Electoral Issue #3: Immigration
Improving authorized migration may be better at reducing unauthorized immigration than additional border enforcement.
Thank you for reading our work! Nominal News is an email newsletter that focuses on the application of economic research on current issues. Subscribe for free to stay-up-to-date with Nominal News directly in your inbox:
If you would like to support us further with reaching our 1,000 subscriber goal by year end, please consider sharing and liking this article!
With the US presidential elections coming up in November, several issues have come to the forefront. Over the next couple of months, we will focus on what economic research has to say about these issues. We will not be focusing on or assessing specific policy proposals, but rather on what the economic literature can contribute to our understanding of the issues. Although the inspiration for these articles is the US election, these issues are broad enough to impact many countries.
Unauthorized immigration (or ‘illegal immigration’) is a topic discussed in many countries. In the US, there are approximately 8.4mln unauthorized immigrant workers, while in Europe, the number is less than half. Typically, the most frequently discussed way of reducing unauthorized immigration is strengthening border enforcement or making authorized immigration more complex. Economics can tell us why these methods are unlikely to work, and are even counter-productive.
I. Modeling a Decision – A Short Intro to Economic Modelling
Kovak and Lessem (2020) (“KL”) looked at how policies could alter the number of unauthorized immigrants. To address this question, KL first built a model to capture the decision making process an individual makes when choosing whether to immigrate from one country (in this paper, it is Mexico) to another country (the US). In this and the next section, we will discuss how such a model is built and in the third section, we will go over the results.
Location and Authorization Status
Each individual in the model knows where they are currently located, as well as their status regarding being authorized to work in the US (either they are authorized or they are not). This status can change – a person unauthorized to work in the US can, with some probability, gain authorization. Without authorization, an immigrant that enters the US faces a deportation probability.
Cost of Moving
An individual also knows the ‘cost’ of moving between countries. Note, the ‘moving cost’ does not have to be identical – that is moving to the US from Mexico can be more or less expensive than moving back to Mexico from the US.
Value of Each Location
Lastly, the individual can approximate the overall value they will have from either staying in their current location or changing locations. This value can be influenced not only by income, but also by personal preferences such as preferring to be in Mexico or the US, proximity to family, etc. Thus, this value captures both financial and non-financial benefits of a location.
Value ‘Shock’
This overall value from being in a location from time to time gets ‘shocked’ – i.e. the value of living in a particular location can change periodically, reflecting the fact that income opportunities can change, or other life events can alter a person’s perspective of being in a location.
Decision
Using all the above facts, each time period (for example, each month), an individual decides their optimal strategy of whether to change locations or not. The individual can choose to remain in the country they are currently in or they can choose to cross the border, paying the ‘moving cost’.
II. Fitting the Model to Data
In the above model, an individual’s decision will be influenced by the values of the model inputs listed above. For example, if the moving cost is lower, an individual is more likely to move. If the value of a location is higher, the individual is more likely to move to that location. So how are these values determined?
Parameters
The values for these model inputs are typically assumed to come ‘randomly’ from a probability distribution. This is because, for example, we cannot ask every person what their ‘moving cost’ is. Instead, we simplify this exercise by assuming the value of the ‘moving cost’ is ‘randomly’ drawn. The probability distribution the number can be drawn from is described by ‘parameters’.
For example, the Normal Distribution is described by two parameters – the mean and the variance (a measure of dispersion, how far numbers are from the mean). The outcome from a six sided die roll is also a probability – specifically, a discrete uniform distribution. The parameters of a die roll are its lower bound value (1) and its upper bound value (6).
Estimating Parameters
Since all the model inputs are influenced by probability distributions and therefore by parameters, all we need to do is figure out what parameter values to use in the model. To do so, economists look at data. In our case, one data set KL looked at is the immigrant flow data – how many people move to the US from Mexico and how many immigrants leave the US and head back to Mexico.
Using this data, KL were able to estimate what parameters of the probability distribution in the model were most likely to generate these observed immigration flows. That is, under what specific parameter values in KL’s model, was the observed data most likely to occur (this is called Maximum Likelihood Estimation). The method of doing this is relatively simple – we guess the parameters, then observe how individual’s make decisions in our model, and if their decisions match what we observed in the data, we have the answer!
III. Running Experiments – How Policy Impacts Unauthorized Crossing
Armed with the above model along with the parameter estimates of the model, KL can now run experiments by looking at what would happen if certain policies changed (i.e. if deportation and legalization rates in the model were changed).
Baseline Current Policies
First, under the current data, the deportation rate ranges from 1%-3% (an unauthorized immigrant in the US has a 1%-3% chance of getting deported) while the legalization rate is about 0.05% (the probability a potential individual interested in migrating could get US authorization). Moreover, KL found that authorized, or legal, immigrants and unauthorized, low education (0 to 5 years of education) immigrants have the highest value of working in the US. High educated (6+ years of education), unauthorized individuals generally prefer to remain in Mexico.
