in interstate lane kilometers increases population in the regions around the interstate by 2% over 20 years and can explain only around 5% to 20% of the increase in VKT. Diversion from other smaller local roads is almost non-existent – only 0% and 10% of the increase in VKT.

These results suggest that construction of additional roads have limited positive impacts – it does not reduce congestion on any roads nor does it have any meaningful increase on population growth, which would translate to economic development.

The increase in commercial vehicles could be a positive economic sign (more goods being transported) but it depends whether this increased truck traffic is driven by increased economic activity or a substitution from other forms of goods transportation like trains.

The main cause for increasing congestion when new lanes are built is that individual drivers choose to drive more. This fact has been termed often as “induced” or “latent” demand. Any time a new road is built or an old one is expanded, it quickly ends up getting congested, without any relief of congestion on other roads. However, the people undertaking these ‘new trips’ must have been doing something before the road expansion. They either used an alternative mode of transport or did not make this trip. This ‘pent-up’ demand keeps congestion constant, as any road expansion will immediately be filled up by cars.

Interestingly, this also means that some solutions that might work – won’t. Duranton and Turner looked at whether increasing public transit can reduce congestion. By studying the provision of buses, Duranton and Turner found that public transit has no impact on VKT. This is consistent with induced demand. Even if people shift some of their car trips to public transit, there are ‘new trips’ ready to take their place.

Duranton and Turner also found that, over-time, regions slowly converge to having the same average daily number of vehicles per lane. This suggests that there is a ‘constant’ level of congestion all roads tend to. People appear to migrate out of regions that have higher than average congestion levels, to areas that are lower, resulting in the convergence of congestion.

Solving Traffic

Research shows that expanding roads will not alleviate congestion. So what will? The only solution demonstrated to have had impacts on congestion is congestion pricing. Congestion pricing internalizes the externality2 caused by using a vehicle. The externalities caused by car usage include increased travel times, increased air and noise pollution, as well as increased probability of accidents. These costs are borne by both drivers and non-drivers. By putting a monetary cost on the travel, we internalize this cost, as the driver now pays for the costs that car usage entails. Research shows congestion pricing does reduce car usage – a holistic overview by Lehe (2019) has shown that congestion pricing does reduce the number of cars (depending on the implementation) and that people are very responsive to the additional congestion cost.

In the US, New York City implemented congestion pricing last year, and, as expected, the benefits have been significant, including a fall in commute times and pollution.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Even Before the Iran War, There Was a Growing Inflation Problem – goes over the inflation picture in the US, and why inflation had been rising even before the Iran war.

2. In view of the above post, March inflation, as being forecast by the Cleveland Federal Reserve, may be very high – at 0.62 percent (normally it should be at 0.15%)

   

3. The Dirty Secret of Election Models: explains why judging election models based on hindsight (i.e. after the result) isn’t helpful for picking good models. What matters more is the transparency and validity of assumptions:

All models have assumptions. There’s no such thing as a “purely data-driven” model – only models whose assumptions are obvious…

The difference between a good model and a bad one isn’t whether assumptions exist. It’s whether those assumptions are valid, explicit, and aligned with the real-world process being modeled.

Subscribe to Our Newsletter

Governments provide various assistance programs to address social issues such as poverty, poor health and poor educational outcomes. Many of these policies often provide certain financial benefits. One would expect uptake of these programs to be near perfect – why wouldn’t a person want ‘free money’. But, often, uptake of such policies is mixed.

A recent study by Yu, Stolove, Riddell, and Mahumane (2026) (“YSRM”) on HIV medication adherence may explain why these theoretically sound policies may have underwhelming results in practice.

HIV Medicine Adherence Incentives

YSRM conducted a field experiment in Mozambique to test several hypotheses around HIV medication adherence.

HIV Medication Adherence

Anti-retroviral therapies (ART) can keep the effects of HIV at bay, allowing HIV positive individuals to live a full life. Typically, as part of ART, an individual needs to take a daily pill. Missing a pill can severely weaken the benefits of the therapy, and a 95% adherence rate is needed for ART to be successful. In Mozambique, although ART is highly available and free to people, adherence is generally low – in the YSRM study, base adherence was about 22%. So what can be done to improve adherence?

The Experiment

YSRM enrolled eligible participants – individuals 18 years or older, living with HIV, who have not been on ART recently and who have a smartphone (the smartphone was needed for payments and medication reminder calls). As part of the ART, individuals were given pills for 30 days, and told to come back for a refill before they ran out of pills.

Individuals were then split into several groups:

1. Control – this group was not given any additional incentive;

2. Financial Incentive – this group was given a payment if they refilled their medication on or before the refill date (30 days after first pill);

3. Reminder – this group was given a call 0 to 6 days before their expected refill date to remind them of refilling their medication;

4. Combination – this group was given both treatments of the Financial Incentive and Reminder groups;

5. Information – this group was given a short video explaining the risks of HIV and importance of ART;

6. Stigma-Relieving – this group was told about how their local community perceives living with HIV, and that, in most cases, individuals overestimated the stigma associated with living with HIV.

Behaviors Tested

The purpose of these hypotheses was to assess the interventions and also to test two behaviors:

• Inattentiveness; and

• Forgetfulness (or Imperfect Memory).

Inattentiveness occurs when individuals do not take the pill every day, as they are not paying attention to daily pill taking.

Forgetfulness refers to the fact that individuals may not remember when they are supposed to go get the refill (no later than 30 days after the first pill). Individuals can guess they may need to refill their medicine based on the number of pills they have remaining, since they had exactly 30 pills to start. However, they also might not remember how many days they didn’t take the pill.

Results – Group Averages

The main outcome YSRM looked at is the Medication Possession Ratio (MPR) – the ratio of total number of pills possessed over the total number of days in the study (since for 100% adherence, you need one pill per day). As mentioned above, ART works best with 95%+ adherence, so for example, an individual will need to have (and take) at least 95 pills for 100 days, which would give an MPR of 95% or more. In order to satisfy that, an individual needs to ensure they refill their medicine on time.

The chart below shows the percentage of individuals that maintained at least 95% MPR/adherence under the different treatments:

After 180 days, the adherence rates (i.e. MPR of >95%) were:

1. Control – 22.6%

2. Financial Incentive – 31.8%

3. Reminder – 31.7%

4. Combination – 47.4%

5. Information – 28.3%

6. Stigma-Relieving – 35.7%

The Combination group clearly performed better than other groups in the raw data.

What Improves Adherence?

Using a regression-based analysis to get more granular impact, YSRM found that financial incentives increase adherence by about 10.1 percentage points. Reminders increase adherence by approximately 11.6 percentage points. The Combination increases adherence by 24.3 percentage points, which is a bit more than the two interventions additively.

The fact that the Combination result is greater than the sum of the two interventions is actually an important result. It means that the two interventions are valuable on their own. Often, each additional intervention has less of an impact when combined with other interventions, since the interventions are targeting the same population. If one intervention leads to adherence, giving a second intervention would be wasteful. Thus, typically we should not expect two interventions to have an additive impact.

Why Do People Not Adhere to ART?

YSMR were also able to determine that inattentiveness and forgetfulness do occur. The Reminder group showed increased adherence by showing up for medicine refills even before the reminder call was made! This means that just the knowledge of a reminder call meant individuals were paying more attention to taking pills and thus, saw based on their pill count, that they needed to refill their medicine.

With regards to forgetfulness, YSMR noted that in the Financial Incentive group, ART adherence increased mainly by individuals who thought they were refilling their medicine before the due date, when in fact they were refilling after the due date (and thus did not receive the cash incentive). This tells us that the financial incentive made individuals think 30 days haven’t passed yet, when in fact 30 days did pass. Thus, the financial incentive acted as a type of reminder that one ought to go to get a refill.

Implementing Policies – Dealing With People’s Behaviors

Free ART on its own is a really beneficial policy – both for individuals and society. Uptake of this policy is, however, quite low, even though it’s extremely beneficial to the individual. Thus, another policy is introduced – financial incentive for ART adherence. Again, the impact of this was relatively muted. The YSMR study shows us that this lack of adherence is driven partially by human behaviors – inattentiveness and forgetfulness. Thus, what may look like a ‘failed’ policy, is actually a policy implementation issue. Adding reminders significantly boosted the adherence rate.This finding has also been established in other research. Banerjee et al. (2025) conducted a large-scale experiment in India on immunizations and found that a policy mix of financial incentives, reminders and outreach ambassadors increased immunization by 44%.

This has implications for many similar policies – especially financial incentives. Many policies are usually based on some form of financial incentive – lowering tax, increasing a transfer, providing a voucher or pre-loaded debit card. Take up of policies can fail as people are inattentive to repeated tasks or forget deadlines for applying for these programs. YSMR shows that it can be relatively cheap to improve uptake of these policies by adding a reminder mechanism such as a call (or maybe an app notification).

Governments on every level may find it beneficial to focus more on connecting citizens to already existing programs, rather than creating new programs or policies to address issues. New programs may similarly suffer from low uptake of old programs, as they do not solve the issue of inattentiveness or forgetfulness. Instead, focusing on outreach or incentives that reminds citizens to existing programs may be much more effective. The creation of such civilian outreach programs has been shown to have large positive impacts on communities, as citizens are made aware of what’s available to support them.

Let me know what else you think reduces the effectiveness of policies in the comments below.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Mind the Failure Gap: discusses research about how people underestimate how often people and systems fail. This has implications for policy, especially in how people may view punishments for infractions.

2. Nobody could tell me what they’d pay for a robot: asks an interesting question, to which the answer is not as easy as it seems – how much would you pay for a robot that can do house chores? Let her know your response.

3. The Bureau of Economic Analysis (BEA) made a change (without announcing it) to how one of the inflation measures is computed. It lowered reported inflation by 0.1 percentage points. The lack of information about this change, however, is concerning:

Subscribe to Our Newsletter

Venture capital (VC) – a form of financing high risk/high reward early stage firms – has led to the creation of many large companies like WhatsApp, eBay, Uber, etc. Although VC investments can be highly lucrative, one of the main concerns of investing in early stage startups is the alignment of interests between investors and company founders, commonly referred to as the ‘principal-agent problem’. VC investments alleviate this problem through complex contracts1 between investors and founders. However, since the 2000s, changes in these contracts may have led to an increase in founder fraud. Dyck, Fang, Hebert and Xu (2026) (“DFHX”) took a closer look into this.

What is Venture Fraud and Why Should We Care?

Venture fraud occurs when the company, most often its founders, deliberately conducts fraudulent activity that harms the VC investor. This is often done when founders misrepresent financials (Spartan Micro), claim to have more customers than they do (Frank vs JP Morgan), or lie about what their product does (Theranos).

One could argue that venture fraud is a private matter since most damages occur between two private parties – investors and founders. There may even be an ‘optimal’ level of fraud, since to perfectly prevent fraud, investors would need to spend a lot of money and time to ensure fraud doesn’t occur, or simply undertake fewer investments. However, the effect of venture fraud may spillover more broadly into society, which then makes it a topic of interest.

The Case of Theranos

Theranos claimed to have designed a compact blood testing machine that, using a small amount of blood, would give quick results. Several US pharmacies planned to open in-store rapid blood testing using this product. However, it turned out that the machine was entirely fraudulent. This ended up hurting patients, who were misled by the tests. The damages to patients, as well as the wasted time/capital of pharmacies are specific social damages – commonly known in economics as negative externalities.

Investors, beyond their own financial windfall, do not take into account possible negative externalities – i.e. they do not internalize these negative externalities. This means that from a society perspective, venture fraud is more damaging than to the investor, and therefore, too much venture fraud occurs.

Preventing Fraud

The main way investors prevent fraud from occurring is by having control in the firm they invest in. This is usually done via having voting rights in the board of the firm. In this role, investors can vote on changing CEOs (who are often the original founders), as well as setting compensation structures for the CEOs. Company boards also approve or block key business decisions. Thus, investors may often want to have control of the board (i.e. more than 50% of the voting rights). But in the last few decades, this investor control appears to have reduced.

Causes of Venture Fraud

The Theories

DFHX posit three different hypotheses on what may make fraud more likely to occur:

• Founder-friendly board structures – the original founders have more voting power in the boards of their companies, allowing them to mask fraud;

• The number of investors in a firm has increased – each new investor in a board may have a different goal (company growth vs. cash distributions to investors) leading to more board disagreements, which in turn reduce the quality of oversight in a firm, allowing fraud to occur;

• Founder characteristics – certain traits of founders, whether innate or developed over time, can lead to fraud occurring more often.

DFHX also looked at whether market conditions mattered for fraud. DFHX hypothesized that under “Hot Markets” – if macroeconomic conditions are such that firm valuations are higher, board control may be weaker, as investors are not as focused on company control. Moreover, there may be more investors in the market, competing for a limited pool of founders, which in turn allows founders to have friendlier board structures.

Data Collection

To test the theories, DFHX turned to data. The data collection process here is actually quite interesting, as there is no simple data set that can be easily looked into. DFHX looked at several sources – Securities and Exchange Commission (SEC) and US Department of Justice (DOJ) official filings, a Class Action lawsuits database, the Westlaw research platform and the Pitchbook investments news data feed.

The main issue for DFHX with data collection is the concern of not capturing all venture capital fraud events. This can skew the results if, for example, only specific venture fraud cases are reported. That is why DFHX looked at multiple databases to capture as many fraud cases as possible. Overall, DFHX built a sample of 614 US based venture fraud cases since 2000.2

Now, to study if venture-capital backed firms are more or less likely to have fraud cases, DFHX looked at a sample of 4,094 VC-backed firms and non-VC-backed firms that had an Initial Public Offering (IPO). IPOs (being listed on the stock market) have stringent disclosure requirements; often, fraud gets discovered in this phase. The fraud discovered during the IPO (or soon after it) can be assumed to have been caused by the company’s management team (i.e. often the founder) prior to the IPO.

Separately, to study what may be driving the differences in fraud level amongst venture capital firms, DFHX looked at a sample of 78,852 VC-backed firms in the 2000-2020 time period.

Predictors of Venture Fraud – Results

DFHX found that VC-backed firms that IPO’d are 54% more likely to have committed fraud than non-VC backed firms. To put it into context, about 8% of firms that IPO’d had a fraud lawsuit, while the rate for VC-backed firms was closer to 12-13%.

When looking specifically at VC-funded firms (not necessarily IPO’d), DFHX found that companies where the founder has control over the board are 88% more likely to commit fraud. Moreover, every additional investor on the board increases fraud likelihood by 8%. If the investors are less involved in oversight, like mutual funds or hedge funds who do not actively participate in board meetings, fraud is also more prevalent.

When looking at founder characteristics (age, gender, education, etc.), DFHX found that these did not matter for the likelihood of fraud. What did matter is whether the market was ‘hot’ when the VC funding round happened. That’s most likely driven by the fact that investors offered founder-friendly terms during these times, as there were more investors chasing fewer founders.

Lastly, one additional question DFHX looked at is whether fraud-involved founders are ‘disciplined’ by the market. Typically, one would assume that if someone committed fraud, this person would be less likely to receive investment in the future. Surprisingly, DFHX found that there is no market discipline – founders that committed fraud are just as likely to receive investment in the future as founders that did not.

Dealing With Fraud

The DFHX paper has several interesting implications. First, it shows the importance of the often-studied topic in economics of the principal-agent problem. The reduction in oversight by investors appears to have led to a real increase in fraud by founders (shown in next section below). It also shows the weakness of market mechanisms, as it appears that the market is not discipling fraudulent founders. This is a surprising result, as a common argument against regulations is that market actors will self-discipline. Here, it appears that investors are willing to take unnecessary losses.3

Regulatory Scope and The AI Boom

If we look at the VC market today, it appears that we’re in a ‘hot’ genAI market. Based on the DFHX research, this suggests that in the future, we may see quite a bit more fraud than usual associated with these VC-backed investments. As shown here by Ewens and Malenko (2020), founder/entrepreneur control has gone up over time at every stage of the company’s financial lifecycle in recent years (lines specify which round of fundraising it was):

It’s worth noting that, historically, founders rarely had 50% control after the first round of funding, as can be seen in the middle chart.

It is worth noting, that in other contexts, such as in the Private Equity-backed companies, DFHX found that fraud has not increased, and actually declined, clearly showing that oversight by investors, specifically in the VC space, is lacking:

From a policy perspective, it may make sense for governments to enhance oversight over VC-backed companies by potentially increasing financial reporting requirements, as VC investors appear to be more tolerant of fraud than socially optimal. This suggests that there may be scope for regulations to correct this market inefficiency.

Let me know your thoughts on this topic and whether you think intervention may be needed to reduce the amount of fraud.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. A Simple Idea That Could Change Things for Kids: Child Impact Statements – provides an interesting idea of requiring policy changes to have a ‘child impact statement’: that is, an explicit statement on how this policy change will impact children.

2. Does Where You’re Born Matter More Than How Hard You Work? – discussed why it appears that the physical location where one grows up explains a large amount of a person’s outcome: “Roughly 60 percent of the variation in outcomes across neighborhoods reflects genuine causal effects of place – not just differences in the families who happen to live there.”

3. Note: An economics professor argues that economists should not come up with solutions. That is not something I agree with – as discussed in our article a few weeks ago.

Subscribe to Our Newsletter

Financial well-being, especially of children, is an issue that many countries grapple with. US President, Donald Trump, announced a policy in which newborns born between 2025 to 2028 will receive an investment account with $1,000 in it provided by the US Treasury (i.e. US tax payers). The money in the “Trump Account” can be invested in certain stock market funds, but the funds themselves can only be accessed after the child turns 18.1

Now, on paper, this policy sounds good, as we’re boosting savings for families with children. But, upon closer inspection, this is a policy that does not actually help much at all – and it’s not because the amount of money invested is small. Let’s dive in.

Alleviating Child Poverty

Before we start assessing the Trump Account policy, we should first determine what it is that we want to achieve with the policy. A policy is just a tool to achieve a goal. The goal, at least to me, is to improve outcomes/welfare for children and, perhaps, teach children and families to save/invest via exposure to markets.

Child poverty in the US is a serious problem. Around 12%-14% of children live in poverty2:

On the above chart, you may notice that, in 2021, there was a dramatic fall in child poverty, which then rebounded to the current highs. The reason child poverty fell in the US was thanks to the numerous government programs helping low income families, of which the 2021 Expanded Child Tax Credit3 (eCTC) was the most important. We’ve discussed this policy extensively here, and argued that it is a policy that pays for itself. The eCTC was a 40 times larger program than “Trump Accounts”, but the reason it worked had to do more with the fact that it immediately alleviated issues for families in need. Per every $100 a family received:

• $28 was spent on food;

• $31 on housing; and

• $15 on child related goods and services.

On the other hand, a savings account will have none of this impact, as the money is locked until age 18.

Some might argue that these families who receive the “Trump Account” could increase consumption by saving less in other accounts. The problem with this argument is that families in poverty do not have savings, and thus they cannot ‘tap’ into other savings.

Therefore, for low income families, this policy will not have any meaningful impact, at least for 18 years.

Shuffling Savings

Turning to non-low income families who already may have their basic needs met, “Trump Accounts” could in theory benefit them by increasing and encouraging more saving. However, as these families may already have savings, it is unclear whether their overall savings will increase by the full $1,000 amount. That’s because a family may choose to save less in other accounts (savings accounts, personal brokerages, etc).

Research on other types of tax-advantaged accounts (i.e. saving accounts that have certain tax benefits) have shown that adjusting savings is common. Benjamin (2003) showed that when a person becomes eligible to use a new tax-advantaged retirement saving account, only around half of the funds that are put into the new tax-advantaged account are truly new savings. The rest is simply shuffling savings from a different account. Other studies found even lower percentages – Fehr, Habermann and Kindermann (2008) showed that only 22% of funds in tax-advantaged accounts are truly new savings, while Pence (2001) argued that the percentage may actually be 0%.

Regardless, this tells us that many families will simply adjust their other savings in response to the “Trump Accounts”. Intuitively, it makes sense. Suppose you have a rule of thumb that you save 30% of your income. If you receive an unexpected $1,000 in income, you may end up spending $700 more now to maintain the same ratio of consumption to saving.

Therefore, the “Trump Savings” account will not increase saving by a significant amount. Furthermore, in effect, tax payers will be subsidizing the consumption of higher income families, as the funds in the “Trump Accounts” come from tax payers.

Negative Impact on Recipients

While researching for this piece, one thing that surprised me is that there may be “Trump Account” recipients that can end up worse off. This surprising outcome is driven by how different government programs interact. Dynarski (2003) showed how similar tax advantaged accounts – the 529 savings plan, which allows families to save money tax free to be used to cover children’s college costs – can end up harming users. That’s because the US also gives college applicants financial aid that depends on the financial situation of a family. Thus, if you save enough, you may no longer be eligible for financial aid, resulting in no benefit whatsoever from saving.

Although this may be somewhat of an extreme case, it’s worth noting that not all the accumulated savings in “Trump Accounts” may end up truly staying with the recipient.

Financial Literacy

So is there an aspect where this policy may have larger positive impacts? One way “Trump Accounts” may be helpful is that it exposes people to saving and stock market investing. This in turn could lead to more saving/investing in the future, as children exposed to savings/investments will become adults that are more financially literate.

The question therefore is – does having an account from birth lead to children and their parents becoming more financially literate? The evidence on this is mixed (and the amount of research on this topic is limited). There have been similar programs enacted at the state level in the US – for example, the SEED Oklahoma program, where several thousand children received a $1,000 savings account at birth that was to be used for college. Results were mixed, as several analyses showed positive impacts (more likely to go to college, higher savings), but under closer inspection, a lot of these outcomes were driven by wealthy families.

An Experiment

A potentially better method to improve financial literacy is to actually teach children in school. Horn, Jamison, Karlan and Zinman (2023) (“HJKZ”) undertook a financial literacy experiment in Uganda. HJKZ took 2800 Ugandan youths and split them into 4 groups:

• Group 1 was given no intervention;

• Group 2 was given access to a savings account at a bank;

• Group 3 received financial education; and,

• Group 4 got the account and financial education.

The outcomes that were studied included – overall financial literacy, trust in financial institutions and the amount that was saved. One year after the intervention, the two groups that received education showed higher financial literacy and trust in financial institutions, while the group that only got the bank account showed no change.

With regards to saving, a similar pattern occurred – groups receiving financial education had higher savings (Note: since the study had a relatively small number of participants, HJKZ could not conclusively reject that the group that only had the account saved less than the education groups).

HJKZ followed up on this study five years later. Surprisingly, the financial literacy of the two education groups reverted to pre-finance education levels, although trust in institutions remained. Savings behavior remained similar to what happened after one year (higher savings than the group that received no account/no financial education, but unclear whether higher than the group that just received the account).

Overall, with the caveats that this is just one study, it appears that any intervention (whether account opening or education) has some positive impact with increased saving. In general, financial education appears to be slightly better than just having an account. Either way, the cost of either helping youths open bank accounts or provide education is relatively low, and thus it may be beneficial to conduct these interventions either way.

How To Improve the “Trump Account” Idea

I will first note that the “Trump Accounts” program is a small program. With about 3.5mln births in the US per year, the direct cost would be $3.6bln annually (for comparison, the Child Tax Credit program currently costs $128bln).4

However, we should still evaluate the program based on whether this is an effective spend of money. I feel quite comfortable saying “no” – the “Trump Accounts” are a poor spend of money. If the aim is to improve outcomes for low income families, then direct unconditional assistance (such as the Child Tax Credit program) is superior. If the aim is financial education, then a better program could easily be designed at the same cost.

For example, opening savings/investment accounts for school children coupled with a financial literacy course would probably be better. These accounts could also start with a government investment, albeit at a lower amount like $100. Moreover, since financial literacy dissipates, perhaps, refresher courses every 3-4 years would be beneficial. As part of completing these courses, students could get another $100 added to their account. This would encourage participation in these financial literacy classes. Students exposed several times during their studies might come out far better equipped.5 This policy would likely be much more effective and cheaper than the “Trump Accounts”.

Financial literacy and saving is an issue faced by all societies. Although today we focused on the US (and Uganda), I believe these findings are applicable globally.

Let me know your thoughts on these accounts and what you think would improve financial literacy.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Stop looking at economic averages! – discusses why focusing on average statistics does not paint the whole picture of the economy – distributions matter.

2. Political journalism needs fewer “takes” and more analytical legwork – highlights the current problems with political journalism, which is similar to what we see in economics.

3. Prediction – Jason Furman asked people what they think average annual productivity growth will be in the next decade, when taking into account AI. He provided a table of 5 possible outcomes, and asked readers for their probabilities. Here is Nominal News’ forecast in red:

   

   Generally, I do not expect any major surprises. I do worry that there might be an ‘accidental’ recession that could happen due to misuse of genAI (similar to the financial crisis of 2008 and the misuse of collateralization calculations). My main reason for forecasting a solid breakout is that energy costs may drop in the next decade allowing for greater production + genAI lowering the cost of trying out new ideas, especially in early phase applied research.

Subscribe to Our Newsletter

Friday, February 20, 2026 was a big economic news day in the US.

1. The US Supreme Court deemed President Donald Trump’s tariffs issued under the International Emergency Economic Powers Act (IEEPA) as illegal.

2. The Bureau of Economic Analysis (BEA) published US fourth quarter Gross Domestic Product (GDP) data which showed that real GDP growth was 1.4% (while various official forecasts were suggesting between 2.5% to 3.5% growth).

3. The BEA also released the Personal Consumption Expenditure Price Index (PCE – the Federal Reserve’s preferred measure of inflation) for the month of December, which showed inflation at 0.4% (month-over-month) and 2.9% (year-over-year).

So how do all these pieces of news fit together?

Tariffs – Soon to Disappear?

The decision by the Supreme Court to deem the tariffs as illegal was widely expected. But this doesn’t mean the end of tariffs, as President Trump quickly reinstated tariffs under a different section of US law. The newly announced tariff rate – i.e. tax rate on imported goods – is 15% on all imports, regardless of origin. These new tariffs can only be in place for 150 days until Congress must approve them through a vote. At the same time, the legality of these tariffs is also being questioned.

The exact impact of the recent changes in tariffs is still ambiguous. Some are forecasting a higher effective tariff rate than was currently present (the effective tariff rate is computed as the total amount paid to the government by importers over the total amount that is imported). On the other hand, the Yale Budget Lab estimates that the effective tariff rate will be lower. Nonetheless, there will definitely be a change in the distributional impact of these tariffs, since the new tariffs are placed on all countries, while the previous tariff regime was targeted at different countries.

Tariff Refunds?

One significant development emerging from the Supreme Court decision is whether the tariffs collected in 2025 will need to be refunded to the importers. From an economic perspective, if the tariffs are refunded, it may lead to the net impact of the tariff policy being a large transfer from consumers to firms.

To understand why, we first note that tariffs are simply a tax on imported goods paid by US based importers. This tax, however, is ultimately paid by local consumers since importers typically increase the price of these imported goods in the US to recoup the tariff (tax) they had to pay. Now, the benefit of this tariff (tax) is that the government (i.e. US citizens) gets the revenues. With the Supreme Court ruling, if this collected revenue gets refunded, it will go back to the importers, since they were the one that paid the tariff tax originally.

Therefore, the net impact of these IEEPA tariffs with the Supreme Court ruling could be a de facto transfer from local US consumers to US importing firms, making it a very regressive policy (transferring from poorer individuals who had to pay extra for goods to richer individuals – owners/shareholders of importing companies).

However, it is worth adding that even absent these potential refunds, the tariffs have adversely impacted GDP growth and increased inflation.

GDP Growth – Negative Impacts of Tariffs

US fourth quarter real GDP growth came in at 1.4%, which was lower than various forecasts suggested. It is worth adding an important caveat. One of the reasons real GDP growth came in lower than expected was due to the government shutdown that lasted 43 days (October 1 to November 12, 2025).

Aside – Government Services Adjustment

Government-provided services are part of GDP. During the government shutdown, the government stopped providing most services, so in calculating total GDP, it can be assumed that the government provided 43 days less worth of services. But at the same time, government employees were still paid for these 43 days. Therefore, during the fourth quarter, the US government provided fewer services, but it cost the exact same amount as in other quarters during the year.1 Thus, although nominal GDP came in at 5.1%, real GDP came in at 1.4% since the amount of real government services was 43 days fewer.

Absent the government shutdown, the fourth quarter real GDP growth would have been closer to 2.5%. This ‘missing’ real GDP will reverse in the first quarter of 2026, as the government will return to a normal level of government service provision.

Overall 2025 GDP Growth

For 2025, real GDP growth was 2.2% (adjusting for the government shutdown, it would have been 2.5%). In 2024, real GDP growth was at 2.8%. Last year, we wrote an article talking about research on the likely impact of tariffs, summarized in the table below:

The above table shows that tariffs were estimated to lower real GDP by around 0.75% in the period after tariff enactment. Interestingly, this means that absent tariffs, real GDP growth in 2025 would have been around 3.3% (after adjusting for the government shutdown and adding the impact from the table), which would be a bit higher than 2024 real GDP growth.

Although this high level comparison is not definitive, it does appear that these initial estimates in the Kalemli-Ozcan, Soylu and Yıldırım (2025) have captured reality well. Even the inflation forecast seemed quite good.

US Inflation

After the release of the December data, US PCE measured inflation for the year was 2.9%. Using the estimate in the table above, absent tariffs, inflation would have probably been around 2.4% (target inflation is ~2%). Some data suggests that inflation may have been even lower absent tariffs. The image below shows a different estimate:

Since March 2025, Consumer Price Index (CPI) measured inflation (a slightly different inflation measure) would have been 1 percentage point lower had tariffs not been enacted.

Tariffs and the Economy

The recent developments show us that tariffs (a domestic tax on imported goods) are really not a good policy. So far, tariffs:

• Led to higher inflation by around 0.5 pp to 1 pp;

• Reduced GDP by around 0.3%-0.5%; and

• May end up being a direct transfer from poorer consumers to richer importers.

I think it’s safe to say that these are not generally desirable outcomes from a policy perspective.

The only positive from this tariff policy is that it has shown that academic economists are pretty good at modelling the impacts of such a policy, and, perhaps, they should be listened to a bit more.

To read more on our coverage of tariffs:

• Yes – Tariffs Have Increased Prices and Inflation in the US (June 15, 2025)

• Modelling the Impact of Tariffs (April 30, 2025)

• Why Import Tariffs are Identical to Export Taxes (April 13, 2025)

• Do Tariffs Cause Inflation? Yes. (March 24, 2025)

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Trust in Numbers – tackles the concerns around the reliability of US government statistics, showing that the statistics are currently reliable, although funding cuts may make collecting data harder.

