Recovering from the Recession: Fiscal Drag Can Explain the Slow Recovery

I.  Introduction

Economic recovery from the 2008 financial collapse and economic downturn has been slow, with unemployment still high in 2012.  Republican political officials, whether the presidential candidates or the Republican leaders in Congress, have charged that this has been due to an unprecedented explosion in government spending during the Obama administration, and that the way to a fast recovery would be to slash drastically that government spending.  Senator Mitch McConnell, the Republican Leader in the Senate, for example, has repeatedly harped on what he has called the “failed stimulus package” of Obama, and has successfully blocked any effort to extend any fiscal stimulus to address the continued weak state of the recovery.

They point to the recovery during the Reagan years from the downturn that began in July 1981.  This was the sharpest downturn the US had suffered until then since the Great Depression, and the Republicans point to the subsequent recovery as an example of how (they assert) cutting back on government spending will lead to a strong recovery.  Unemployment reached a peak of 10.8% in the Reagan downturn in late 1982, surpassing even the peak of 10.0% reached in the current downturn.  But the unemployment rate then fell to 7.2% by election day in November 1984, and Reagan was re-elected in a landslide.

Recovery from the 2008 downturn has indeed been slow.  But it is simply false to say that fiscal spending has exploded during the Obama years, while it was contained during the Reagan years.  In fact, it was the exact opposite.  Understanding begins with getting the facts right.  And once one does, one sees that fiscal drag can fully explain the slow recovery from the 2008 downturn, while the recovery during the Reagan years was helped by the largest expansion of fiscal spending in any downturn of at least the last four decades.

II.  The Path of Real GDP

Start with the path of real GDP at each of the downturns since the 1970s (with the data here and below from the Bureau of Economic Analysis of the US Department of Commerce):

The graph shows the path of real GDP in the three years (12 quarters) before each business cycle peak since the 1970s (as dated by the NBER, the recognized entity that does this for the US), to four years (16 quarters) after the peak.  Recessions start from the business cycle peaks, with negative growth then for varying periods (of between 6 months and 18 months in the recessions tracked here) before the economy starts to grow again.  Six recessions have been called by the NBER in the US over the last four decades.  The NBER stresses that it concludes whether or not the economy went into recession based on a wide range of indicators, and not simply real GDP decline, but generally, negative GDP growth of two quarters or more is commonly associated with a recession.

As the graph shows, the downturn that followed the business cycle peak of December 2007 was the largest of any of those tracked here, with most of the decline occurring in 2008, before Obama took office, and in the first quarter of 2009, when Obama was inaugurated.  The economy stabilized quickly under Obama, and then recovered but only slowly.  By 16 quarters after the quarter of the business cycle peak (i.e. by the fourth quarter of 2011, the most recent period for which we have data), real GDP was barely above where it had been at the peak.  This was the worst performance for GDP of any downturn of those tracked here, and indeed the worst since the Great Depression (other than the fall in output after 1945, which is a special case).

The recession that began in July 1981 (shown in green in the figure) at first followed a strongly negative path, but after five quarters began to recover.  The economy then recovered faster than in any other US downturn of the last four decades, and by 16 quarters out, real GDP was over 14% above where it had been at the pre-recession peak. Republican officials argue that government cut-backs under Reagan account for this strong recovery.

III.  The Path of Government Spending

But what was in fact the path of government spending?  There are a couple of different measures of government expenditures one can use, but they tell similar stories.  First, one can look at direct government expenditures, for either consumption or investment.  This is the concept of government spending which enters directly as one component of demand in the GDP accounts:

In the downturn following the December 2007 peak, government expenditures rose in the first six quarters following the business cycle peak in the middle of the range seen in the other downturns.  But then it flattened out, and then it fell.  By the 16th quarter following the peak, real government spending was barely (less than 1%) above where it had been at the business cycle peak, four years before, and was the lowest for any of the six downturns of the last four decades (although close to that following the 1990 downturn, which was then into the Clinton years).  In contrast, government spending in the 1980s, during the Reagan years, continued to grow rapidly, so that by 16 quarters out it was almost 19% above where it had been four years earlier.  This growth in government spending during the Reagan period was by far the most rapid growth in such spending seen in any of the downturns.  Government spending growth following the March 2001 business cycle peak (i.e. during the Bush II years) was second highest after Reagan, but only a bit over 10% higher rather than 19% higher.

