The False Promise of Stimulus Spending: Lessons from the Great Recession

Report Markets and Finance

The False Promise of Stimulus Spending: Lessons from the Great Recession

May 6, 2020 15 min read Download Report
Adam Michel
Adam Michel
Former Senior Policy Analyst, Grover M. Hermann Center
Adam N. Michel focused on tax policy and the federal budget as a Senior Policy Analyst in the Grover M. Hermann Center.


The economic damage following COVID-19 has already led to bipartisan calls for more aggressive federal stimulus programs to jump-start economic recovery. A program of stimulus spending would be misguided. Estimates of the effectiveness of fiscal stimulus show that it does not boost private activity or increase total output. Following the Great Recession, stimulus policy experiments illustrate how direct transfers and government purchases fail to jump-start the recovery—and may even slow it. Individuals and businesses react to new government programs by scaling back their personal spending and shifting—rather than expanding—production, canceling out any theoretical benefits. Instead, Congress should allow the private sector to drive the recovery by removing barriers preventing businesses from expanding, hiring, and creating new products.

Key Takeaways

History shows governments cannot spend their way into economic recovery; stimulus measures are at best ineffective and at worst, delay recovery.

As economic damage is realized across the country, calls for more aggressive federal programs will increase in intensity.

Congress should consider past stimulus failures and allow the private sector to drive recovery by removing barriers to investment, work, production, and trade.

As the U.S. economy has undergone a severe suppression of activity to fight COVID-19, a novel coronavirus that originated in Wuhan, China, federal programs of direct payments, tax credits, and loans are acting as a kind of floor for the economy to rest on as we give the pandemic health response time to work.

Timely, temporary, and targeted relief to help people and businesses bridge the pandemic-related shutdowns has short-term benefits that can ease immediate hardships. However, policymakers should not think of these policies as “stimulative.” Current spending programs will have future costs in the form of poor incentives, misallocation of capital, new public debts, and likely future tax increases.

As the full extent of the economic damage is realized across the country, calls for more aggressive federal programs will increase in intensity. For example, current calls for multi-trillion-dollar infrastructure spending programs to provide jobs and a boost to the economy are already bipartisan and likely to grow louder as work restrictions are eased in the coming weeks and months.

Policymakers must not heed the siren song of stimulus spending. Instead, they should learn from the Great Recession, which demonstrated the inability of government spending programs to boost private activity or increase total output. Government spending tends to displace existing projects and employment, rather than add to them—and there is evidence that high levels of government debt could make additional spending even less effective at stimulating a recovery.

As described in a Heritage Foundation Backgrounder,REF reducing barriers to private activity which enable people to work, trade, and invest without needless impediments are all better suited to economic recovery. Additionally, as a recent Heritage Foundation Special Report noted, improving economic freedom in the United States carries far greater promise in returning the country to growth and prosperity.REF

The Argument for Fiscal Stimulus

The theory of effective stimulus depends on each dollar of new government spending resulting in more than one dollar of increased private-sector activity. In the simple Keynesian economic model, this multiplier effect can jump-start falling consumer demand, which in turn drives an expansion of supply, creating jobs and increasing private sector output.

Economists often use a single number—a “multiplier”—to communicate how much gross domestic product (GDP) will increase or decrease for each additional dollar of government purchases. A multiplier of one means that government spending creates no additional private sector economic activity; the resources are simply shifted from the private sector to the public sector. A multiplier below one means that additional government spending would shrink private activity and could slow total economic output. Multipliers larger than one predict the traditional Keynesian view that government spending increases economic activity more than the direct outlay.

Keynes’ simple 1930s model of directly propping up falling demand has since been revamped. New Keynesian economists have modernized the theory by adding households and firms that make decisions based on expectations of the future. In these more complicated models, the impact of additional government spending on employment and economic output is not so straightforward. The results are highly dependent on other variables, such as the timing of the government purchases, persistence of the policy, the type of purchases, and the interaction with monetary policy, among many others. Most of these variables are not reasonably controllable by democratic legislatures.

New Fiscal Research Shows Stimulus Ineffective

Following the economic crisis of 2008, the United States and governments around the world enacted large fiscal stimulus programs, guided by economists predicting large multipliers and promising shorter, less painful recessions. The programs would prove to be an excellent opportunity to put the theory of stimulative government spending to the test. In the years that followed, a new cohort of fiscal researchers reinvestigated historic examples of fiscal action and the new data from the financial crisis, finding the evidence does not support the Keynesian or New Keynesian theory.

In a 10-year retrospective on new research following the financial crisis, Valerie A. Ramey investigates the effectiveness of government spending programs as a response to recession. Chart 1 shows Ramey’s sampling of government spending multipliers based on aggregate data. The multipliers come from researchers using a wide range of models, techniques, data, and time periods. Summing up the results, Ramey explains, “The bulk of the estimates across the leading methods of estimation and samples lie in a surprisingly narrow range of 0.6 to 1.”REF She concludes that stimulus spending likely does “not stimulate additional private activity and may actually crowd it out.”REF


BG3493 Chart 1


Two different phenomena could further complicate the application of general multipliers to the current economic downturn: Interest rates are near zero, and the U.S. has a high debt-to-GDP ratio. Calibrated New Keynesian models often show multipliers can be larger when monetary policy is constrained by the “zero lower bound”—times when interest rates are near zero. However, the calibration can effectively build in the desired result because strong assumptions about how the real world works are necessary for the model to identify the fiscal policy effect. Thus, the economist’s preexisting worldviews can easily drive the results.

There is also compelling evidence that spending multipliers are zero or negative (perhaps as large as –3) when a government’s debt-to-GDP ratio is above 60 percent.REF In 2019, before the current crisis, U.S. federal debt held by the public was 79.2 percent of GDP.REF In the first quarter of 2020, Congress added an additional $2.5 trillion to the national debt, increasing projected debt held by the public to over 101 percent of GDP by the end of the 2020 fiscal year.REF Such high debt levels will make any additional stimulus spending more likely to shrink the economy than expand it.

The big-picture historical evidence on government stimulus shows it to be ineffective across historical episodes and implementation strategies. Examples from the U.S. Great Recession shed light on specific failings of stimulus programs.

Direct Payments: Ineffective Stimulus

During the financial crisis of 2008, Congress sent checks to most individuals on two separate occasions through the Economic Stimulus Act of 2008REF and the American Recovery and Reinvestment Act (ARRA) of 2009.REF

In a review of his research 10 years after the Great Recession, John Taylor explains that the 2008 “temporary rebate did little or nothing to stimulate consumption demand, and thereby aggregate demand, or the economy. In fact, the data show that consumption began declining in the following months.”REF The 2009 stimulus package resulted in similar trends. Failure to boost aggregate consumption demand breaks the key link that would predict increased consumption leading to a broader government-induced economic recovery.REF Evidence shows that similarly motivated indirect transfer programs, like the 2009 “cash for clunkers” program, which subsidized new vehicle purchases, also have no measurable medium-run effect on purchases or overall economic conditions.REF

Research following stimulus payments does find many recipients spend a majority of their one-time payment,REF but this micro-level analysis misses the longer-run picture described by Taylor. One reason one-time payments may have little to no impact on aggregate trends is that many individuals spend and save their income based on expectations about their future income.REF Looking over their life cycle, individuals factor in things like the possibility of future tax increases to pay for current period benefits and temporary versus permanent changes in income.

Government Purchases: Ineffective Stimulus

The ARRA was also intended to create new jobs in the hardest-hit sectors of the economy during the Great Recession through direct government purchases. The vast majority of the ARRA spending was allocated to state and local governments for infrastructure, health, and green energy projects. These federal payments were misused, poorly targeted at reviving struggling firms, and crowded out private activity.

Temporary stimulus programs are more successful at shifting resources within industries, rather than expanding the industry. Fieldwork from Garett Jones and Daniel M. Rothschild shows that “stimulus funding went to firms that were already busy, not those that suffered the most from the downturn.”REF Several surveyed firms turned down private-sector non-ARRA funded work, highlighting that government spending was directly competing with private activity. The same researchers found that among ARRA subsidized employers, only 4.4 percent of laid-off workers were rehired, and the plurality (47 percent) of the measured ARRA-created jobs were hired from the ranks of the already employed at other competing firms.REF

Most jobs, especially infrastructure construction jobs, require skills specialization and training to be effective, safe, and efficient. Knowing the government money is only temporary, training unemployed workers to expand payrolls is often not worth the costs, so instead, federal contractors hire skilled workers from private-sector contractors at inflated wages. This is an example of government-sector activity “crowding out” or displacing private-sector activity. Federal dollars can also crowd out state and local project funding by allowing lower levels of government to decrease debt issuances or fund other priorities.REF In Maryland, former Governor Martin O’Malley cut spending and raided the state’s infrastructure trust fund for other priorities following the receipt of ARRA infrastructure money so that net state funding for transit infrastructure decreased by $90 million.REF

Federal spending also funnels money that could have been invested by the productive private sector to federal bureaucrats who are not disciplined by the market (profit and loss), and thus often misallocate resources. ARRA funding financed projects such as new sidewalks to replace similar sidewalks built just five years earlier and a Nevada biomass plant intended to generate electricity, which was closed after the federal funds dried up.REF Better known malinvestments include the $535 million loan to the failed solar manufacturer Solyndra and similarly sized grants to Abound Solar, which subsequently filed for bankruptcy, and First Solar, which laid off workers and paid out large sums to its executives following large federal investments.REF Not only did government stimulus projects crowd out other existing projects and their employment, instead of adding to them, moving employment to government priorities can add additional costs when the public money dries up and the industry must again reshuffle to meet private-sector demands.