2. Removing Truth & Sanity from American History (United States) – conducts an interview with Michiko Quinones (Co-Founder and Director of Public History for the 1838 Black Metropolis Project), where they discuss the re-writing of history via the recent removal of display panels from the President’s House in Philadelphia, which included information about the role of the U.S. slave trade in the development of the country.

3. America’s Pensions Can’t Beat a Vanguard Account Or Finance a Transmission Line, But They Can Close Your Hospital – discusses how pension funds can invest in important projects for society (energy projects, additional housing). Pension funds in other countries often do this, while US pension funds appear to be less efficient with their use of capital, spending a lot more on fund managers.

Subscribe to Our Newsletter

On February 4, Jon Stewart – a well-known American comedian/writer/political commentator – interviewed Nobel Prize-winning economist Richard Thaler. The discussion touched on several economic topics like healthcare and climate change. The interview itself was relatively inconsequential, but there was significant fallout post the interview with numerous economists and economics commentators criticizing Jon Stewart harshly for what was viewed as very dismissive comments about the field of economics.

I listened to the interview and I see it as a classic case of economists and non-economists talking past each other. But I don’t think the avalanche of criticism against Stewart heaped on by economists was justified.

Economics in Trouble

I am a PhD-trained economist, and I believe economics is both an important and useful subject. Economics studies human decision-making; these insights are invaluable for policy discussion. I also believe that the majority of PhD-trained economists conduct research in good faith, with the aim to gain a deeper understanding of human behavior, which often results in better policy making.

In the interview during the a discussion on climate change policies, however, Jon Stewart said the following about economists:

“So my point is, since when is economics about improving the human condition and not just making money for the companies that are extracting the fossil fuels from the Earth?”

Stewart’s view of economics is starkly different from mine. It is an incorrect view, but the reason Jon has that view is because, as a field, economics has failed to accurately distinguish what we actually do. Unfortunately, the blame for this should fall on us, economists, as we have failed to properly demonstrate what we do, allowing for others to use economics as cover for their opinions. Let me explain.

Economics Research, Policy and Politics

In the interview, Stewart and Thaler focus a lot of their discussion on climate change and reducing pollution. Stewart asks about what economists would recommend to deal with climate change, to which Thaler responds by mentioning carbon taxes as the solution. Stewart says that such a policy would never pass.

We have several issues occurring in this exchange, which is predominantly driven by talking about different things. Thaler responds to Stewart’s question by proposing a very good policy (a policy I also would personally support). Stewart, however, changes the question by bringing in a constraint – he’d like to hear a policy that would be politically feasible. Stewart was a bit vague with his initial question, which caused the confusion.

Economists, however, can tackle both of Stewart’s questions. If you, as a personal preference for a social outcome, want to see less pollution, economists may recommend the carbon tax policy. If you add a constraint to this preferred outcome – for example, I want less pollution but I don’t want taxes – economists may provide a different policy.

Basically, if you have a preference for a social outcome for anything involving people such as less pollution, more technological innovation, more factories, etc., economists can give you a list of policies that would achieve the stated goal. Moreover, each policy would have a particular cost associated with it. In the climate change case, it may look like this:

• Carbon Tax: short/medium run things will be more expensive, which may have distributional consequences;

• Subsidy: taxes need to be raised to fund subsidies; it is difficult to set subsidies to achieve desired goals;

• Regulations: costly to enforce (need to create a new body); regulation evasion can be significant.

Armed with this, a politician/policy maker/voter can pick their preferred policy. None of the policies above are wrong, as long as it is aligned with the preferred outcome of a politician or individual. Economists, in their capacity as experts, should focus on how to best achieve your stated goals/outcomes (we can think of this as the ‘fiduciary’ responsibility of economists).

However, economists are also people, and they often have their own preferred goals/outcomes, which is where things get murky.

When Policy Recommendation and Personal Preferences Get Murky

The problem for economics occurs when economists or economics commentators start intentionally mixing policy recommendations with their own preferences. Unfortunately, these are rarely easy to spot.

One notable example to me is the recent proposal by New York City mayor of lowering transit cost by providing free buses. Numerous articles and comments came out criticizing the program. But the reality is that economics research is quite clear that free buses (and free transit) is a good policy to address this issue. However, the field of economics has not pushed back against the false narrative being spun around the policy.

Another topic where the policies and personal preference got murky is the debate around wealth inequality. Many economists and commentators are often very critical of any changes to taxation or the introduction of wealth taxes. These commentators often point to the costs of implementing the policy, as if the costs themselves are sufficient reason for not implementing changes. Instead, it should be the choice of the public about what costs they want to bear and trade-offs they want to undertake.1 It is not for economists to decide which policy to choose – just to inform about the cost/benefits to the best of our knowledge.

Thus, the problem for the field of economists is that many of these economists/economic commentators covertly pass on their personal preferences and views, under the guise of economics. This is why Jon Stewart ultimately viewed the field in the negative context – he’s seen plenty of headlines and news pieces, which do that, and assumed that the entire field of economics must be like this. Economists, unfortunately, ceded the conversation on what economics does and let it be usurped by fancy looking opinion pieces. We shouldn’t be surprised that an opinion like Stewart’s is out there – we haven’t done much to combat it.

What Should Economists Do?

Naturally the next question is what should economists do, to be accurately depicted.

First – an easy one – don’t pile on people like Stewart (or anyone for that matter – as mentioned by Claudia Sahm). Stewart has his beliefs and preferred social outcomes. We, as economists, should engage people like Stewart and can help explain which policies can achieve his preferred social outcomes (Thaler did a good job with this in the podcast).

Moreover, since we economists are trained to think in particular frameworks which focus on preferences, in discussions like the one with Stewart, we should make sure we understand people’s preferences over social outcomes. Too often the debate focuses on a particular policy (free buses, wealth tax, etc), which is a tool and not the outcome the policy is supposed to generate.

But what else should we do? This is where it gets harder, but economists should start a conversation about how to properly communicate what we do. As mentioned earlier, the vast majority of academic economists and commentators act in the way I described economists should. Often, however, the most vocal ones do not. I do not have an easy answer on how to deal with this. Perhaps there is a need for a centralized body/committee that can clearly present economic findings, or present how different policies will affect the economy. Or maybe handbooks (similar to the ones in medicine) that may act as solid starting points for policies.

Taking One Step at a Time

This post is one of the reasons I started Nominal News, as I have seen this problem grow over the last decade. With Nominal News, I try my best to present economics as the important and useful field it is – a field that is filled with many great and brilliant economists and educators that have designed plenty of great tools to use under different circumstances.

I will finish this post with a humorous post by Professor Ben Ho:

Let me know if you have any thoughts on how economics can improve its image and what I, at Nominal News, could do better?

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. What is Critical Thinking? And why is it Important? – discusses what it is ‘critical thinking’ and what steps we can take to improve our critical thinking skills.

2. Why Affordability and the Vibecession Are Real Economic Problems – synthesizes various strands in the affordability/vibecession debate. This includes the rising prices of essentials, higher costs to own a house, and the overall impacts of higher economic uncertainty.

3. AI and the Economics of the Human Touch – wrote an important piece highlighting how to approach AI from an economic standpoint, and uses historical examples to explain why technological change is unlikely to oust humans.

Subscribe to Our Newsletter

In the US, a recent report by Challenger, Gray and Christmas Inc (the “Challenger Report”) made significant headlines, as it showed a surge in US layoff announcements in January 2026. The report stated that 108,435 job cuts were announced, a level not seen since 2009. Companies like Amazon and UPS have announced major layoffs. What to make of these layoffs, and, more broadly, are firms making the correct decision when conducting layoffs?

Layoffs in Context

First, it is worth putting the layoffs in context of the US Economy – weekly, around 100,000 to 300,000 new people file for unemployment. This suggests that these job cut announcements are not likely to impact national unemployment figures due to their relatively small size in comparison to the number of unemployed (around 7.5 mln currently). Moreover, as explained by labor economist here, the Challenger Report’s layoff announcement figures do not track actual layoffs well. It is directionally correct (i.e. layoffs have increased), but the magnitude in the Challenger Report is usually much

higher than what actually materializes. The reason for this discrepancy is unclear.

But how do layoffs, in general, impact:

• the wider economy;

• the individuals affected (which includes both laid off workers and workers that were not laid off); and

• the company making the decision to conduct layoffs.

I’ll start off from the latter.

Firms and Layoffs

The decision to layoff employees is an independent choice by the management of the company. Therefore, we would expect that this decision should bring some positive impact to the company, preferably in a quantifiable way. Answering this question is not as easy as it seems due to the fact that it is difficult to establish the counterfactual, i.e. what would have been had the company not laid off employees. Furthermore, as each company’s circumstances are unique, undertaking a ‘natural experiment’1 is not straightforward. It is difficult to argue that the company that did a layoff is similar to a company that did not do a layoff. However, attempts to address these issues and develop some insights have been made.

Layoff Impacts on Company Management and Stock Performance

Carriger (2016) summarized the research on the impact of downsizing (layoffs). There are two key ways to look at how downsizing might impact a firm – a management perspective, which focuses on how metrics such as profitability and return on equity or assets react to downsizing, and a finance perspective, which focuses on how a company’s publicly traded stock responds to the downsizing.

From a management perspective:

• De Meuse et al. (2004) found that the financial metrics of downsizing firms worsened in the years after a layoff event, but over time, improved back to match the financial performance of non–downsizing companies;

• They also found that companies that did fewer layoff events performed financially better than companies that had more layoffs;

• Guthrie and Datta (2009) found that downsizing reduces firm profitability, and is especially worse in industries characterized by large research & development investment, high growth, and low usage of tangible capital, such as machines (note: interestingly, this seems to describe the characteristics of the tech sector);

• Chalos and Chen (2003) found that the cause of downsizing matters. If the reason to downsize is cost-cutting (a reactive downsizing), then firms have worse financial performance; if it is revenue-refocusing (i.e. pro-active downsizing in response to future market trends), then financial performance improves.

Even with the caveat that it is difficult to ascertain what would have happened had these firms not downsized, which would allow us to establish a causal link, most research shows that there is generally no evidence or correlations suggesting downsizing for cost-cutting purposes will improve financial performance. Turning to the financial perspective:

• Chen et al. (2001) found that on the layoff announcement day, stock prices respond negatively, especially if the cause behind the layoffs is lowering demand for the company’s goods or services. They do find that profitability and labor productivity increases for companies that conduct layoffs. However, these results occur for companies that have been underperforming to their industry peers for the prior three years. Three years after the layoff decision, operational metrics of firms that conducted layoffs outperform industry peers, while employment reverts back to pre-layoff levels. Interestingly, this improvement does not get reflected in stock market performance – the stock performance three years after the layoff period remains similar to its peers. Hiller et al. (2007) find similar results;

• Brookman et al. (2007) found that companies in which CEO’s pay is linked to the performance of the stock are more likely to announce layoffs; this layoff decision does generate positive returns to shareholders. They find that this is potentially due to CEOs “adopting operating strategies that improve operating profits and stock performance”. Furthermore, similar to the Chen et al. (2001) study, Brookman et al. (2007) found that company internal performance metrics improved within two years of the layoff announcement;

• Marshall et al. (2012) found that stock prices respond positively to layoffs in good times, but fall in bad times;

• Capelle-Blanchard and Couderc (2007) conducted an overview of the entire literature of stock prices reacting to layoff announcements and found that generally stocks respond negatively to these announcements. As with research mentioned above, if the layoffs are driven by ‘defensive’ reasons (the firm is facing a difficulty), then the stock performance is even worse.

From the above, it appears that layoffs have negative impacts in the short-term on company stock prices. If the layoffs are ‘defensive’ in nature, most evidence points that both company stock performance and internal performance will suffer. This does not necessarily determine causality – i.e. do layoffs make the firm worse, or was the firm going to be faring poorly regardless of layoff decision. If the layoffs are done as a strategic decision, evidence suggests that companies fare somewhat better after them. Again, the same causality issue applies as above.

Longer Time Horizon

To further address some of these contradictory results, Carriger (2016) tracked companies for a longer period of time during the 2008 financial crisis. He partitioned companies into 4 groups by two categories: financially healthy and unhealthy companies based on the amount of available cash they were generating, and whether companies downsized in staff or not, specifically in 2008.

He found that companies that downsized in 2008 did worse in the period 2008-2014 than non-downsizing companies, regardless of financial health. However, any difference between these two groups became negligible in 2014. Furthermore, companies that downsized in 2008, immediately began re-hiring, and within 3 years were back to their pre-downsize employee count.

One important issue with this analysis is “Survivorship Bias”. Survivorship bias is the fact that when doing data analysis, we can only use data points of companies that existed (i.e. survived) throughout the data analysis period (in this case, from 2008 to 2014). More specifically, in the case of this particular study on layoffs and firms, we do not observe firms that ceased to exist in 2014. Thus, we do not have data points on companies that either chose or didn’t choose to downsize, but went bankrupt in that time frame. For example, a company that downsized in 2008 and in 2014, appeared to have poor financial performance, maybe would have gone entirely bankrupt if it had not downsized.2 In this instance, survivorship bias may make downsizing and layoffs seem like a less beneficial tool to the firm than it really is.

Layoffs and Individuals

Layoffs impact specific people. Economists are very interested in this impact because layoffs can potentially create many costs. For example, being laid off destroys accumulated human capital3 that was specific to the firm (i.e. by working at a firm, you already know how everything works and because this knowledge is specific to the firm, upon getting laid off, this knowledge is wasted). Furthermore, there are potentially many impacts on laid off workers, including changes to long-run career prospects, health outcomes, and even impacts on their families.

Unexpected Job Loss

Davis and von Wachter (2011) summarize the research around the various impacts of job displacement, an unexpected job loss initiated by the firm, usually involving multiple people being laid off. In their own research, they find that for laid off workers during mass layoff events, lifetime earnings of displaced workers drops by approximately 1.4 years of pre-displacement earnings (i.e. the drop in lifetime earnings is equivalent to you not working for 1.4 years, regardless of length of unemployment). This effect is twice as large, if at the time of the layoff, the unemployment rate in the US is high, above 8%.

There is debate, whether the act of layoffs itself results in lower earnings, or whether a different underlying cause is responsible. Fallick et al (2021) find that it is the joblessness duration after the layoff that explains all the drop in earnings – the longer one is unemployed, the worse their future earnings outcome.

Why Length of Unemployment May Matter

There are several hypotheses why duration of unemployment may have a significant impact on future earnings. The mains ones are:

1. worker differences (there is something inherently different about workers that stay unemployed longer, which results in them earning less in the future);

2. human capital depreciation (the longer one is unemployed, the more they ‘forget’ about the job and become less effective and thus, receive a lower wage);

3. changes to local labor markets (if many of those laid off are in one market, there is an over-supply of workers, leading to a reduction in earnings); and

4. moving to a worse firm (this is called the job ladder model: good companies hire from worse companies, but worse companies hire from the unemployed).

The authors find that the most likely reason is number 4. Workers that stay unemployed for longer than four quarters see significant earning reductions because they end up at generally worse firms.

Layoff Impacts – Lower Consumption, Worse Health and Less Child Education

Browning and Crossley (2001), using Canadian job loss data, find that consumption drops significantly for laid off individuals who are in the lower end of the income distribution, as they do not usually possess savings to use for consumption during unemployment spells. Stevens (1997) finds that job displacement lowers job stability, increases earnings instability, increases job and industry switching, as well as makes future unemployment spells more likely. These effects are very persistent, lasting 10 years after the layoff, resulting in earnings 9% below expected levels. If, however, a displaced worker is able to find a job in the same industry very soon after being laid-off, they experience an increase in earnings in comparison to their previous job.