If one believes in the Republican assertions that fiscal spending will restrain growth and that fiscal cuts in periods of high unemployment will spur growth, then one would have seen slow growth during the Reagan period and rapid growth during the Obama period.  In fact, one saw the opposite.  Fiscal austerity is bad for growth, while fiscal expansion when the economy is in recession will spur growth.

It should be noted that government spending considered here is total government spending, at the state and local level as well as federal.  State and local spending is on the order of 60% of the total in recent years.  The cut-backs in recent years in total government spending has been driven by cut-backs in spending at the state and local levels, many of whom have followed conservative fiscal policy either by statute (balanced budget requirements when revenues are falling) or by choice.  In the four years following the December 2007 cyclical peak, state and local spending indeed fell by 6%.  This was offset to a degree by growth of 13% in spending at the federal level (of which 8.8% occurred in the last year of the Bush administration, while 4.1% total occurred over the three years of the Obama administration).  A similar breakdown during the Reagan years following the July 1981 downturn shows that state and local government spending rose by a total of 13.6% in the four years, while federal spending under Reagan rose by a total of 25.4% in those four years, for overall growth in government spending of 19%.  This large increase in federal spending does not support the common assertion that Reagan followed a policy of small government.  Federal spending grew almost twice as fast under Reagan as it did in the Bush II / Obama period following the 2008 downturn.

We therefore find that government spending on consumption and investment grew sharply during the Reagan years following the July 1981 downturn, more than in any other downturn in the US over the last four decades, and that the economy then recovered well.  In contrast, overall government spending (including state and local) was flat in the most recent downturn (rising at first in the last year of the Bush administration, but then flattening out and then falling), and the recovery has been weak.

It could be argued, however, that the government accounts that appear as a component of GDP (as direct consumption or investment of goods or services) do not capture the full influence of government spending on the economy, as it leaves out transfers.  Transfers include amounts distributed under Social Security, Medicare, unemployment insurance, and similar programs, which are indeed a significant component of government expenditures.  The purchases of good or services (or the amounts saved) are then undertaken by households, where it will appear in the GDP accounts.  But including transfers in the total for government spending will not change the story:

One again sees government spending increasing over the initial periods in each of the six downturns tracked, including that following the December 2007 peak.  This continues into the first complete quarter during the Obama administration, after which it flattens out and then falls.  By quarter 16, government spending including transfers in the Obama period is tied for the smallest growth seen in any of the six downturns (all at about 9% higher), along with that following the 1990 and 2001 downturns.  Once again, sharply higher growth is seen following the 1981 downturn, with total government spending including transfers was 21% higher after four years in this Reagan period.

By either measure of government spending, therefore, with or without transfers, government spending rose rapidly during the Reagan period while it flattened out and then fell during the Obama term.  This has produced a strong fiscal drag which has harmed the recovery.  In contrast, strong growth in government spending during the Reagan period was associated with strong GDP growth.

IV.  The Path of Residential Investment

There will, of course, be many other things operating on the economy at any given time, so it would be too simplistic to assign all blame or credit for GDP recovery on government spending alone.  Government spending is powerful, and does have a major impact on growth, but there are other things going on at the same time.  In the most recent downturn, the most significant other factor affecting aggregate demand for output has been residential investment.

Residential investment had risen sharply during the Bush II presidency, as has been discussed previously in this blog (see here).  The housing bubble then burst after reaching a peak in prices in early 2006 (see here), housing construction fell to levels not seen in decades, and such construction remains stagnant.

The collapse in residential investment seen in the most recent downturn in unprecedented.  While residential investment has typically fallen at the start of the downturn (and indeed was often already falling before the start of the overall economic downturn), the falls had never been so sharp and extended as now.  Four years since the business cycle peak in December 2007 (and six years since the peak in residential investment), residential investment remains weak, with no sign yet of recovery.  There are two reasons:  1)  The housing bubble reaching a peak in 2005/06 was unprecedented in size in the past half century, with residential investment reaching 6.3% of GDP in the fourth quarter of 2005, and then collapsing to just 2.2% of GDP in 2010/11; and 2) The downturn that started after the overall GDP peak in December 2007 was fundamentally caused by this bursting of the housing bubble, which led to mortgage delinquencies, financial stress in the financial institutions who held such mortgages, and then to collapse of the financial system.