At the macro-level, government purchases financed with taxes or debt can also crowd out private spending. If the government taxes or borrows a dollar from individuals, the amount they have to spend or invest is reduced proportionally. The trade-off of fiscal stimulus through government purchases is a choice between private activity or government activity, as clearly articulated by John Cochrane in 2009: “We can build roads instead of factories, but fiscal stimulus can’t help us to build more of both.”REF

The Current Crisis

As with any government intervention, there are both benefits and costs. Through increased spending and decreased revenues, the more than $2.5 trillion the federal government has allocated thus far for the COVID-19 pandemic has a direct and immediate benefit to the recipients. Much of the aid will serve as a floor for the economy to rest on while non-essential functions remain closed and the pandemic can be contained. These temporary programs, however, are not immune to the practical and theoretical problems that undermine the effectiveness of fiscal stimulus over the medium term. Government spending ultimately has real economic costs that will show up in a number of known and unknown ways throughout the recovery.

The newly expanded unemployment benefits included in the CARES ActREF allow a majority of Americans to earn more from unemployment than a job and will thus create a more sluggish labor market, deepening the recession and delaying the recovery.REF Businesses receiving loans and other government subsidies will face restrictions and public pressure against changing employment levels or buying out investors to retool for the post-crisis economy.REF Higher levels of public spending and debt can also crowd out private investment directly and high levels of budgetary uncertainty can push investors into safer, low-return assets and away from more productive private-sector options.REF Each of these distorted incentives will accumulate to slow the recovery and depress necessary levels of innovation following the crisis.

Policymakers have determined the immediate life- and livelihood-saving support to keep the economy appropriately shuttered is worth the medium- and long-run costs. These current policies, however, are not costless and should not be confused with economic stimulus. As Congress contemplates the best response to the post-pandemic recession, the lessons of past stimulus spending will become ever more salient. Historical evidence tells policymakers that new government programs will not boost the recovery—and will more likely divert its course, prolonging the downturn and constraining growth and prosperity.

An Alternative Approach

Despite the apparent confidence with which many economists embrace stimulus spending, there is little consensus about the correct theory of booms, busts, and what can be done to counteract them. For example, Matthew Mitchell likens the government’s ability to revive an ailing economy to early 19th-century surgery: “[T]he instruments are blunt, we’re not very adept at wielding them, and there’s a good chance the intervention will cause more harm than good.”REF

Policymakers would be wise to abide by the same principle we ask of our doctors: First, do no harm. Rather than spend billions or trillions of dollars on fiscal stimulus with high costs and dubious benefits, Congress should first remove unnecessary restrictions and regulations that prevent businesses from expanding, hiring, or creating new products.

State and local governments restrict the supply of credentialed professionals by requiring unnecessary occupational licenses, the federal government has erected barriers that keep small businesses from accessing new private funds to retool or expand, and federal programs like unemployment and disability insurance create incentives for people not to work.REF Congress and state legislatures should remove the barriers we know are depressing economic growth, not add to them by diverting activity with government checks that come with greater costs than the benefits we are promised.

As Congress formulates the next economic recovery package, it should:

  • Avoid stimulus spending because the economic costs outweigh any benefits;
  • Remove barriers to economic activity, allowing individuals and businesses to more easily adjust to meet the needs of the post-crisis economy, including through new business formations and expansions; and
  • Reject new impediments to growth in the form of additional regulations, expanded redistribution, or mandates that raise costs and slow the recovery.


The trade-off of fiscal stimulus is, at its core, a choice between private activity and government activity. In fact, by simply shifting private activities to government, stimulus spending does not create additional growth—and likely depresses it. Lawmakers should resist the seemingly easy, and counterproductive, fix of spending additional money to jump-start an economic recovery. This strategy could do more harm than good by creating economic incentives for the private sector that would ultimately slow the recovery. Instead, Congress should allow the private sector to drive the recovery by removing barriers to investment, work, production, and trade.

Adam N. Michel, PhD, is Senior Analyst for Fiscal Policy in the Grover M. Hermann Center for the Federal Budget, of the Institute for Economic Freedom, at The Heritage Foundation.


Adam Michel
Adam Michel

Former Senior Policy Analyst, Grover M. Hermann Center