Regarding health, workers that experience job loss have higher incidences of strokes and heart attacks (Burgard, Brand, and House, 2007). Sullivan and von Wachter (2010) use administrative and social security data that allowed them to track workers that were displaced and found that mortality rates significantly increased for this group. Based on their analysis, a job loss during a recession reduced life expectancy for middle-aged men by 1 to 1.5 years.

Besides impacts on the individual, job losses also spread to the family of the impacted workers. Stevens and Schaller (2011) established that parental job loss reduced children’s educational achievement, with children being 15% more likely to repeat a grade. Wightman (2009) further shows that both educational attainment and cognitive development of the children suffer after job loss. Other studies have shown that job loss increases divorce rates4 (Charles and Stephens, 2004), reduces fertility, and decreases homeownership (von Wachter and Handwerker, 2009).

Layoffs have a significant cost on employees. These costs are both direct with falling earnings and consumption, but also indirect, as health, life, and family outcomes are also adversely affected.

Layoffs and the Wider Economy

The prior two sections focused on the impact of layoffs on firms and individuals. These impacts, however, also have costs to the economy. From the firm side, layoffs might lead to lower profitability, which means we are generally using resources less efficiently, and the government has lower tax revenues. The negative consequences to the workers are also costly to the economy; as a society, we have less effective workers due to losses in health, human capital, and job match. Furthermore, government services are also additionally strained due to the adverse health and family outcomes of impacted workers, as they will seek assistance from government programs.

Local Community Impact

Beyond these economic costs, how do layoffs affect the communities and areas they are located in? Foote et al. (2018) studied mass layoff events from 2000 to 2011 to establish how impacted local labor markets respond to such events. They found that, if 1% of a county’s population gets laid off, the county’s labor force drops by 0.19 percentage points. This is caused by two key channels: migration (that is local population moving to other parts of the US) and quitting the labor force (i.e. permanently not looking for a job, retiring, or going on disability). Typically migration was the main driver of local labor market adjustments to layoffs. However, the authors found that after the Great Recession of 2007, quitting the labor force grew substantially and became the predominant way local labor markets adjust to shocks.

Layoffs – Costly to Firms and Employees

Summarizing:

• Current layoff announcements are relatively small and their magnitude overestimates the true size layoffs;

• It is unclear if firms that do layoffs are better off after layoffs from both operational and stock price perspectives;

• Workers that are laid off see real negative impacts – lower consumption and lower health outcomes.

I would preface the above conclusions with the fact that research on layoffs and its impacts still appears to be in its infancy, in part due to data difficulties and assessing the counterfactual (i.e. what would have happened if the firm did or did not conduct layoffs). Overall, though, it appears that the impact of layoffs on the overall economy is very murky – it is unclear if firms truly know that layoffs will improve their own performance, while costs to workers are significant.

Interestingly, the US unemployment insurance program requires firms that have undertaken layoffs to pay larger unemployment insurance taxes for a period of time. The bigger the layoffs, the larger the tax. This policy appears to focus on making the layoff decision a bit less frivolous. During the Great Recession of 2007-2009, this mechanism is estimated to have reduced layoffs by around 825,000.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. When “Efficiency” Means Admitting You Were Wrong – writes in that the recent layoff announcements may be driven by ‘malinvestment’ (when capital gets allocated based on distorted market signals rather than genuine economic fundamentals). In other words, firms mistakenly over-hired several years ago and now are incorrectly conducting too many layoffs.

2. When “Preferences Changed” Stops Being an Explanation – tackles the commonly used explanation for observed economic phenomena (my preferred example being the increase in working from home) is that people’s “preferences” have changed. As Sebastian points out, one needs to explain how and what caused preferences to change before claiming they did.

3. Will Trump’s Fed Nominee Kevin Warsh Be Chair In Name Only? – discusses the appointment of the new Federal Reserve Chair with a slightly different perspective from mine, arguing that Kevin Warsh will be able to push through his preferred interest rate policy.

Subscribe to Our Newsletter

What Happened?

On January 30, 2026, US President Donald Trump picked Kevin Warsh to replace Jerome Powell on May 15, 2026 as the Chair of the Federal Reserve.

Why Does It Matter?

The Federal Reserve (the US Central Bank) sets US monetary policy. Unlike other central banks, the Federal Reserve has a dual mandate:

• Maintaining price stability by keeping inflation stable (typically at 2%);

• Maximizing employment.

Decisions made by the Federal Reserve, therefore, influence inflation and employment. These monetary policy decisions are made via a vote in the Federal Open Market Committee (FOMC). There are 12 voting members in the Federal Reserve:

• 7 Federal Reserve Governors (of which the Chair is one);

• The head of the New York Federal Reserve;

• 4 rotating presidents of the regional Federal Reserve Banks (there are 11 such banks).

The Chair of the Federal Reserve has one vote like all other voting members.

Why Are We Writing About This?

Independence Concerns

The reason this nomination garnered more media attention than usual is due to comments from President Trump about wanting lower interest rates. These comments have raised concerns that the Federal Reserve may no longer act independently of the government, which may result in economic instability.

Monetary policy, as mentioned above, influences both inflation and employment – high interest rates, which lower inflation, also slow down economic activity, which increase unemployment. This can naturally create a problem for the government – a slow, high-unemployment economy can reduce the chance of winning an election. That is why monetary policy is often not controlled directly by the governments, as governments may be tempted to protect their electoral position by keeping interest rates lower. This, however, can result in runaway inflation, due to the ‘unanchoring of inflation expectations’.

‘Inflation expectations’ is one of the three key drivers of inflation (if you’d like to learn more, we have a guide on inflation here). In simple terms, what people and firms believe inflation will be in the future impacts today’s inflation. Thus, if people believe that the Federal Reserve will no longer focus on keeping inflation stable in the future, then today’s inflation will rise. Restoring the ‘anchor’ can take many years, as you need people to credibly believe that the Federal Reserve will prioritize keeping inflation in check. (Somewhat worryingly, here is an observation by that inflation expectations may already be trending higher).

Should We Be Concerned?

I personally don’t think that inflation expectations will un-anchor. Monetary policy is voted on by 12 voters and thus, it is unlikely that a majority of these voters would go against what economics would broadly suggest (i.e. allow for inflation expectation to un-anchor). Moreover, although Warsh has stated that he would prefer to lower interest rates today which could increase the already high inflation of 3%, some of his other monetary policy preferences would lead to lower inflation. Let me explain.

Kevin Warsh’s Thoughts and Preferences

In one article (“The Federal Reserve’s Broken Leadership”), Warsh argues against the ‘large’ Federal Reserve balance sheet and would prefer to shrink it (commonly referred to as “Quantitative Tightening”). This is actually a disinflationary policy.

An Aside on Federal Reserve Balance Sheet Operations

The Federal Reserve holds assets – predominantly US Treasuries (government bonds) and Mortgage Backed Securities (home loans). The reason the Federal Reserve holds these assets is to increase the amount of ‘liquidity’ in the financial system, as well as lower the interest rate that needs to be paid on long-dated borrowing (especially home loans).

To understand how this happens, let’s think of a home loan. Initially, a bank issues a loan to a home buyer. The bank now owns an asset – the loan – which pays interest and principal in regular installments. However, the bank is ‘stuck’ with this asset, as it only slowly gets the money from the home buyer.

In comes the Federal Reserve, which can buy this home loan from the bank. The bank instantly receives all the cash and can re-lend this money, stimulating the economy. The Federal Reserve now holds the asset (the home loan) and slowly gets paid back. Since the bank knows that the Federal Reserve will be committing to buying the home loan, the bank can also offer a lower interest rate on future home purchasers. In this way, the Federal Reserve can both increase liquidity in the market (the bank now has cash it can lend again) and also reduce the long term interest rates (home loans have lower interest rates).

Reducing the Balance Sheet

Turning back to Warsh’s article, he stated a preference of reducing the Federal Reserve balance sheet, which would entail selling off the assets that the Federal Reserve has. This will have a disinflationary effect, as it would do the opposite of buying assets – increasing borrowing costs and reducing the amount of cash in the market.

So it appears Warsh prefers two policies – lowering interest rates and reducing the balance sheet – that work in the opposite directions, meaning inflation is unlikely to un-anchor.

What Does Concern Me

Although I am not concerned about inflation spiralling out of control, I am a bit concerned by the approach Warsh has towards analyzing certain economic events. Here are a few examples:

On AI

Warsh believes “AI will be a significant disinflationary force, increasing productivity and bolstering American competitiveness.” This is a strong and risky belief, as evidence for it is very minimal. Warsh predicts a 1 percentage point higher productivity growth, which is nearly doubling the long run productivity growth of 1.5%. Again, this appears to be very unlikely and more wishful than academically rigorous.

On Recent Inflation

“Second, inflation is a choice, and the Fed’s track record under Chair Jerome Powell is one of unwise choices. The Fed should re-examine its great mistakes that led to the great inflation”.

Warsh puts blame squarely on the Federal Reserve for high inflation. However, as we have discussed before, keeping inflation at 2% at all costs during the pandemic would have ended up severely harming workers. The Federal Reserve followed the latest research in not increasing interest rates by as much. Its action during the Covid pandemic resulted in large real wage growth for many workers, especially on the lower end of the income distribution. At the time, the choice for the Federal Reserve was either higher than usual inflation or large unemployment, and they wisely chose the first.

On Tariffs and Inflation

In this op-ed, Warsh states “So long as Mr. Powell’s Fed hasn’t suffered from a loss of inflation-fighting credibility, the effect of a 10% tariff [on inflation] shouldn’t be statistically significant.”

This is a surprising statement, as we know today that the tariffs have led to at least 0.65% extra inflation:

Tariffs do lead to higher inflation, as we discussed here.1 Tariffs (and potentially current immigration policy) are some of the main reasons we are at 3% inflation today rather than 2% inflation.

Concluding Remarks

Overall, I don’t have any strong concerns on Warsh, as ultimately the Federal Reserve Chair is just one vote in a committee of 12. My main concerns surround how he appears to view certain economic developments through one perspective without looking at all the data, with the tariff example being especially relevant. This dogmatic view of economic events may put him in an interesting spot, where his preferred policy will not be chosen by the FOMC (the interest rate voting committee), leading to some very awkward press conferences.

Tell me in the comments what do you think of the appointment of Warsh. Am I being too harsh on him? Or am I not concerned enough?

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Article – The Means-Testing Industrial Complex: shows how means-testing programs (such as work requirements) end up costing the government a significant amount of money.

2. Article – Young College Graduate Unemployment Is Worse Than the Slowdown Explains: explains that there may be truth to the fact that this slowdown might be worse for young college grads than previous ones.

3. Note: An interesting debate on X/Twitter, where Oliver Blanchard suggested that people in France prefer more leisure to Americans. This ended up with a discussion – do the French prefer more leisure or are they prevented from working more by laws/regulations?

Subscribe to Our Newsletter

The US Supreme Court is the highest judicial body in the US, and opines on federal and constitutional issues. One unique feature of the Supreme Court is that it consists of 9 judges that are appointed for life, making judicial appointments extremely important. These Supreme Court justice appointments, which do not occur often, are usually made by the party in power, which has led to questions about the politicization of the Supreme Court. Surveys from the Pew Research Center show the public’s change in perception of the Supreme Court over time:

So is the public right in their view that the Supreme Court has polarized? Prat, Morton and Spritz (2026) (“PMS”) set out to answer that question.

Do Judges Rule in Favor of the Wealthy?

PMS started with a simple observation – a Supreme Court judge either rules in favor of the wealthier or poorer party of a court case. Different Supreme Court judges vote differently in the same court case. Many theories, involving legal interpretations, are proposed to explain this difference. However, what if instead, judges themselves have preferences over the wealth distribution? PMS decided to build a model of judges’ decision making.

Judges’ Preferences

A judge is assumed to have a preference over the ideal wealth distribution in society (in theory, every one of us has a preferred ideal wealth distribution). This preferred wealth distribution is referred to as the judge’s “ideal point”.

When faced with a court case, a judge can choose to side with the wealthier or poorer party. A judge gets utility (i.e. happiness) depending on the decision the judge makes. If the judge’s utility is higher when they side with the wealthier party over the poorer party, they’ll rule in favor of the wealthier party (and vice versa).

The amount of utility a judge gets from a decision depends on the court case specifics (i.e. it may be clearer which party should ‘win’ the court case) and how close the decision is to the judge’s “ideal point”.

Based on the above model, we can infer the judge’s “ideal point” based on their decisions. For example, if the outcome of a specific court case leans in favor of the wealthy party, while the judge rules in favor of the poor party, that would imply that the judge’s “ideal point” of the wealth distribution leans towards poorer individuals.

Judge Nominations

Supreme Court judges in the US get nominated by the majority party – the Republican or Democratic party. When the party in power gets to pick a judge, it usually chooses a judge from a pool of candidates. The pool of potential candidates is not random, as it inherently reflects the preferences of the party. In our particular case, we assume that both Republican and Democratic parties have preferences over the wealth distribution, and therefore the average judge in the pool of candidates from which they select would reflect this preference.

By looking at which judges are selected for the Supreme Court, PMS infer whether a party’s preferences over the wealth distribution have changed. For example, if judges selected by the Democratic party are more likely to side with the wealthier party in court, that would imply that the Democratic party’s preferences have moved towards the wealthy.

Now that we have the model, in order to determine how party preferences have changed, we need to look at data.

Data Collection

Contentious Court Cases

PMS, using undergraduate and graduate students, went through all Supreme Court rulings since 1953. First, PMS excluded all court cases in which the decision was unanimous, i.e. all 9 judges voted in the same way. The reason for this is that when all judges vote the same, it is not possible to infer the judges’ “ideal points” (preferences). To give a simpler example of why these court cases need to be excluded, if you and a group of friends all pick pizza when choosing dinner, then you cannot figure out from this observation who likes pizza the most (the technical term for this issue is we cannot ‘“identify” the preference).

PMS therefore focused solely on court cases where there is at least one Supreme Court judge that voted for the wealthier party and at least one judge voted for the poorer party.

Cases with Wealth Transfer

Next, PMS looked at court cases that resulted in a wealth transfer from one party to another. For example, if the case involved an employee receiving a payment from an employer, this court case would be selected for analysis, since the ruling would result in a direct transfer of wealth between the employee-employer.

Lastly, the selected court cases also had to have litigating parties that can be described as ‘wealthy’ or ‘poor’. For example, in the employee-employer case, the employer would be coded as ‘wealthy’, while the employee would be coded as ‘poor’. Court cases where this was not clear, such as a court case between two companies, were not included in the data set.

PMS laid out the protocol for classifying these cases quite clearly and imposed several checks, such as having multiple assistants classify the same cases to determine if there are any mistakes being made.

So Did Judges Become More Pro-Wealthy?

After classifying the cases, PMS looked at raw data:

The above chart shows Republican-selected judges (red) and Democratic-selected judges (blue). The y-axis shows the share of decisions made in favor of the wealthier party. The dots signify the date the judge was appointed. The line by party, at any point, represents the most recent 7-judge average (i.e. the red line is the average pro-wealthy vote share of the last 7 Republican judges).

