Nature of Budgets
Earlier, I focused on the US government’s infrastructure spending bill. One issue highlighted during the negotiation of that bill was how much it would cost and the impact on the US budget. The way the program cost was publicly presented (as a 10 year cost) was not accidental – it was driven by the way the US does its budget. Although the way governments construct the national budget might appear to be a topic that is not important and already resolved, that is far from true. By budget construction, I am not referring to the political questions around budgets (i.e. how much to spend on which priority), but how the budget is constructed (for example, the current US budget lists revenues and expenditures by department for the upcoming year).
The way national and local governments undertake the spending and budgeting processes ends up significantly influencing government decision making. We are often unaware of the consequences that the form of the budget has on these decisions. In order to understand the issues stemming from the form of budget, and discuss alternatives, it is helpful to define what is the current way of budgeting.
The process of government spending is almost universally conducted via established budget procedures and rules. It is not done, for example, in an ad hoc fashion, where governments could decide at any time what to spend. Typically, we think of government budgets as an income and spending list for one year, divided into different programs and departments. As defined by Aaron WIldavsky, “Traditional budgeting is annual (repeated yearly) and incremental (departing marginally from the year before). It is conducted on a cash basis (in current dollars). Its content comes in the form of line-items (such as personnel or maintenance).” Most governments and local governments follow this style of budgeting.
The US Federal Budget, which is a traditional budget, is set on October 1 each year for the next fiscal year and allocates spending by department, even for programs and projects that are clearly cross-departmental. Furthermore, due to US budgetary rules and procedures, 10-year forecasts for both revenues and outlays are estimated. However, these two choices: one year budgeting and spending by department are not as innocuous as they seem, as there are many different ways a budget could be structured.
Budget Types
The traditional budget approach focuses on one calendar year revenue and expense forecasts. Many government projects, however, do not fall neatly into this time frame. Thus, it comes with no surprise that such a limited time frame can cause issues in multi-year undertakings, such as infrastructure spending. Since a new budget is written each year, the fate of certain multi-year projects is never guaranteed.
To counter this problem, alternative budgeting methods have been suggested and implemented at different levels of government. One such proposed change is a different time frame for the budget – rather than having a budget for one year, budgets could be designed to span multiple years. Budgeting over a period longer than a year can result in more efficient allocation to spending projects and increase the likelihood of completing these projects. A key drawback though, is that forecasting longer time periods is generally a difficult endeavor, with forecasts often being significantly off from reality. This can lead to spending shortfalls if government revenues are reduced or project costs increase unexpectedly. Some of these issues are also present in the current US budget procedures, as although the budget is set for one year, the mandatory 10 year forecasts that are used to assess the long term impact of new government programs are estimates with very low certainty.
In the US Federal Budget, the spending procedures are further complicated by the division of spending into three categories – mandatory, discretionary, and tax expenditures1 (Kasdin, 2018) – as often different spending priorities can fall in more than one category, while each category requires very different legislative procedures to undertake the spend.
Another key budgeting choice is whether to allocate funds by a specific project (for example, building a specific hospital) or by department (allocating money to the Department of Health). Generally, funding is allocated to departments, which then decide how to distribute the funds across their responsibilities. This leads to potentially several issues. One – if the department budget is insufficient, certain projects may never be completed or significantly altered from what was agreed upon to fit within the allocated budget. Two – departments may have an incentive to spend all their allocated funds in order to not have their budget reduced in next year’s budget plan, even if the spending would be wasteful. Three – departments do not internalize the fact that certain decisions they make can impact budgetary costs of other departments. For example, the Supplemental Nutrition Assistance Program (SNAP) spending has been shown by Hong and Henly (2020) to have significant effects on reducing malnutrition, which improves both health and educational outcomes. Thus, any reduction in the SNAP budget will potentially increase the costs incurred by the health and education departments. These cross-impacts are not easy to estimate and the US 10 year forecasts does not address them, even though their effects might be significant for estimating the cost and benefits of certain government spending.
The Optimal Budget?
The question of how the optimal budget should be formed has been studied in a limited fashion. One study by Kasdin (2016) attempted to create a framework to evaluate budget type proposals. Kasdin argues that a budget process should have several key goals that it needs to meet such as:
macroeconomic stabilization (keeping macroeconomic volatility low);
allocative efficiency (the ability of the government to shift spending if necessary);
aggregate efficiency (the capacity of the budget to control spending);
timeliness (budgets are made in a routine manner); and
participation (reflecting the preferences of the public).
Based on this framework, altering the current US 1-year budget to a partial 2-year structure would probably not lead to a significant improvement.2 However, merging certain US budget procedures (such as the authorization process and appropriation process3) could lead to significant improvements.
The question on optimal budgeting has not been studied extensively. Most research on this topic was conducted a few decades ago. It would seem helpful for academics to revisit the question of how a budget should be structured given the advancement in forecasting techniques, technological coordination, and improvement in both data collection and analysis. Some of the budget limitations, such as cross-departmental impacts, could probably be overcome today.
Although budgets might not be too interesting or exciting of a field of study, the impact of our accounting method on societal outcomes is significant. The budget type impacts how we perceive spending, which influences our decisions, and also changes the likelihood of accomplishing public goals. Therefore, the budget type is not an innocuous decision.
Tax expenditures are various forms of tax credits to individuals and businesses.
A biennial budget would improve timeliness and transparency of the budget, potentially also increasing participation. However, the biennial budget suffers from difficulties of responding quickly to sudden changes in the economy, reducing allocative and aggregate efficiency.
The authorization process generally refers to spending that is mandatory such as entitlement programs like Social Security (retirement benefits). The appropriation process refers to discretionary spending, where there is a need to determine the ‘appropriate’ level of funding for various legislative goals.