V.  The Impact on GDP Recovery of the Fiscal Drag and of Housing

Finally, it is of interest to see what the impacts on GDP recovery might have been, had there not been the fiscal drag seen in recent years, or if residential investment had followed a different path.  There are many counterfactuals one could envision, but of particular interest would be the path of government spending seen during the Reagan years, as this is often held up by Republicans as the example to follow.  The following table shows what the impact would have been, along with a counterfactual scenario which keeps residential investment flat at the level it had been in 2007Q4.

Counterfactual Scenarios    All figures as a share of GDP.  Note that the shortfall from Full Employment GDP was 5.8% of 2011Q4 GDP (CBO estimate) Direct Impact Multiplier of 1.5
Government Consumption & Investment increases                                   as in 1981Q3-1985Q3 3.4% 5.1%
Total Government Spending (incl. Transfers) increases                             as in 1981Q3-1985Q3 4.0% 6.0%
Residential Investment flat at 2007Q4 level 1.4% 2.1%

The results indicate that the fiscal drag in recent years can by itself account for the shortfall of current GDP from its full employment level.  The estimate of full employment GDP comes from the January 31, 2012, economic baseline forecast of the Congressional Budget Office (see here), and is 5.8% above where current GDP was as of the end of 2011.  That is, if GDP were 5.8% higher, we would be at full employment (or “potential”) GDP, as estimated by the CBO.

In the first line of the table, government consumption and investment expenditure is increased by how much it was in 1981Q3 to 1985Q3 during the Reagan years, rather than the much slower growth (indeed essentially no growth) of 2007Q4 to 2011Q4.  The direct impact would have been to increase GDP by 3.4%.  Assuming a very modest multiplier of 1.5 (many would argue that the multiplier when the economy is in recession as now, with Federal Reserve Board interest rates close to zero, would be 2 or more), the impact would be 5.1% of GDP.  This is close to what would be needed to make up for the 5.8% gap.

In the scenario where total government spending (including transfers) is increased as it was during the Reagan years rather than the slow growth that has been chosen in recent years, GDP would have been 4.0% higher by the direct impact alone, and by 6.0% with a multiplier of 1.5.  This would have fully closed the gap.

One can also look at scenarios for residential investment.  Given how high the housing bubble had gone in 2005/6, it would be unrealistic to believe that housing would bounce back up quickly, and certainly not to the bubble levels.  But in a scenario where housing had simply recovered in real terms to the level seen in the fourth quarter of 2007 (when it was 4.0% of GDP), the direct impact on GDP would have 1.4%, or 2.1% assuming a multiplier of 1.5.  While still quite significant, these impacts are less than that resulting from the slow growth of government spending in recent years.

VI.  Conclusion

In summary:

1)  The recovery of GDP in the recent economic downturn has been slow, with unemployment still high.  Not only is a vast amount of potential output being lost, but long periods of unemployment is particularly cruel to the lives of those who must suffer this.

2)  Prominent Republican leaders have repeatedly asserted that the slow recovery has been due to an explosion of government spending under Obama.  This is simply not true.  Growth in government spending since the onset of the recession in December 2007 has been slower than in any other downturn of the last four decades in the US, and has been far less than the growth seen following the 1981 downturn during the Reagan years.

3)  Growth in government spending following the 1981 downturn during the Reagan period was in fact the highest by a substantial margin of any of the six downturns.

4)  If government spending had been allowed to grow in the recent downturn as it had during these Reagan years, the economy by the end of 2011 would likely have been at or close to full employment.

5)  Residential investment has also collapsed following the housing bubble that reached its peak in 2005/6, and has contributed substantially to the current downturn.  Its impact, while significant, is however quantitatively less than the impact of slow government spending, since residential investment is normally (and even at its peak) well less than government spending as a share of GDP.