The Uses and Misuses of Budget Data

Alternative presentations of the same budget data tend to offer very different impressions, creating opportunities for a deeper understanding of fiscal health. However, these measures sometimes reflect hidden assumptions about government finances, so even a seemingly neutral way of presenting data often isn’t so neutral. The ability to give different impressions with the same budget data creates the opportunity for policy mischief, as one can tell very different stories about fiscal policy depending on the measure used. We can assess the appropriateness of various spending adjustments by understanding the underlying assumptions in the measures, how to use the measures analytically, and how they might be used strategically. The paper looks at measures of government spending over time, as well as budget forecasts, to demonstrate this logic in practice. It concludes with a case study of President Obama’s fiscal year 2014 budget.

How can analysts using the same set of budget figures come to dramatically different conclusions about the course of federal government spending? It’s because they start with different notions of what’s important in the spending outlook. While their presentations may all be equally accurate, it is important to understand that each reflects various assumptions about government finances, and is thus subject to manipulation.

In a new study published by the Mercatus Center at George Mason University, David M. Primo reviews the various measures of government spending and examines how they can be properly or improperly employed. Below is a brief summary. Please see “The Uses and Misuses of Budget Data” to read the study in its entirety and learn more about the author.

 KEY POINTS

Budget Presentations

The same budget data can yield different conclusions about the condition and course of the federal budget—depending on the way the data are presented.

  • A model using absolute dollar figures will show a dramatic increase in spending over several decades. Adjusting the same figures for inflation, however, allows for more accurate comparisons of spending levels over time. Any trends that remain are due purely to the growth in government and not to the changing value of the dollar.
  • Another option is to measure government spending per person. Spending over time will appear to be declining using this measure if the population grows faster than spending, even if spending itself has increased in absolute or inflation-adjusted dollars.
  • Presenting the figures as percentages of gross domestic product (GDP)—that is, as shares of national economic output—allows one to evaluate the weight of government spending on the private sector. Here, too, spending will appear to be declining if the economy is growing faster than spending.
  • When the population or the economy is growing, measuring trends using population- or GDP-adjusted spending may mask the magnitude of spending increases.

Baseline Bungling

Budget presentations may also vary depending on the starting point used for measuring budgetary changes, known as the “baseline.”

  • The Congressional Budget Office (CBO) uses the “current law” baseline to show the budgetary implications of existing law, including changes in law that are scheduled to occur.
    • For example, CBO must assume a reduction in Medicare physicians’ payments, as required under existing law.
    • But Congress has regularly rejected these cuts in an action called the “doc fix,” which then results in higher Medicare spending than the current law baseline.
  • The CBO also produces an alternative scenario using a “current policy” baseline, which shows what will happen if existing policies are maintained indefinitely without change. Continuing with the Medicare example, CBO uses this baseline to show what will happen to Medicare spending if Congress acts as it has in the past and passes a “doc fix” to cancel the scheduled spending cuts.
  • Neither of these models is intended as an exact prediction, because Congress may change laws and policies; they are simply intended to show the likely course of spending under different scenarios.
  • Any model can create confusion about what constitutes an “increase” or a “cut” in spending; it depends on what baseline is used for comparison.

A Case Study: The Obama Budget

The fiscal year 2014 Obama budget can be presented in different ways—absolute dollars, inflation-adjusted dollars, inflation-and-population-adjusted dollars, and percentage of GDP. Thus, its spending may be seen as either increasing or decreasing depending on which set of values is used.

  • While none of the analyses are wrong, they can be misused if their underlying assumptions are hidden, or if those citing the analyses fail to explain their various nuances.
  • In addition, because the president’s budget—like congressional budgets—shows only 10 years of spending, it fails to reflect the long-term implications of current laws and policies.
Mercatus AI Assistant
Ask questions about this research.
GPT Logo
Mercatus AI Research Assistant
Ask questions about this research. Mercatus Chatbot AI More Details
Suggested Prompts:
Ask us anything. We use OpenAI's ChatGPT 4o base model to answer any question about Mercatus research.