Fudging Predictions on the Impacts of Government Spending is Surprisingly Easy
June 27, 2017
Politics, at its heart, is a number puzzle.
Even beyond the counting of votes, delegates and electors, lawmakers need to know how every new law, budget, executive order and legal ruling will affect citizens and the economy.
Take the battle over the Affordable Care Act and the American Health Care Act as an example. Before Congress voted on the American Healthcare Act the first time, the Congressional Budget Office reviewed the law to see how much it would cost the Federal government, how many people were likely to benefit and how many people were likely to be burdened.
Other organizations took that information and determined how the new law would likely affect things like public health, the overall American economy and how much money Americans would have in their pockets. Everyone from the legislators voting on the bill, to special interest groups and citizens watching the news rely on these numbers, and ones calculated for the Affordable Care Act to decide which plan they should support.
It is impossible to predict down to the dollar exactly what effect a given law will have, but economists can use models to estimate the effect of the law, and those are the numbers we see on the news. New research from NC State University, however, shows that one of the most general far-reaching models used by economists worldwide may not be as accurate as advertised.
NC State economist and co-author of the study Nora Traum and colleagues from Indiana University reviewed the most popular models for how government spending relates to gross domestic product (GDP)—the sum of all the goods and services produced in a country in a given time period—and found that it can be significantly fudged to suit whatever interest is producing the model.
The relationship between spending and GDP is a lot like the relationship between investment and profit. Using the example of the Affordable Care Act, the idea is that spending a dollar subsidizing a health insurance plan leaves a dollar in the pocket of the person with the subsidy that he can spend somewhere else, or use to buy a better insurance plan that could make him or her healthier and more productive.
Economists and interest groups could use a model to work out whether the United States would see more or less than a dollar’s worth of return for each dollar invested. The researchers found that there are several assumptions worked into the model, like how the government would choose to pay back any money it has to borrow, which can affect both the investment and return numbers.
Traum and her colleagues found that by tampering with the assumptions, even within their normal limits, one can dramatically change whether the return is positive or negative.
This research is important because this particular model gets at one of the most fundamental debates in American politics: whether spending more Federal money will benefit or hinder our economy. It is a question that affects everything from healthcare, to infrastructure, taxes, the military and scientific research. Those of us without advanced economics degrees might use models to help us decide whether we think some of these programs are worth funding. That could determine how we vote or whether we decide to protest or write our Congressmen.
The researchers developed a new version of the model that minimizes the influence of the assumptions and allows for less play in the numbers. They published that model, and their other results in the journal American Economic Review. Their model predicted that every additional dollar spent by the Federal government would return about $1.30 in GDP in the short term.
If the new version of the model gains traction, we may get a more accurate picture of new laws and how they might affect the U.S. bottom line in the months and years to come.