From the graph, it can be seen that recently, there’s been a big shift towards pro-wealthy voting by Republican-nominated judges and slight fall in the pro-wealthy vote share for Democratic-nominated judges.

If we look at the specific type of cases (the case type is coded by the Supreme Court), the shift in voting occurs for all case types:

For example, in the top right, contentious cases involving unions have always been voted in favor of the wealthier party by the three most recent Republican-nominated judges.

Judges’ “Ideal Point” (Preference)

The above results show the share of decisions made in favor of the wealthy party. As mentioned in the model section, using the observed decisions, we can attempt to infer what is each judge’s ‘ideal point’ or preference for wealth distribution. The chart below shows that:

Values above 0 suggest a pro-wealthy preference, while values below suggest a pro-poor preference. The black line shows the current implied median justice preference. The implied “ideal point” of Republican nominated judges has increased significantly, which also implies that the pool of potential judges from which the Republican party selects has also become more pro-wealthy.

More specifically, the implied probability a Republican-appointed judge in 1953 would side with the wealthy party was 44% (for Democratic-appointed judges, it was 40%). In 2022, the probability that a Republican-nominated judge would side with the wealthier side is 74% (for Democratic-appointed judges, it is 27%).

What Does it All Mean?

The paper I discussed today, although political in nature, is mainly meant to demonstrate the variety of questions economic methods can be applied to. A recent post by talked about what is economics –

“Students arrive at class thinking it [economics] is “the study of money.” It’s not. Economics is the study of decision-making under scarcity.”

I liked this definition, and, moreover, I think that economics is more broadly the study of decision-making.

Today’s paper studied two decision processes – the decision of nominating a judge and judges’ voting decisions. As mentioned previously, judges’ decisions are often analyzed based on the legal school of thought (like textualism or originalism) judges prefer to use to interpret law. However, not all decisions by the same judge seem consistent with the school of thought. The research discussed today may explain this discrepancy, as it appears that judges’ wealth preferences may help explain why judges make different decisions in similar legal cases.

The Benefits of Clarity

Given the political nature of the question, there can be a lot of agreement/disagreement with the findings. The benefits of this paper by PMS is that the approach to the question is clearly laid out – which court cases were looked at, how the cases were coded, etc. This enables for a much more productive discussion, as any and all assumptions made by PMS are clearly stated.

One minor concern I noted is that the authors cannot control for the possibility that the cases brought to the Supreme Court may have fundamentally changed over time. For example, if in the 2000s cases, the law was more ‘objectively’ on the side of the wealthy parties, then some of the observed shift in Republican judge’s preferences may be due to the change in cases rather than an actual shift in preferences. However, some of this concern is alleviated by the fact that the Democratic-nominated judges sided more with poorer parties. Nonetheless, thanks to clarity of the model and the research, we can have such discussions.

The study by PMS is definitely not the last one on this topic and any future research will have to grapple with their findings of large shifts in judges’ decisions favoring the wealthy. Maybe this article will inspire one of our readers to tackle this issue!

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Article: discusses how the Prime Minister of Canada, Mark Carney, may be utilizing his economics and game theory background in foreign policy.

2. Article: talks about how much is spent on immigration enforcement – $170bln – and how that compares to other US government programs.

3. Article: and discuss how preventing immigrant students from getting work in the US will result in more visas being allocated to outsourcing companies.

Subscribe to Our Newsletter

In the US, rewards for using credit cards are immensely popular – Americans received around $40bln in rewards in 2022. However, although ‘rewards’ make it sound like a positive program, research shows that these rewards-based credit cards increase costs for all and exacerbate inequality by transferring from poorer to richer individuals (as well as ‘naive’ individuals to ‘sophisticated’ individuals).

Credit Card Rewards

Credit card usage in the US is ubiquitous, accounting for around $6trln in spend in 2022. These credit card transactions come with a cost – around 1.5% to 4% of the sale price is the cost to the merchant of conducting a credit card transaction. This fee is split between the network provider (like Visa and Mastercard) and the credit card issuer (usually a bank or credit card company).

To encourage usage of credit cards, credit card issuers offer rewards in the form of ‘cash back’ or points (these points can be redeemed for cash or used to purchase items or services).

Naturally, since corporations are in the business of making profits, who ultimately pays for these rewards and why do rewards cards exist at all?

Different Payment Methods - Different Costs

To tackle these questions, Felt, Hayashi, Stavins and Welte (“FHSW”) looked at US and Canadian transaction data made with cash, debit cards or credit cards. Each of these methods incurs a different cost - cash payments may require ATM fees for individuals to get the cash, while debit cards and credit cards have various transaction and network fees. In the credit card world, these fees are actually quite varied.

Credit Card Fee

When a customer uses a non-rewards credit card, the merchant pays on average 1.89% of the transaction amount as a fee (this is referred to as the ‘merchant discount rate’). A basic rewards card typically has a 2.04% fee, while a premium rewards card has a 2.49% fee.

Credit Card Reward Rates

FHSW also computed the average rewards consumers get as a percentage of spend (after fees) – a basic rewards card nets around 0.75% in rewards, while a premium rewards card nets around 1.5% to the user.

Do Merchants ‘Discriminate’?

Based on the above data, it would be logical to ask if merchants charge different prices based on the method of payment. Briglevics and Shy (2014) looked at this question and determined that merchants would likely not benefit from such behavior.

Merchants could at most offer around a 1% discount to encourage other forms of payment (cash or debit cards). However, the profit this would generate is unlikely to cover the costs of offering different prices (confusing customers with different prices; sales delay by having to explain the price difference; customer distrust when seeing different prices).

Given that merchants are unlikely to price discriminate, who then pays the cost of the credit card fees?

Cash Payers Subsidize Card Payers

FHSW, under the assumption that 90% of the card fees gets passed on to consumers via a flat increase in price on all goods/services, found that the majority of the total cost is borne by high income households:

As the chart above shows, after accounting for rewards, high income individuals, on average, spend $22.61 a month to cover credit card merchant fees, while low income households spend $11.25.

This fact isn’t surprising as high income individuals spend the most. However, as a percentage of spend, the lowest income households pay the largest amount:

On the left we see the US data, which tells us that 1.41% of the spend undertaken by individuals with an income of less than $25,000 goes to cover transaction costs. As can be seen, higher income individuals (above $150,000 annual pay) pay only 0.82% of their spend in transaction costs.

As merchants charge the same sticker price, the higher reward credit cards, which are costlier to merchants, are in effect subsidized by lower income households who often do not have access to these credit cards in the first place, as they rely on cash instead. It’s also worth noting that this analysis does not include any credit card interest payments that may be due.

Subscribe to Our Newsletter

Share

Sophistication Between Credit Card Users

FHSW research looked at all payers – cash, debit cards and credit cards, and focused only on the transaction costs. What if we focus solely on credit card users and the interest payments associated with credit cards? Agarwal, Presbitero, Silva and Wix (2025) (“APSW”) did just that.

APSW used a US dataset that contains granular information on individual’s credit cards, including information on credit limit amounts, how much has been borrowed, the interest rate on the credit card, the repayment amounts and the FICO (Fair Isaac Corporation) score (Note: the FICO score is a metric that aims to capture the credit riskiness of an individual borrower).

APSW looked at “Net Rewards”, which is the rewards a credit card user gets minus the fees and interest they paid on the card. What they found is that FICO scores predict whether an individual benefits from the card or doesn’t.

The chart above shows that income is not a predictive metric of reward benefits, as high income (green line) low FICO users seem to have quite high costs with reward cards. So what explains why FICO scores may matter regarding benefiting from a card?

Testing for “Sophistication”

APSW argue that high FICO users are more ‘sophisticated’ financially, as they are better at optimizing their finances, than low FICO users. APSW test sophistication in the following way.

In their data, APSW observe whether an individual received a credit limit increase initiated by the bank (i.e. the credit card holder can now borrow more). APSW found that both low and high FICO users increase the amount they borrow after a credit limit increase, but only high FICO users also increase how much they repay of the borrowing, while lower FICO users repay the same amount as before. This suggests that high FICO users are able to forecast their cashflows better than low FICO users.

Additionally, high FICO users appear to optimize which credit card they repay, focusing on higher interest rate cards first. On the other hand, lower FICO users use ad hoc rules of thumb such as repaying a fixed percentage of each credit card balance, which leads to an overall higher interest payment.

The chart below presents the share of ‘misallocated’ credit card payments (i.e. the individual would have been better off to use their payment to pay off a higher interest rate credit card):

Bank Profits from Credit Cards

APSW also looked at the profits banks receive from different types of credit cards – reward credit cards vs non-reward credit cards. The chart below shows the profitability by card type and FICO score for banks:

Reward cards generate significantly higher profits than non-reward cards. What’s worth noting is that reward credit cards often have lower interest rates on them vs non-reward ones. The higher profits banks make on rewards cards mainly come from the higher interchange fee (the fees paid by merchants) and the fact that people appear to borrow more on rewards cards.

APSW showed how these different sources of revenues are split by FICO scores – the table below shows that banks receive most of their revenues from low FICO users from interest payments, while interchange income (i.e. the fees merchants pay) is the main driver of income from high FICO users:

Rewards Cards – The “House” Always Wins

The findings from FHSW and APSW show that credit card reward programs are mainly funded by transfers from cash payers to credit card payers, and transfers from ‘naive’ individuals to ‘sophisticated’ ones, with the credit card issuer taking a profit margin.

Now, a certain profit margin makes sense – the banks and credit card networks provide a service for which they ought to be compensated. The concern to me is that rewards programs, with their higher interchange fees, simply increase costs on all consumers with no added value. Non-rewards cards are cheaper to operate and provide an identical level of service from the credit card companies to the service offered by rewards cards. The only difference is the ‘rewards’ feature which is ultimately a transfer of money from one group of individuals to another, akin to what happens gambling. The credit card issuer pockets a higher total fee from merchants under rewards cards than non-rewards cards, which is why they choose to issue such cards.

Rewards cards take advantage of people’s behaviors by increasing indebtedness and inducing extra spending through the promise of rewards. At the same time, merchants also have to pay a higher cost for rewards cards. Neither of these aspects appear to be welfare improving, and the only beneficiaries are, on average, richer and more ‘sophisticated’ people.

Thus, there is reason to ban rewards credit cards, as they appear to be more predatory than welfare improving. Generally, I don’t like bans as a policy tool, so another possible solution is simply placing a tax on rewards and cash-back, treating it in the same way as income. This would likely significantly dissuade the usage of rewards cards.

P.S.

Incidentally, right as I was finishing this article, the US announced plans to limit credit card interest rates to 10%. The stated goal of this policy is to improve outcomes for consumers. Without getting into much detail, such an interest rate will likely result in reducing access to credit, which is negative to people’s welfare. On the other hand, banning rewards cards would likely be a better option as it may encourage more competition amongst credit card companies on interchange fees they receive from merchants. Since there will be no varied rewards programs, credit card companies would have to differentiate themselves with their level of interchange fees to encourage merchants to use their card and gain more market share. These lower fees would be passed on to consumers.

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Subscribe to Our Newsletter

Share

Interesting Reads from the Week

1. Article: discusses how research findings sometimes may not replicate and its important to update ones beliefs. In this case, research which suggested that turning on subtitles for children improves their reading ability may not necessarily be true.

2. Article: goes over how utility and preferences can impact our choices (especially why someone may prefer bad coffee). What might seem ‘irrational’, in economic terms, might be perfectly rational!

Subscribe now

If you would like to support us in reaching our subscriber goal of 7,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

As we are about to enter 2026, we wanted to do a small recap of how 2025 has been for Nominal News, as well as introduce to you a new chat feature we have launched (if you’d like to skip to the new feature – please scroll to the Looking Forward section).

Nominal News – By the Numbers

At the end of 2024, we wrote the following:

“Last year, we set our target to reach 1,000 subscribers. For this 2025 year, we’re going to set ourselves a target of 3,000 subscribers.”

As of this article, we have 4,799 subscribers so we ended up achieving our goal and then some! We grew by about 3,600 subscribers this year, and for that we want to thank all our readers – every read, open, share, like and comment are extremely meaningful to us, as it helps grow our Substack.

Most of the growth came in a significant surge at the beginning of the year (when one of our Notes on the Substack platform was widely shared):

Since then growth on a net basis has stalled, but we still couldn’t have forecast such high subscriber count and retention.

With regards to the number of articles we wrote – here is a summary graphic from substacktools:

We wrote 32 articles in 2025, which is 4 more than in 2024. Consistency is something we’ve been trying to improve upon, so publishing a higher number of posts than last year is a personal success.

Our most liked and discussed articles, as can be seen in the graphic above, are all related to inflation – a topic that remains in the headlines. Links to those three articles can be found here:

• Inflation - a Short Primer

• Do Tariffs Cause Inflation? Yes.

• The Federal Reserve is Right

Another highly viewed article was our discussion of the research on work requirements for benefit eligibility, specifically how work requirements are an ineffective policy.

The Substack Network

Our growth would not be possible without the Substack Network. We have discovered several new interesting Economics Substacks this year:

• The Peeples Economist – a fellow economics PhD, writes about economic research on important social topics such as education, affordability, and, most recently, on trust in the field of economics.

• Mike Konczal’ Substack – former Special Assistant to the President and Chief Economist of the National Economic Council under President Biden and current Senior Director of Policy and Research at the Economic Security Project, writes about economic and social issues with a focus on policy making and policy outcomes.

• Housing Hell – , a former economist at Freddie Mac (a US government-sponsored enterprise that specializes in residential mortgage liquidity), focuses on real estate economics and the research in this field.

Moreover, I continue to recommend the following great newsletters:

• Monday Morning Economist – a newsletter by that covers a broad set of economic topics in a very accessible way. Topics include the limited cost savings from government bureaucracy, the economic impact of Taylor Swift and how universities are competing on price.

• American Inequality – presents very important data via maps to explain how inequality impacts the US. Along with economic insights, the American Inequality newsletter elucidates the distributive consequences of economic inequality in the US.

• Can We Still Govern – presents actionable economic policy insights on how to resolve certain economic and social problems.

• Emancipatory Research – undertakes deep dives into important historical events, as well as the effects these events have to this day. Keshler also interviews individuals related to these events which gives a personal perspective on the topic.

• Risk and Progress – synthesizes research and knowledge from a variety of fields to discuss how human progress and development continues to evolve, as well as how to foster human innovation.

• Alpha in Academia – a newsletter focusing on academic research in finance, which distills key and actionable findings of recent research papers.

• MacroMostly – analyzes US labor market data, providing key insights about the state of the economy.

There are many great Substacks and the list above is just the tip of the iceberg.

Looking Forward

Nominal News is a 2-person team – a writer and an editor – that started in November 2022. We started Nominal News with the intent of bringing and sharing the tools and skills that I’ve learned during my Economics PhD, and use them to discuss current economic issues. (Actually, my favorite piece to write was the “Reader’s Guide to Economic Headlines”, which gives a three-question approach to filtering out the noisy headlines from the valuable ones. The editor’s favorite piece was “H-2Bs - Immigrants Without Degrees Spur Economic Growth”, which shows the positive economic effects of temporary non-agricultural workers)

Currently, this is a publication we do in our free-time (one of the reasons why we have not enabled a paid feature), but we hope that as we continue to grow, we will be able to turn this into a full-time project!

With regards to goals, for 2026, we’re going to set ourselves a target of 7,000 subscribers. I think this will be a tough goal to achieve. As always, your support is always the most impactful in helping us with this goal and the longer goal of making this a full-time project, which is why if you do find our writing interesting, please consider liking, commenting and sharing our work.

