Close Nav

The 'Government Sector of the Economy' Canard


The 'Government Sector of the Economy' Canard

June 7, 2012

A recurring theme of commentary appearing in the New York Times and elsewhere is that the government is shrinking as a share of the economy. According to Floyd Norris, “the government sector of the American economy has shrunk during the first three years of a presidential administration” for the first time since the 1960s. This dangerous decline in government spending is thought by Norris and others to be responsible for the slow growth since the recovery officially began in July 2009.

President Obama’s economic policy can be criticized on many fronts but insufficient government spending is not one of them. The most straightforward way to assess the burden of government is by comparing total government outlays to gross domestic product (GDP). By this standard, the federal government is currently larger than at any point in post-War history. Between fiscal years 2009 and 2012, federal outlays averaged 24.4% of GDP, the highest government-to-GDP ratio since 1946 and 22% (4.4 percentage points) larger than the average size of government since the military demobilization following World War II of 20.0%.

Rather than measure actual government spending, which is at record levels, Norris instead focuses on the “government sector of the economy,” which is different conceptually from the actual size of government. Each dollar of government spending does not automatically contribute to GDP. The Bureau of Economic Analysis (BEA) only counts direct government purchases of goods and services or investments in capital equipment or infrastructure in the National Income and Product Accounts (NIPA). When the government buys an airplane from Boeing, for example, it counts the same in the GDP as if the aircraft were purchased by United Airlines. The same accounting generally works for employee salaries, as government purchases of services provided by Commerce Department employees, for example, count as government consumption spending in GDP.

It is important to note that the “government sector of the economy” argument is substantially different from the one made in a recent MarketWatch piece, which contended that President Obama was not responsible for the large spending increase. This piece inappropriately attributed fiscal year 2009 spending to President Bush simply because the fiscal year began October 1, 2008. The Obama Administration’s reluctance to return spending to 2008 fiscal year levels, as recommended by the Bowles-Simpson Commission, suggests that either government spending has increased substantially or the President is not willing to countenance relatively small cuts to discretionary spending.

The difference between government spending and the government contribution to GDP is largely attributable to transfer payments, entitlements, and subsidies. In recent years, transfer payments have exploded upwards. As of March 2012, transfer payments were running at a $2.3 trillion annualized rate, or 15.2% of GDP. Over the past four years, transfer payments have grown at a compound annualized rate of 9.1%, or about 3.5-times faster than the economy. Since passage of the Obama Administration stimulus, transfer payments have accounted for more than 18% of household income. As the composition of government spending has shifted away from capital investments and towards transfers, the “government sector” of the economy has fallen even as government spending has reached record highs.

The biggest contributor to transfer payments is entitlement spending. Social Security, Medicare and Medicaid spending are considered transfer payments that do not contribute directly to GDP. While spending on these programs has increased in recent years due to aging, they have actually fallen as a share of total transfer payments. Figure 1 provides the change in transfer payments as a share of personal income and entitlements as a share of total transfers. As transfer payments have risen from 14 cents of every dollar of income in 2008 to 18 cents in 2010, entitlements fell as a share of transfer payments from 77% to 70%. The rise in transfers was explained instead by an increase in unemployment insurance and new refundable tax credits billed as “tax cuts” but counted as government outlays. More than one-third of the “tax cuts” contained in the President’s 2009 stimulus bill were, in fact, transfer payments where the government credited the bank accounts of households who had no income tax liability.

Entitlements, Transfer Payments and Personal Income

It is important to recognize that even though transfer payments don’t directly contribute to GDP, they generally add to the economy indirectly. In the case of Medicare and Medicaid, the transfers go towards medical consumption expenditures, which are counted in GDP. Likewise, when the recipient of unemployment insurance, or the President’s Making Work Pay refundable tax credit, buys a good or service with that transfer, it also gets counted in the GDP. For these reasons, it is absurd to blame the anemic recovery on insufficient government spending, since the rise in transfers has likely pushed aggregate household consumption spending higher than it would otherwise be.

While it would be easy to completely dismiss Norris’ critique, it does have some resonance for three reasons:

(1) The decline in the “government sector” of the economy highlights how government’s role as a provider of public goods like infrastructure has fallen over time. As e21 explained previously, the growth in state and local government has come with no corresponding increase in public goods like infrastructure to show for it. Today, the federal government’s primary role is to collect money and redistribute it. The result of the redistributive state is either less public goods like infrastructure, or a much higher tax burden.

(2) Not all spending cuts are created equal. Spending cuts for sequestration aimed at defense, for example, will reduce GDP on a dollar-for-dollar basis. Spending cuts on transfer payments and subsidies only reduce GDP to the extent that the dollar would have been spent or not otherwise earned. A dollar devoted to unemployment insurance that lengthens the duration of unemployment, for example, may actually reduce GDP. Thus, it’s possible that cuts in such programs could actually boost GDP growth. Indeed, recent research from Chicago economist Casey Mulligan finds that the disincentive to work created by some safety net programs is so great that removing them would improve labor market efficiency and economic growth. This is an important lesson for Congress as policymakers consider bills to address the looming spending cuts under sequestration.

(3) President Obama’s stimulus was very poorly constructed. In 2009, Republicans criticized the stimulus as a “spending bill.” The President responded that increased government spending was the “whole point” of a stimulus. But based on the analysis of Norris and other commentators, the spending increase was obviously too oriented towards transfers instead of real purchases of goods and services. An effective stimulus based on government spending would have looked like the plan advocated by Martin Feldstein, which would have increased government purchases of military equipment and hardware. While left-leaning economists often lament the size of the President’s stimulus (i.e. wishing that it was even bigger), the composition or relative share of the type of spending was likely a much bigger problem.

In one sense, the “government sector” of the economy analysis is a canard designed to obscure the record spending levels of the past four years. Yet, when taken at face value, the critique provides some interesting insights into the changing role of government in society and the economic mismanagement of the past several years.

e21 Partnership

Stay on top of the issues that matter to you most





























Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed