March 5, 2015

March 5, 2015 | Issue Brief on Budget and Spending

CBO Should Update Its Methodology Before Dynamically Scoring Spending Bills

Congress could soon ask the Congressional Budget Office (CBO) to dynamically score a new transportation bill or other spending bills it will consider. Dynamic scoring has been a frequent topic of debate in recent months. It considers how major pieces of legislation affect the macro economy, whereas current static scoring does not.

The House passed a rule earlier this year instructing the CBO and the Joint Committee on Taxation (JCT) to dynamically score legislation, including both tax and spending bills, that has a sizeable effect on the economy.[1] The debate continues because it remains unclear how and when the CBO and the JCT will carry out those instructions.

The CBO and the JCT should dynamically score spending and tax bills, respectively, that have large macroeconomic effects. When it applies to spending, however, the CBO must do the dynamic scoring completely. Otherwise, the result will wrongly inflate the benefit of government spending.

The CBO Overestimated the Benefits of the Stimulus

In the past, the CBO has not dynamically scored spending correctly and therefore overestimated the benefits of certain government spending programs. A high-profile example was the 2009 American Recovery and Reinvestment Act, better known as the stimulus.

Shortly after Congress passed the stimulus in 2009, the CBO estimated that by 2010 it would increase gross domestic product (GDP) between 1.1 percent and 3.4 percent, reduce the unemployment rate by between 1.9 and 0.6 percentage points, and increase employment by between 3.6 million and 1.2 million.[2] The economy’s prolonged anemic state in the years after the recession proved the stimulus fell well short of the CBO’s prediction. For instance, the unemployment rate at the end of 2010 was 9.3 percent,[3] which was 0.6 percentage points above the CBO’s baseline estimate of what it would be had the stimulus never become law.

Methodology Is One-Sided

The CBO’s estimate of the stimulus was flawed because it only assumed the beneficial impact of higher government spending. It did so by using the multiplier effect. That effect shows how an extra dollar of spending by the government increases economic activity by more than a dollar because it triggers a cascade of additional spending.

For instance, if the government spends money on a new road, the construction workers building it have more money to go out to dinner, which means the wait staff at the restaurant have extra money to spend as well, and so on and so forth. The additional layers of spending are what lead to higher levels of growth. Different types of government spending have different multipliers.

The multiplier effect does not work as cleanly in reality because government cannot spend money without first taking it out of the economy, either by taxing it away from families and businesses, or borrowing it from the same credit markets they use. When government takes money out of the economy, it prevents the private market from spending it and therefore prevents a similar chain reaction of spending from occurring. This foregone chain of spending is known as the negative multiplier.

The positive and negative multipliers are equal in magnitude, but opposite in sign. As such, they cancel each other out. At best, this means government spending has no impact on the economy. Since government spending must go through the bureaucracy before it is spent, it is likely that much of it has a negative impact on the economy.

The CBO overestimated the impact of the stimulus because the same dollar cannot be spent twice, which is in effect what the CBO assumes when it scores spending bills without considering the negative multiplier. To accurately describe the net effect of additional government spending, the CBO needs to subtract the negative multiplier from the positive multiplier. By failing to do so with the stimulus, it only told half the story and biased its estimates in favor of the stimulus.

Positive Multiplier Never Has a Chance to Take Hold

Accounting for the negative multiplier would be a significant improvement in the CBO’s modeling. However, it would still give too much credence to government spending because it assumes that the positive multiplier actually occurs.

In reality, the demand destroyed by the government taking money out of the economy immediately snuffs out the demand created by the spending. As such, the positive multiplier never has a chance to get started.

In contrast, once the government takes money out of the economy to spend, the negative multiplier does take hold and has a lasting negative impact. The private foregone investment is never realized, leaving a permanent hole in the economy.

The CBO Repeated Error Again and Again

Although the stimulus has long passed from the consciousness of the American public, largely because it had such a negligible impact on the economy, the CBO is still required under that law to assess the law’s impact. Quietly, it recently released its latest report on how the stimulus supposedly helped the economy in 2014.[4]

This most recent analysis was not based on actual measurements of economic activity. Instead, like all the CBO’s previous periodic analyses, it used the model it originally used to estimate the impact of the stimulus before it became law.

The CBO used the model to predict what would happen if Congress passed the stimulus. Now, many years later, the CBO is using the same model to say that because the law did what it said it was going to do, it had the effects the CBO anticipated it would have. This is faulty methodology because it assumes the result before running the analysis, committing the fallacy of begging the question.[5]

Same Flaw Would Bias Dynamic Score of New Spending

Since Congress could soon ask the CBO to dynamically score new spending bills, the CBO’s error in estimating the impact of the stimulus is still relevant because the methodology it used remains unchanged. Unless the CBO updates its methodology to account for the negative multiplier, it will overestimate the impact of those bill’s effects on the economy the same way it did the stimulus.

It is also possible that the CBO would report rates of return on government spending that goes to investment, such as on roads, bridges, and other transportation infrastructure in a dynamic score. This is a useful statistic for lawmakers to know, and if the CBO has the capabilities, it should include those returns in its report.

However, to have any meaning, lawmakers would also need to know what return the money would have earned had it been invested by the private market. The CBO would also need to report the private returns the money the government spent could have earned had the private sector invested it so that lawmakers can weigh the relative economic benefit of spending the funds through the public sector versus allowing the private sector to allocate those resources.

The baseline the CBO uses will also be important when assessing spending bills. The CBO’s economic forecast assumes a certain baseline of government spending. Maintaining spending up to that level would have no new economic impact, positive or negative, given that assumption. A dynamic estimate should only evaluate the economic impact of spending that exceeds or reduces baseline spending. For instance, if the transportation bill maintains the same level of spending as the CBO already assumed it would, then there is no dynamic impact by definition. If the CBO reports a strongly positive impact, it would be because it assumed that all federal transportation spending would stop without a new bill, which is unrealistic.

Dynamically Score Spending the Correct Way

Dynamic scoring is proper for both spending and tax legislation. If the CBO dynamically scores spending bills, it should update its models to account for the negative multiplier. If it does not, it will improperly inflate the benefit of such spending and bias its estimates in favor of bigger government.

—Curtis S. Dubay is Research Fellow in Tax and Economic Policy in the Thomas A. Roe Institute for Economic Policy Studies, of the Institute for Economic Freedom and Opportunity, at The Heritage Foundation.

About the Author

Curtis S. Dubay Research Fellow, Tax and Economic Policy
Thomas A. Roe Institute for Economic Policy Studies

Show references in this report

[1] H. Res. 5, 114th Congress, January 6, 2015, (accessed February 25, 2015).

[2] Douglas W. Elmendorf, Director, Congressional Budget Office, letter to Senator Charles E. Grassley, March 2, 2009, (accessed February 25, 2015).

[3] U.S. Department of Labor, Bureau of Labor Statistics, “Labor Force Statistics from the Current Population Survey,” (accessed February 25, 2015).

[4] Congressional Budget Office, “Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output in 2014,” February 2015, (accessed February 25, 2015).

[5] See, for example, Curtis Dubay and Patrick Louis Knudsen, “CBO’s Stimulus Review: As Good as a Horoscope,” The Daily Signal, February 27, 2012,; Curtis Dubay, “CBO Stimulus Report Begs the Question,” The Daily Signal, May 31, 2012,; Curtis Dubay and J. D. Foster, “CBO Appears Immune to Embarrassment,” The Daily Signal, August 24, 2012,; and Curtis Dubay, “CBO Again Repeats Faulty Methodology to Estimate Impact of Stimulus,” The Daily Signal, February 21, 2014,