Moreover, using the fact that US border enforcement expenditures have varied over time, KL could estimate how much additional border expenditure impacts immigration patterns. Increasing border enforcement by $10bln (currently, the US border enforcement budget is close to $20bln), only increases the moving cost for potential unauthorized immigrants by about 5%. Thus, it appears border expenditure has limited impact on whether a person chooses to undertake unauthorized immigration.
Changing Policies
KL experimented with changing the deportation rate, changing the legalization rate and providing a new temporary visa.
If deportation rates were increased to 10% (from the current level of 1%-3%), the probability of a migrant choosing to move to the US as an unauthorized worker falls by about 30%, while the total number of unauthorized migrants present in the US would be cut in half.1 However, it is worth noting that the cost of enacting such a policy is unclear and could be significant, much larger than the current $20bln border enforcement budget.
On the other hand, increasing the legalization rate to 1% (from 0.05%) would only reduce the probability of unauthorized migration by 5%. However, this would reduce the number of unauthorized migrants by about 15%. This is due to the fact that waiting for authorization is more preferable than losing the right to a legal pathway in the event of getting deported. At the same time, with the increased legalization rate, the number of authorized workers would significantly increase from the current amount. Overall, the population of authorized and unauthorized migrants inside the US would double from the current rate, with the outsized increase of authorized migrants offsetting the decrease of unauthorized migrants.
Temporary Visa
KL, however, noted that 65% of immigrants from Mexico (whether authorized or unauthorized) stay in the US no longer than 3 years, suggesting that most are not interested in permanent migration. Thus, KL experimented by proposing a temporary visa program. In this program, immigrants would be allowed to apply for a temporary 3-year work visa, which an immigrant could obtain with 5% probability (significantly higher than the current 0.05% legalization rate).
Under this temporary visa scenario, the total number of unauthorized immigrants falls by approximately 20%, while the total number of immigrants (both authorized and unauthorized) remains the same as it is currently (unlike the the previous example, where the authorization rate was increased to 1% which doubled the number of individuals in the US, as this was a permanent legalization). This temporary visa program appears to have significant benefits, even with what can be seen as a very low probability of being granted a visa.
IV: Increasing Legal Avenues
So why does increasing legal entry probability appear to reduce unauthorized immigration? The intuition is as follows. For a potential immigrant, if the probability of getting authorization is almost zero (as is now), there is no real adverse impact of being deported if the immigrant enters illegally. The entry ban due to deportation simply turns the already low probability of 0.05% into 0%, which has negligible impact on decision making.
However, implementing a legal temporary visa that gives an immigrant a 5% probability of getting a visa encourages the immigrant to apply via this program rather than enter the US illegally. If the immigrant were to be deported, they would lose the privilege of applying for the temporary work visa, which is a real cost to the immigrant. Under the current probability of 0.05% of getting legal authorization, there is practically no deterrent effect of being denied the ability to gain legal status once an immigrant is deported.
Reducing unauthorized immigration that is primarily for work purposes also has the benefit of making it easier to identify the individuals that are crossing the border illegally for drug and weapons smuggling purposes. KL’s research appears to demonstrate that there is significant value in improving legal migration avenues, even in a very minor way, as it both reduces unauthorized migration and may also have other economic benefits for both countries (for example, increased tax revenues due to authorized status). Unlike other solutions that may be much more costly, a temporary visa program may be far more effective and cheaper to implement.
Interesting Reads from the Week
- discusses his newly released research paper on how Generative AI may have already impacted the labor market. Moreover, the historic job polarization (growth in high and low income jobs), a trend observed in the 90s and 00s, appears to have slightly subsided in the 10s.
- talks about the development of the Bangladesh economy and the growth of microfinance, fishing and garment industries.
US Labor Market Update:
and gives us their US labor market updates for the month of September. Links here and here.
The Electoral Issue Series:
Electoral Issue #1: Investing in the Future – Child Tax Credit
Electoral Issue #2: Why We Demand Bad Policy
Electoral Issue #3: Immigration
Electoral Issue #4: Minimum Wages
Electoral Issue #5: Tariffs, Tariffs…
The total number of migrants is measured in person-years, that is if an unauthorized migrant would stay for 2 years, they would count the same as if 2 unauthorized migrants stayed for a year each.
There is something funny about the relation of expenditure on border enforcement and it effect on illegal immigration. "Border enforcement" is very heterogeneous. A "wall" v an army of asylum processing agents are very different.
Also, the model seem pretty reasonable for Mexico, but the asylum issue is not from Mexico