New Feature Launch

On that note, we have recently launched a Nominal News chat/discussion board on Substack, where we post interesting research papers that we might not always get to write about. If you enjoy research, please join, comment and ask questions in the chat linked below:

Thank You!

Lastly, Nominal News would not be possible without you! So we would like to take this opportunity to thank all our subscribers and followers, as you are the reason we keep working and writing new articles. We hope that you find our articles useful and interesting and we are always interested in making them better for you!

Thank you for being a valued reader of Nominal News and a Happy New Year!

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

Note from Nominal News

Wishing all our readers a Happy Holiday period and thank you for supporting our Substack! Next week we will do a New Year special, recapping our year.

Every year around the holidays, the paper by Joel Waldfogel (1993) on the inefficiency of gift giving gets widely circulated in social media and the news (Wired, New Republic, Marginal Revolution, WSJ). Waldfogel, using survey data from a wealthy college, determined that people value the gifts they received between 10% to 33% below the price of the gift. This finding implied that gift-giving is ‘inefficient’ and value destroying.

But a follow-up series of papers with much more sophisticated methods, showed that gift giving actually generates positive value! This suggests that the original Waldfogel paper was most likely influenced by a biased sample (wealthy college students that took economics). If you’d like to read about this research, we’ve covered these papers here:

Interestingly, in addition to the value of gift-giving between friends and family, it appears that gifts from employers, rather than cash bonuses, also influence workers positively as well.

Gifts in the Workplace

Kube, Maréchal and Puppe (2012) (“KMP”) looked at how gift giving to employees impacts worker productivity. KMP recruited workers to a one off task of cataloging books for a library for 3 hours. The job paid 12 Euros (~ 14 USD) per hour. The outcome KMP were measuring was how many text characters were entered in the online catalog. The workers did not know that their work was being measured as part of an experiment.

As part of the experiment, KMP tested out giving ‘unexpected’ bonuses to the workers. Prior to starting the work, different groups of workers in separate work sessions were each told the following:

1. They will earn just 12 Euros per hour (no bonus was announced);

2. In addition to the 12 Euros per hour, the workers will be paid a 7 Euro bonus (20% of total income), which was given to the workers at the start of the work sessions;

3. In addition to the 12 Euros per hour, the workers will receive a 7 Euro water bottle, which was given at the start of the work session.

In theory, being given option 2 (cash) and option 3 (water bottle) should have the same impact on workers. Option 2 may even be preferred. Surprisingly, KMP found the opposite – people that received a cash bonus did not have any statistically significant increase in output, while people that received the bottle entered nearly 25% more characters!

To study this puzzling behavior and rule out any other potential explanations, KMP, in a follow-up experiment, made sure that the 7 Euro price tag on the bottle was visible, in case workers thought the bottle was more valuable than it really was. Just as with the group that received the bottle without the price tag, this group of workers that received the bottle with the price tag also ended up being 20-25% more productive than the group that received the cash.

A Choice of Gifts

To see whether the gift (the bottle) was really seen as a valuable item, KMP conducted another experiment. Workers, prior to starting the work, were told that as a bonus, they can choose either to receive 7 Euros in cash or the bottle worth 7 Euros. Nearly unanimously, workers chose the cash. Interestingly, unlike in the cash only bonus scenario, workers were again 25% more productive. Thus, although the workers did not really value the bottle since they chose cash, the idea that the employer had prepared a physical gift appears to have influenced the workers. This led to one final experiment.

Origami Money

KMP gave one session of workers just the 7 Euro bonus at the start of the work session with a slight twist – the 5 Euro bill was folded into an origami shirt, while the 2 Euro coin had a smiley face drawn on it.

It turned out that the workers ended up again being 25% more productive compared to the baseline of no bonus, and 23% more productive compared to the situation where just money was given:

As the chart above shows, the rate of character entry was higher throughout the duration of the work session (the blue line is the baseline with no bonus). So how can we explain all these findings?

Gifts – Value Creating

Based on the KMP research, employees appear to value when their employer puts in time and effort when giving a bonus. These employees may reciprocate this employer effort – in the case of the KMP experiments, employees increase output by more than the bonus (25% more productive, while their income only went up by 20%). Even presenting cash in a more ‘artistic’ way appears to signal effort and care from the employer, which alters workers’ attitudes.

This result is in stark contrast to the conclusions from Waldfogel’s research – if employers were to follow that advice and just give cash , the output generated would be lower than if employers followed the intuitive tradition of giving gifts. Unlike the news articles we often see, gift giving is actually ‘efficient’. Regardless, KMP shows that gift-giving is a far more complex economic question than many present it to be.

Subscribe now

Share

Interesting Reads from the Week

1. Article: discusses how the current US policies are panning out, and why, as expected by economists, the policies did not achieve the goals they were intended for.

2. Article: argues why, in economics, we need both statistical methods and models to properly answer economic questions.

3. Article: and talk about ‘tree inequality’ in the US. From their article, the chart below shows how greenery differs between neighborhoods of different incomes. This also translates into higher ambient street temperatures for neighborhoods with fewer trees.

   

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

I am currently renting in New York City. Whenever I tell my parents and friends how much I spend on rent, I am often told that it is a ‘waste’ of money – I should buy a home instead, as then the monthly payments partially go into the home I get to keep.

The situation above is a common occurrence – the question of renting vs owning a home is a topic that many consider at some point in life. I will not answer the question today but instead, I will discuss an often overlooked part of the discussion – how do renters vs homeowners respond to recessions in their local markets.

Should I Stay or Should I Go

Bojeryd (2024) (an economics PhD student that is currently on the economics job market!) studied how the decision to rent vs own a home impacts a person’s response to economic shocks. Theoretically, if the economic prospects of a city get worse, especially due to a large outside shock, we should expect the city’s population to fall significantly and rapidly. However, we generally do not observe such an immediate strong reaction. How come?

Norway, Stavanger and Oil

Thanks to very detailed Norwegian data we can answer that question. Bojeryd looked at the behavior of people living in the Norwegian city of Stavanger. Stavanger is a city where a lot of the economic output is dependent on the oil industry. During 2014-2016, global oil prices fell by around 50% due to newly developed technologies allowing the US to significantly increase oil production.

This oil shock deeply affected the Stavanger economy:

In Panel A of the figure above, we see how the earnings of workers in Stavanger changed relative to those in other areas of Norway. The baseline year is 2013 (both the orange dots and red dots are on zero). A dot below (above) the zero line tells us that in that year, the relative income between Stavanger workers and other Norwegian workers was lower (higher) than in 2013. Thus, prior to the oil shock, Stavanger workers saw relatively faster income growth than the rest of Norway. After 2014, relative incomes fell significantly, with oil workers (red) seeing much larger declines.

Panel B shows the change in workers taking up unemployment benefits relative to the rest of Norway. Prior to the 2014 shock, unemployment benefit uptake in Stavanger stayed relatively the same. But after the oil shock, relative unemployment spiked in Stavanger, impacting oil workers more than others.

House prices and rents also reacted in Stavanger –

Both home prices and rents in the Stavanger area (Rogaland) dropped during the oil shock, while the rest of Norway had an increase in rents and prices. Home prices and rents in Stavanger fell by around 30% compared to other areas.

Migration Responses

With such a large local job shock, we should expect to see the migration patterns in Stavanger change. With fewer and worse job opportunities, it may no longer make sense to move to Stavanger or remain there.

Bojeryd found that the leaving rate in Stavanger only slightly increased and only for a few years before returning back to normal (and even falling below the previous leaving rate). Migration to Stavanger, however, fell sharply – far fewer people than usual moved to Stavanger. Moreover, the in-migration rate never rebounded.

Who Leaves?

Bojeryd found that renters and low house-wealth individuals increased their probability of leaving. Renters did so substantially, with a 40% increase in probability that they’d leave.

Surprisingly, high house-wealth individuals actually reduced their probability of leaving Stavanger. To ensure that it is housing wealth that appears to reduce migration in response to a negative economic shock, Bojeryd looked at other possible explanations. These were:

• proximity to family in the area;

• age;

• unemployment status;

• current income level.

None of the above explained the increased likelihood to stay, as housing wealth did.

The Wealth Shock

One possible reason why high house wealth individuals are less likely to move is due to the fact that they can afford ‘less’ housing in other regions. For example, if previously a person that would sell their home in Stavanger and move to the capital, Oslo, could afford a two bedroom home in Oslo, they may now only be able to afford a one bedroom home, as the price of their home in Stavanger fell.

Because the oil shock was specific to one market, individuals become relatively poorer in comparison to other regions. This makes it more difficult to move, as an individual is taking a loss in ‘real’ terms – if they were to move, they would have to downsize.

Who Moves In?

Bojeryd conducted a similar analysis for the people that migrated to Stavanger. Interestingly, there’s an increase in low income renters and low income-low house-wealth individuals. On the other hand, the probability of high house-wealth people moving in significantly falls.

Based on Bojeryd’s analysis, the increase in low income individuals moving in to Stavanger can be explained in the following way:

• Low income individuals are likely to rely on government support (which is fixed typically at a national level);

• Falling rents in Stavanger mean the income received from government support can get an individual more housing (and more consumption since less needs to be spent on rent).

This analysis tells us that the people that are not as reliant on labor income, and, therefore, are not impacted by labor market conditions as much, choose to move to Stavanger.

Homeownership Comes with Costs

Bojeryd’s research shows how a local adverse economic shock (fall in oil prices) resulted in a ‘lock-in’ effect for high house-wealth individuals. These individuals that have locked in a lot of their wealth in their homes are stuck between two not good decisions – either lose wealth (by selling their home cheaper) or sacrificing future income growth.

A homeowner has their labor income and wealth exposed to the same risk - the local labor market. As we discussed previously, this should not be surprising as home prices are very closely linked to local income. Thus, any fall in income will translate quickly into a fall in house prices. Potential home buyers are rarely made aware of this risk when deciding to purchase a home.

Without a doubt, it can also be beneficial to be a homeowner over a renter. But it is important to know all the risks/costs related to purchasing a home. Homeownership is often presented as an ‘obvious’ decision and may push individuals into a bad financial situation based on social pressures.

Homeownership is considered to be an important life milestone. Renting is often viewed a bit more negatively. However, the decision to own is rarely presented as taking a ‘bet’ on how the local economy will develop over the next 10-20 years. Adverse local economic developments can be crippling to a person’s wealth and long run earning prospects – you either lose a lot of wealth (selling a home cheaper) or you sacrifice your future income growth. This risk is never mentioned to potential home buyers.

Subscribe now

Share

Interesting Reads from the Week

1. Article: On the topic of housing, explains why removing taxes on the gain in home value will probably only push home prices higher.

2. Article: explains what the recent data on jobs in the US tells us about the US labor market.

3. Article: goes over how economics (specifically game theory) can help explain video game player behavior in the recent and very popular Arc Raiders game.

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article

Share

Since 2018, sports betting in the US has become legal after the US Supreme Court overturned a federal ban. Sports betting in the US grew extremely quickly – by 2023, Americans placed over $120bln in bets, which generated close to $12bln of revenues (this revenue number is net of payouts) for sports betting companies. 

As with all forms of gambling, the concern with sports betting is how it impacts the financial well-being of people participating in it. Thus, economists asked the question – did the legalization of sports betting adversely impact people’s finances.

Gambling and Vices

Gambling can be viewed as a form of consumption entertainment given the fact that most participants lose money. In this sense, gambling losses are a cost paid to partake in the entertainment derived from gambling. Since sports betting is a form of gambling entertainment, it can be viewed as part of the spending a person allocates to gambling activities like playing the lottery or casino games. Thus, the legalization of sports betting could have no negative impact on a person’s finances if that person bets more on sports, but then spends less on lottery tickets or other forms of entertainment. 

If, however, individuals allocate other money – such as savings – to sports betting, then there may be a risk to their financial situation. Baker, Balthrop, Johnson, Kotter and Pisciotta (2024) (“BBJKP”) set out to determine exactly how the legalization of sports betting impacted finances.

Sports Betting – the Data

BBJKP looked at financial transaction data from 230,000 US households (from an original data set of 60mln households). BBJKP are able to see the income a household earns, as well as the spending on sports betting, along with how much money is being transferred to ‘classic’ investment brokerages (such as Vanguard, Fidelity, etc.), ‘gamified’ brokerages (such as Robinhood) and cryptocurrency platforms. 

Moreover, for some of the households, BBJKP are also able to see credit card and other loan balances from a separate data set. This is important, as it allows us to see if households tap into additional debt when sports betting is legalized.

Overall, BBJKP find that nearly 8% of households in their sample participate in sports betting. Interestingly, sports bettors are more than twice as likely than non-bettors to have used a cryptocurrency brokerage or overdrawn their bank account. Moreover, sports bettors are also 4 times more likely to have played online poker. 

In terms of sports betting, the average amount transferred for sports gambling is about $100 per quarter. However, the top third of bettors contribute nearly $300 per quarter (almost 2% of their income), while the bottom third of bettors deposit only $1.39 per quarter. The data suggests that about 2-3% of the households in the sample are betting significant amounts of money. BBJKP also note that if you split the data set in half – into low and high saving households – low savings households that participate in sports betting allocate a larger share of their income to betting and are also more likely to overdraw on their bank accounts. 

Betting Trends

To determine the impact of legalized sports betting, BBJKP used the fact that US states legalized betting at different points in time. Thus, it is possible to compare the effect of legalizing sports betting by comparing households in states where sports betting was legalized vs states where sports was not yet legalized.

BBJKP find that after sports betting is legalized, the probability of becoming a sports bettor is 14%. Moreover, after legalization, households that actually bet, contribute around $179 per quarter to sports betting (the overall average, which includes non-bettors, is $25 per quarter). Interestingly, BBJKP also find that the amount contributed per quarter steadily grows after legalization. The amounts start off low – at around $4 to $5 per quarter – but grows to $73 per quarter (both amounts are averages that include non-bettors as well) after 4 years.

This trend in increasing deposit amounts can be, in part, driven by the fact that sports betting companies offer ‘free betting money’ to encourage new sign-ups. However, the increasing tendency in deposits appears to suggest that bettors have a minimum betting amount in mind when sports betting (initially the amount is met by the free sign-up bets, but after that, the bettor has to contribute their own money).

Saving or Consumption

Where does the money for sports betting come from – does it substitute for investment/savings or for other forms of consumption such as entertainment? BBJKP find that after the legalization of sports betting, net investment (deposits minus withdrawals) in brokerages starts to fall. On average, the amount of investment falls by $53 per quarter. The trend is, however, increasing over time, as in the chart below (sports betting is legalized at quarter zero):

More specifically, a $1 increase in sports betting, reduces net investment by $0.99 – so nearly perfect substitution. For low saving households, a $1 increase in sports betting reduces net investment by $3 (this additional drop in savings can be driven by the fact that bettors increase other entertainment expenditures such as watching the games). 

Now, it is worth noting that not all investments are the same – investing can also be a form of ‘betting’ if the investments are high risk such as speculative stocks or options (which often have negative expected value similar to betting). Thus, the fall in investments might simply be more ‘sports betting’ but less ‘stock gambling’. BBJKP looked at investments made via more ‘gamified’ brokerages, like Robinhood, and found that, although investments in these brokerages also fall, the size of the fall compared to ‘traditional’ brokerages is 10 times smaller. This tells us that most of the sports betting funding comes from long term investments.

Lastly, BBJKP studied the impact of legalized betting on credit card balances (note – the data for this analysis was limited) and found that credit card balances for low savings households increase, and repayment of credit card bills falls.

Overall, BBJKP find that the legalization of sports betting increases gambling at the expense of savings in traditional brokerages rather than other forms of consumption, and the effect is more pronounced for households that already have a tougher financial situation.

Demand for Gambling

The impact of sports betting shows that there is demand for this form of gambling. BBJKP actually also found that sport betting legalization appears to have increased participation in lotteries. Interestingly, spending on online poker does not increase, suggesting that not all gambling activities are treated the same.

Researchers have also studied if high-risk stock trading can also be seen as a form of gambling. It turns out that there is some evidence for this hypothesis. Dorn, Dorn and Sengmuelle (2015) found that both in the US and Germany, trading in the stock market appears to drop in a week during which lotteries have large jackpots. Moreover, in Germany, there is also a drop in options trading when lottery jackpots are high (options trading generates much greater losses and gains than normal stock trading, so options can be seen as similar to playing the lottery).

Kormanyos, Hanspal and Hackethal (2023) (“KHH”) further investigated the link between gambling activity and stock trading. KHH, using German data, found that people that gamble do not necessarily invest in more speculative stocks (stocks that have a much higher risk) compared to active traders (individuals that buy and sell stocks rather than invest in a stock market index). However, on average, gambling individuals do buy and sell stocks more often by 7%, which results in investment returns being about 1% lower than the portfolio returns of non-gambling individuals. 

Alternatives to Gambling

In their research, KHH also looked at whether gambling individuals focus on particular sports events, which would suggest that gambling individuals focus their gambling on sports they may possess significant knowledge about. KHH found that in Germany, it appears that gambling individuals do not focus on particular sports but rather their choice of what to gamble on is random. This suggests that gambling is a form of entertainment/consumption for gambling individuals.

Combining this research with the findings of BBJKP (that gambling individuals fund gambling by reducing their investments), new ideas have been proposed that merge the entertainment value of gambling, while also protecting investments. In 2009, Michigan rolled out the first large scale Prize-linked savings account (PLSA). These PLSAs merge savings accounts with a random chance of a reward. That is, rather than earning a guaranteed rate of return like a typical interest on a savings account, a bank will pay out the interest payments as a prize with some probability of winning. 

Cookson (2017) found that PLSAs rolled out in Nebraska in 2012 resulted in a 4% to 10% reduction in the amount spent on gambling. Moreover, the likelihood of visiting casinos fell by 15 percentage points in counties that had at least one institution offering PLSAs. Deposits into savings also increased as the raffle date for the PLSA prize drew closer (note – each dollar deposited into a PLSA increases the likelihood of winning the raffle for the prize). Cookson thus found that PLSAs did confer some of the entertainment value of gambling, while also protecting people’s investments.

The Complexity of Gambling

As the research above shows, gambling is a complex issue. With the growth of ‘gamified’ brokerages, even the definition of gambling itself is no longer clear. Gambling does appear to be a form of entertainment that occasionally can be very costly, as with any vice. When governments choose to make access to gambling simpler, whether via legalization of sport betting or brokerage gamification, it can come with significant costs. At the same time, governments can offer or encourage solutions that enable individuals to get the entertainment value of gambling, while also protecting people’s finances.

Subscribe now

Share

Interesting Reads from the Week

1. Article: discusses the evolution of economics from theoretical schools to the more formal and structured economics of today.

2. Article: goes over how banks ‘create’ money and why it’s not that mystical.

3. Article: provides a good way to explain two facts – the fact that people today are richer, but at the same time, they think they’re poorer. Hint: it has to do with the fact that households have dual earning adults.

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

The debate around improving safety in cities often revolves around whether this means deploying more police or not. In New York City, the recent mayoral election had certain candidates promise an increase in police patrols, while another candidate proposed the creation of a non-police unit to deal with certain issue (homelessness and mental health).

But do these non-police approaches work? A recent paper by Blattman, Duncan, Lessing and Tobon (2024) (“BDLT”) tackles this issue with a fascinating research design and find somewhat surprising answers!

The Experiment

Background

BDLT conducted an experiment in Medellin, Colombia to study the impact of increasing civilian and government intervention on the levels of crime and governance. Many cities around the world have been attempting to deal with disorder and crime, preferably without having to resort to weapons. However, often it is the police who are the first responders to a range of everyday disputes including domestic violence, homelessness, regulatory breaches, or other civil disputes. 

In Colombia, the police are a national institution – that is local governments and mayors have limited control over the police, especially regarding staffing levels, police training, and police guidelines. Thus, large cities such as Medellin, have created their own ‘alternative’ security agencies with neighborhood-level government staff dealing with local minor problems. These security agencies are controlled by the local government and deal with street disorder, domestic violence, and public space regulation. In Medellin, the agency is called the Secretariat of Security and has roughly 1 employee per 1,000 residents. In contrast, the police staffing level in Medellin is approximately 2.7 police officers per 1,000 residents. 

Although Medellin is a relatively wealthy city, with a well developed bureaucracy, it suffers, just like many cities around the world, from a street gang problem. These street gangs, called ‘combos’, do not only undertake illegal activities but also provide a variety of basic services for the local community such as security, dispute resolution, and debt collection. These services are often provided by combos for a fee, or weekly taxes, called vacunas. 

Intervention

BDLT worked with the municipal government of Medellin and identified 80 low to middle-income neighborhoods that would participate in the experiment. One very unique element about this experiment was that the municipal government of Medellin allowed BDLT to conduct a randomized control trial. Out of the 80 neighborhoods, only 40 were to receive the intervention of an increased civilian government expenditure, while the others would receive the same level of government service as before. This type of experiment design is very similar to medical trials, where one group of individuals gets the intervention (medical treatment), while the other group – the control group – does not. The difference between these two groups would then be attributed to the intervention. In social settings, these types of experiments are extremely rare, as institutions such as local governments would be worried about any political costs of such a decision. 

BDLT, with the assistance of the local government of Medellin, conducted an intervention on 40 neighborhoods, shown on the map of Medellin below: 

The intervention (or treatment), that was to be conducted by the staff and activities of the Secretariat of Security, had two main components:

1. The 40 neighborhoods would be prioritized by the city’s current main service agencies. This included quicker access to dispute resolution officers for problems between and within families, as well as priority responses to basic services requests such as garbage pickup and street light repairs.

2. Each of the 40 neighborhoods received a full-time ‘liaison’ – a local government representative that would assist in linking people with government agencies, identify public service issues, as well as assist with business and family dispute resolution.

The overall intervention can be viewed as an increase in the provision of typical municipal services. Thus, the experiment can determine the impact of such an increase in the provision of municipal services. The intervention in these neighborhoods lasted for 20 months. In terms of how large this intervention was, BDLT estimated that street-level attention to problems was increased 60-fold (measured as the increase in liaison officers per block, which went from 1 liaison per 540 blocks to 1 liaison per 9 blocks), while the attention from the municipal government to the particular neighborhood increased two to three fold. 

Results

The outcomes BDLT focused on measuring were quantifiable ones such as the instances of crime and number of emergency calls stemming from each neighborhood, and also certain qualitative, survey-based measures, such as reported governance (whether citizens believe the local government responds to issues) and legitimacy (whether citizens trust the local government to resolve issues, especially in comparison to combo street gangs). 

The overall results from this study were surprising – on average, there was no impact on any measured outcomes. Crime and emergency calls did not decrease, while reported governance and legitimacy did not improve either (it actually may have decreased!).

This has been interpreted as evidence that increased community interventions do not work and are a waste of resources. However, BDLT assessed this data through another perspective. Prior to the intervention, BDLT collected information on the level of local government capacity and presence in all the neighborhoods. BDLT assumed, prior to the experiment, that in neighborhoods where the local government was weak (i.e. limited municipal interventions and participation), the impact of additional local government intervention would be the highest since even the slightest increase in civilian program expenditure would result in a large positive impact.

The results showed the opposite! In areas where the local government had a strong presence and was seen as legitimate, the returns to the intervention were high and positive. Crime fell by 28%, emergency calls reduced by 41% and government legitimacy increased. In neighborhoods with weak local government presence, there was no positive impact. Part of this was driven by the fact that in these neighborhoods, the new liaison officers failed to deliver on promises they made to the communities. The liaison officers were also less likely to show up to meetings, did not refer problems to the necessary agencies, and did not address the issues such as fixing street lights. Many citizens of these neighborhoods were alerted about the local liaison officer meetings too late, did not see any of the new intervention employees or even know about the increased intervention scheme. The graph below shows the relationship between strength of local government and delivering on promises, after increased intervention.

Based on these results, BDLT believe that for these weak local government presence neighborhoods, it might take a longer time of consistent intervention in order to change the perception of the local government. At first, the local intervention in these neighborhoods may not generate any returns (or even generate negative returns when factoring cost), but over time, as local government presence strengthens, positive returns from additional local government intervention can occur. This is known as the S-pattern of returns, shown below.

Whether this is true or not, would require additional research.

Conclusion

The study conducted by BDLT has a lot of valuable insights not only related to the question of local government intervention, but also methodological issues regarding interpreting data and experiment design.

First, the headline results stemming from this study can be deceiving. The reason there was no impact of increased local government intervention was due to the fact that in high local government presence areas, the impact was positive, while in low local government presence areas, the impact was negative. Therefore, the average came out as zero. But clearly, once we condition on local government presence, the study showed that there are positive impacts from increasing local government presence. By just focusing on the headline results, policymakers can end up making wrong decisions. 

Secondly, the above also shows how important it is to consider a wide variety of factors prior to conducting a study and when evaluating study results. Had the authors not collected information on local government presence prior to the study, we would erroneously conclude that increased local government interventions have no impacts.

Regarding the overall impact of increasing local government intervention, this study shows that it is a complicated issue. In neighborhoods that have the weakest local government presence, which city officials would typically want to target the most to improve, initial investments in local intervention might not pay off for a while. Improvements might only be visible after many years of consistent government intervention, and that any initial lack of any improvement might be expected. However, in areas that have higher local government presence, additional investment appears to have large positive impacts on crimes and government legitimacy. Thus, in such areas, additional investment would improve outcomes immediately. 

Naturally, this finding can impact local government decisions – if a mayor of a city wants to see immediate results to improve their electoral chances, they will invest in areas that already have high local government presence and spending. On the other hand, under-invested areas would continue to be neglected since they require multi-year continuous commitments, forever keeping these communities under-invested.

Subscribe now

Share

Interesting Reads from the Week

1. Article: discusses why price-controls (i.e. rent controls, goods price caps) can be a policy tool to tackle certain issues. At least, they should be studied quite a lot more rather than dismissed.

2. Note: points out an interesting new research which shows how liquid wealth alters people’s ability to find jobs, exacerbating wealth and income disparities.

3. Note: Tariffs imposed by the US continue to push US inflation higher, adding 0.7 percentage points to the inflation measure so far.

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

Solving the problem of housing affordability is one of the top social issues. One often-touted solution is removing various supply constraints on housing such as zoning and building regulations. Theoretically, it sounds like such a policy change should lead to a fall in house prices and rents. But recent research by Louie, Mondragon and Wieland (2025) (“LMW”) suggests otherwise.

Supply Constraints

LMW set out to answer the following question – can housing supply constraints, such as zoning laws and other regulations, explain the observed house price and house quantity differences between cities (more specifically, metropolitan statistical areas) in the US. It is commonly believed that the reason certain cities are ‘more expensive’ is that it is much harder for new housing to be built in them due to onerous regulations.

To test this hypothesis, LMW decided to look at how housing supply in a city reacts to changes in housing demand. If something causes housing demand in an area to go up (for example, new job opportunities), we can look at how the quantity and price of housing change in response to this ‘demand shock’. In cities where it is ‘easier’ to build, housing quantities should go up by more and prices should not rise as much as in cities where it is ‘harder’ to build. Visually, this would look like this:

Here we can see the standard supply-demand graph – with price on the y axis and quantity on the x axis. If something such as income growth or better jobs causes a shift in demand (from line D-0 to line D-1), in ‘easy’ to build cities (denoted as orange S-H), we’d be at point B-H (from point A). In ‘difficult’ to build cities (green S-L), we’d be at B-L.

Looking at these shocks to housing demand, we can figure out what the slope of the supply curve looks like – the flatter it is, the more elastic housing supply is; meaning that, for a small change in price, significant new construction will occur (orange S-H line). Cities with greater supply constraints should therefore show larger price growth with smaller quantity growth as compared to cities with lower supply constraints.

Measuring Supply Constraints

Multiple economists have developed measures of supply constraints that include factors such as regulations and geography. Below is a graph that shows the correlation between house prices and the four different measures of supply constraints (referred to as the elasticity of housing supply in the graph):

As can be seen in the above graphs, regardless of which of the four measures we use, as housing supply becomes more elastic (x-axis), i.e. it is ‘easier’ to build, house price growth falls (y-axis). But what happens to quantities?

Do Supply Constraints Matter?

This chart shows the same four elasticities as in the previous chart, but on the y-axis, we have house quantity change. As can be seen, there does not appear to be any pattern – especially not in the expected direction – that cities with higher supply elasticities have more housing (a positive relationship).

Combining house price growth and house quantity growth, in the chart below, we can see that areas that had most housing quantity growth also saw the highest price growth:

This analysis by LMW tells us that supply constraints do not appear to explain the price and quantity differences in housing across cities. Alternatively, although less plausible, the four different measures designed by economists to capture supply constraints, do not actually reflect supply constraints.

So What Explains Differences in House Prices/Quantities?

LMW argue that one of the major explanations of the observed house price differences is income growth in cities. In the charts below, LMW show the relationship between total income growth in a city (annualized percent change after inflation), and house price change (after inflation). In blue are “difficult” to build cities, while in red, “easy” to build cities.

The house price growth relationship is basically identical, regardless of the difficulty in building. LMW illustrate this result with the following example – from 2020 to 2022, house prices in San Francisco grew by 2.4% annually (after inflation) compared to 1% for Houston. This put San Francisco in the 90th percentile of house price growth, while Houston was around the median. At the same time, per capita income in San Francisco grew by 2.2% annually (99th percentile of the US!) while in Houston it was 0.83% annually (10th percentile). In both cases, ‘affordability’ got worse by approximately the same amount.1

This suggests that supply constraints do not really impact ‘affordability’ and it feels like ‘affordability’ appears to be constant – that is, how much one spends on housing is closely related to income.

Housing Affordability – City Data

To see this pattern in the US, let’s look at ‘affordability’ in different US cities. First, it’s worth noting that ‘affordability’ should not only include housing costs, but also transportation costs. If we look at affordability measured as a percent of income spent on housing and transportation, affordability appears to be quite similar across cities:

Moreover, the list of cities that are more affordable may come as a surprise to readers aware of the US housing market. The cities listed as most affordable on this list are often the ones considered to be the most supply constrained. The LMW paper can partially help explain why, as regardless of the supply constraints, these cities have similar affordability to other less supply constrained cities (if not actually more affordable).

So What To Make of All This

The LMW paper caused quite a stir in the economics profession and was immediately heavily criticized for making mistakes. Two counter commentary papers were written quickly after publication – Wiebe (2025) and Furth (2025). LMW responded to both criticisms (here and here) and addressed their concerns effectively. Moreover, LMW did something completely new to me, and issued an FAQ (Dropbox) addressing concerns and explaining their results!

Overall, it appears to me that the typical supply constraints like zoning regulations do not appear to influence affordability much. At Nominal News, we’ve previously discussed a simple outline of a housing model, which argued that deregulation is unlikely to change housing affordability, as demand for houses should be strongly linked to local income, as LMW showed in a much more formal way.

Moreover, I believe that not only does current local income matter for house prices, but also how much people expect income to grow in a city. If you believe wages in a particular city are expected to significantly increase, you’d be willing to pay more for a house today in order to have access to this job market.

It is worth noting that, regardless of the lack of impact on affordability from reducing housing supply constraints, building more homes would still be beneficial. In part, this is because cities may have an ‘agglomeration’ effect, i.e. there may be additional economic benefits just from the fact that people concentrate in a specific geographic area. The agglomeration effect leads to higher productivity (doing more with fewer resources), which pushes incomes higher (which in turn pushes house prices up).

My one key takeaway for anyone interested in housing policy is that there is no easy fix that can make housing more affordable (spending a lower share of one’s income on housing+transport). Suggesting that certain policy changes in certain cities result in large affordability increases is reckless economics.

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

On October 13, 2025, the Nobel Prize in economics was awarded to Joel Mokyr1 – “for having identified the prerequisites for sustained growth through technological progress” – and Philippe Aghion and Peter Howitt – “for the theory of sustained growth through creative destruction”.

The work by Aghion and Howitt is very timely, as it may help explain some of the rise in genAI investment spending. Today, we will focus on this research.

Growth

The main motivation behind the work of Aghion and Howitt is economic growth – the increase in the quantity and/or quality of goods and services produced. Commonly mentioned causes of economic growth are institutions (forms of government and laws) and overall knowledge accumulation.

Aghion and Howitt (1992) (“AH”) looked at knowledge accumulation through the lens of technological adoption and obsolescence. Many new technologies often fully replace former technologies – like email replacing the fax machine, or online streaming replacing DVDs.

This type of technological development leads to economic gains – higher efficiency – but the process itself results in certain losers (fax makers and DVD producers no longer exist). The discovery of new technologies that replace old ones is known as “creative destruction”, a term popularized by Joseph Schumpeter:

“The fundamental impulse that sets and keeps the capitalist engine in motion comes from the new consumers’ goods, the new methods of production or transportation, the new markets,... [This process] incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one. This process of Creative Destruction is the essential fact about capitalism”

Creative Destruction

To model creative destruction, AH propose the following approach.

Suppose the final good – for example, a car – is produced using a worker and a machine (a factory robot). The machine, on the other hand, is produced by a single firm that owns the intellectual property (IP) behind it. This “IP firm” produces this machine using skilled workers. Moreover, this IP firm is a monopoly, as no other firm knows how to produce the machine. This allows the IP firm to earn monopoly profits (i.e. higher profits than a firm in a competitive market).

Lastly, there are ‘research’ firms. These ‘research firms’ aim to develop a machine that’s better than the machine the current IP firm produces. Since this better machine would be preferred by the car manufacturer, the ‘research’ firm that develops this new machine would usurp the monopoly status of the IP firm, becoming the new “IP firm” and earning the monopoly profits.

‘Research’ Firms – the Source of Growth

It shouldn’t be surprising that all the economic growth in this economy will come from the ‘research’ firms that are trying to develop a new machine.

To conduct this research, ‘research’ firms employ the same type of skilled worker as the IP firm. The ‘research’ firms and IP firm are competing for the same type of worker. Moreover, the more skilled workers working at the ‘research’ firm, the higher the chance of discovering a new technology.

Quickly summarizing:

• Final good (car) is produced by a worker and a machine;

• The machine is produced by an IP firm that is a monopoly and uses skilled workers to make the machine;

• ‘Research’ firms try to develop a new machine to become the monopoly IP firm. To conduct this research, they employ skilled workers.

Making Profits

From the above structure of the economy (side note: in economics, we call this the ‘model environment’), it can be seen that what motivates ‘research’ firms to develop a new technology is the hope that they will get to be a monopoly IP firm one day and earn monopoly profits.

So how much do these ‘research’ firms invest in research, i.e. in this particular instance, how many skilled workers should they hire?

Choosing Research Level

The amount a ‘research’ firm will choose to invest really depends on how much profit the ‘research’ firm can make in the future as an IP firm, once it discovers the new machine. This future profit, just like the current profits of the current IP firm, depends on two crucial elements:

• How much the IP firm must pay skilled workers to produce the machine;

• How long the IP firm remains a monopoly, i.e., how long before a new ‘research’ firm discovers a new technology.

It can be seen from these two elements that the more a ‘research’ firm invests in trying to find a new technology, the less it will earn as an IP firm. Why? This is because in this model of the economy, everything is symmetric – any decision made by one ‘research’ firm, will be made by all ‘research’ firms. So the following two things occur:

• Firstly, to increase the likelihood of developing a new technology, a ‘research’ firm hires more skilled workers, pushing their wages up. This in turn means that wages for skilled workers are higher, resulting in lower IP firm profits. Once the ‘research’ firm discovers the new machine and becomes the IP firm, all other ‘research’ firms will continue to try to dethrone it. This means skilled worker wages will remain high, keeping monopoly profits lower.

• Secondly, as the ‘research’ firm hires more skilled workers to increase the probability of discovering the new machine, it means that the rate of discovery of new machines is permanently higher. But that implies that once you become an IP firm, you will be an IP firm for a shorter time, as other ‘research’ firms now also have a high probability of discovering the new machine. Since you will be a monopoly IP firm for a shorter time, you will earn lower profits.

Both of the above elements imply that as you increase spending on research, your potential profit from this research falls.

Outcomes for the Economy

So what is the rate of growth/innovation in this economy? It depends on the parameters of the economy – things like the number of workers, general likelihood of discovering a new technology, costs of capital.2

AH find that the level of research increases when:

1. Interest rates are lower (as this increases the value of all future profits one will earn as a monopoly IP firm);

2. The discoveries are more likely to be more substantive, i.e. the new machine is much more productive in making cars (as this increases future IP firm profits);

3. There are more skilled workers (as these workers are cheaper, resulting again in higher IP firm profits);

4. The probability of a discovery is higher, all else constant, i.e. if we keep the number of researchers fixed, the probability of discovery goes up.

The last point – higher discovery probability – is not obvious. This is because with a higher discovery probability, the ‘research’ firm is more likely to make a discovery (increasing profits), but once it becomes a monopoly IP firm, it is also more likely to be usurped by another ‘research’ firm (decreasing profits). AH show that the former effect dominates, resulting in ‘research’ firms conducting more research when discovery probabilities are higher.

Socially Optimal Research Levels

Economists also like to look at what the socially optimal outcome is. A socially optimal outcome is modeled by assuming that there is a ‘social planner’ that can pick and choose what every person does – i.e. the ‘social planner’ chooses how many people work in ‘research’ firms and how many work in the monopoly IP firm. This is in contrast to a ‘market’ world, where each individual chooses to do what is best for them without taking into account how their decision impacts others. Do you think a social planner would want more research, and thus more innovation or less?

The answer to this question is surprisingly unclear. That is because the social planner realizes that a new innovation makes the old technology obsolete, destroying the value of the old technology. To protect the old technology, the social planner may reduce the amount of research compared to a ‘market’ world.3

Moreover, the more research that is conducted, the more expensive it is to produce the machine since both research work and machine production use the same skilled workers. More research work implies higher wages paid to skilled workers, making machine production more costly.

On the other hand, the ‘social planner’ values the future benefits of new technological discoveries more than the ‘research’ firms. That’s because better machines allow for cheaper final goods (cars) in the future for everyone. Thus, the ‘social planner’ may increase the amount of research. Overall, whether the level of research in a ‘market’ world is optimal is unclear in an AH model, as it depends on the parameters.

Application of AH to genAI

The beauty of the AH model (and probably the reason for the Nobel Prize) is that with the relative simplicity of the model, many insights can be had. The AH model shows the importance of benefitting from discoveries (the duration of monopoly status) and how it can impact research levels. Moreover, the tension between how easy it is to discover new technologies and how much research is conducted is an important factor to consider when designing research incentives. Some of these insights appear to be quite pertinent to our current focus on genAI technologies.

As AH mentioned, if the probability of discovery of a new technology increases (point 4), more research is conducted. Some have posited that the probability of new discoveries within a field of research can go up after a particular breakthrough. We may have seen some of this happen with genAI – after the release of ChatGPT in December 2022, investment in genAI research has skyrocketed. Additionally, another implication from AH’s model based on point 4 is that if the probability of discovery changes over time, it is unclear whether this will lead to more research being conducted. In the extreme, if new discoveries are too easy find, no research will be conducted! This is because you will never benefit from your discovery, since your technology will immediately be replaced by a better one.

The simple AH model may reflect the world of genAI quite well – the maker of the best genAI model may become a monopolist, as everyone will only use their technology. Knowing this, many firms have invested significant capital into research hoping to become the genAI monopoly. Once some of these discoveries occur, the amount of investment in genAI research might start to fall, as new discoveries become less likely. This helps explain the meteoric rise in genAI expenditures.

Subscribe now

Share

Subscribe now

If you would like to support us in reaching our subscriber goal of 10,000 subscribers, please consider sharing this article and pressing the like❤️ button at top or bottom of this article!

Share

The US government has made headlines over the past year by taking an equity stake in Intel Corporation, along with 4 other companies. At the same, the US government has also levied significant tariffs, and continues to issue new debt (the current total debt to GDP is 125%).

So how are all of these topics connected? They’re all a form of tax.

The Intel Stake

One of the more unique recent government policies was the announcement in October by the US government that it is taking a 10% equity stake in the computer chip company Intel Corporation. It has been framed as a ‘fair’ offset for the subsidies the company receives through government programs.

On first glance, it might sound reasonable – the US taxpayer subsidizes Intel (note: a subsidy is a negative tax) and, in return, the US taxpayer gets 10% equity. But in reality, it’s probably an inefficient way of taxation. Let’s look into how societies fund themselves.

Government Budgets and How We Fund Ourselves

Although governments are often presented as a separate entity from society, it is ultimately “us” that control the government. From a big picture perspective, “we” (via the government) decide that we should provide ourselves with certain goods and services – like education, healthcare, defense.

To fund these goods/services, we collect money from ourselves, done most commonly via taxes (but also via issuing bonds – a promise to pay back the person that gave the government money).

“We” can choose exactly how to do this – we can set tax rates and tax types (income taxes, corporate taxes, congestion taxes, etc). In theory, “we” are completely free to do as “we” choose with regards to what and who “we” tax. Anything that raises money for “us” (the government) can be interpreted as a tax.

Taking the Intel Equity Stake

Thus, taking a 10% stake in Intel by the government is a tax. What type of tax? It’s simply a wealth tax – a tax which is often derided.

How is it a wealth tax? By forcing Intel to give the government 10% of the equity in the firm, Intel has to dilute all current Intel shareholders – the current Intel shareholders give up 10% of the company, making themselves 10% poorer by construction. Thus, taking this equity stake is putting a very specific 10% wealth tax on Intel shareholders.

Should We Care?

The main reason economists care about taxes and tax types is the efficiency implications of taxation. Any form of tax levied on an economic activity potentially distorts the decisions of the people impacted by the tax. An income tax, for example, may discourage work, as a worker gets less money for their work and thus may choose to spend more time not working. A capital tax (a tax on investment earnings) discourages savings and investment. Taxes can also distort decisions positively – a congestion tax discourages usage of cars which generates many benefits (less traffic, less pollution, fewer accidents).

So economists mainly focus on figuring out how much taxes actually distort decisions and how this affects our total output. This is called ‘efficiency’.

Tariff Inefficiencies

Tariffs are a tax – a tax on goods or services being brought in from abroad. This tax, especially on intermediate inputs such as steel or wood, can significantly distort the decisions domestic firms and people make, resulting in a significant efficiency loss. Suppose without tariffs, firms would choose to do A, while with tariffs they choose B. Since firms could choose B before tariffs were imposed, the fact that they didn’t choose B suggests they were better with Option A (i.e. more efficient).

The reason we see nearly all reasonable economists criticize tariffs is because of this fact – tariffs have significant distortions and reduce economic efficiency. Economists will argue that there are better taxes to use to raise revenues for the government.

Taking Equity is Probably Inefficient

Coming back to the Intel equity stake tax, it is probably a very inefficient tax. Picking a firm to impose a wealth tax on its shareholders may have significant impacts on how people save and invest in the future. For example, raising money for firms may become costlier, as investors may be wary of having to pay a wealth tax. This can result in smaller firms with less hiring, shrinking the economy.

If the purpose is to raise revenues from investors, then it would be much simpler for the government to raise the tax on dividends or capital gains – a tax which would be less distortionary compared to equity taking.

Regardless of one’s preferences, the overarching point is that taking an equity stake is a tax.

Government Debt

This brings us to the last topic – government debt. In the beginning of today’s post, I mentioned that any form of revenue raised by the government is a tax. Government-issued bonds are similar – the government, or “we” take money from ourselves.

The only difference is that “we” promise the specific people who gave “us” the money a negative tax in the future in the form of interest payments and a return of the amount of money the person gave “us”.

When governments raise revenues by traditional taxation, we also collect money from specific groups of people and then provide goods/services to these or other people (education, healthcare, infrastructure). Note the sole difference with government debt is the very specific contractual obligation (the lender to the government will get x% and principal back), rather than an implicit society agreement (the taxpayer will receive certain benefits)

Repaying Debt

A big topic of concern is the ability of a country to repay the debt. When we talk about repaying the debt, what we fundamentally mean is whether we give “ourselves” the negative tax (interest payments, principal).

A government can repay the debt by doing three things – raising taxes, cutting spending or defaulting. If it raises taxes, then “we” are contributing money to pay off “ourselves” (or more specially a subset of “us” – the bond holders). If it cuts spending, then “we” are no longer receiving a service/good, i.e. we removed a particular negative tax.

Lastly, if “we” default on our debt, then in effect, we are raising a wealth tax on “ourselves” (more specifically, the bond holders) to pay off our own bond holders.

As you may notice, the main impact of these three approaches is who ends up getting taxed – i.e. a distributive question. If we raise taxes – someone’s paying. If we cut spending -- someone’s negative tax amount decreases (a form of tax raise). If we default, we tax the bond holders.

Thus, whenever you hear people talking about ‘reigning in the deficit’ or ‘cutting debt’, it is essentially a discussion about re-distribution.

It’s All a Question of Tax

I hope the above has shown how everything boils down to who is taxed and whether a particular tax is efficient. When the US announced the receipt of an equity stake, it was a tax on the equity holders. Tomorrow the US could announce a 10% tax on all government bond holders and reduce the outstanding government debt by 10%. It would mechanically be no different from the equity stake (from a financial perspective).

Naturally, instead of either of the above, the US can just raise traditional taxes – income taxes, property taxes, etc. – to pay for the debt. Unfortunately, we often see inefficient taxes, like tariffs, get implemented, because they sound good, but are really very costly.

Subscribe now

Share

Next Week Preview

On October 13, 2025, the Nobel Prize in Economics was awarded to Joel Mokyr – “for having identified the prerequisites for sustained growth through technological progress” – and Philippe Aghion and Peter Howitt – “for the theory of sustained growth through creative destruction”.

We will go over Aghion’s and Howitt's contributions to our understanding of economic growth, and how competition may slow growth!