Government Spending and Economic Expansions
by Mike Kimel
Cross posted on the Presimetrics blog.
Government Spending and Economic Expansions
It is conventional wisdom that raising taxes, particularly during and just after a recession, will harm the economy. Last week I checked whether that was true. (The post appeared in the Presimetrics blog and the Angry Bear blog.) The post looked at every recession since 1929, and it showed that recessions that were accompanied by marginal tax rate cuts were followed by shorter, slower expansions than recessions that weren’t accompanied by marginal tax rate cuts. (Expansion, btw, is the term for the period between recessions.)
This week I will look at the effect of cutting back on government spending during and just after a recession. I’m going to do that with three graphs. The first shows the length (in months) of every expansion since 1929. The second looks at the annualized growth in real GDP per capita for each expansion period, and the third looks at the total growth rate in real GDP per capita over the length of the expansion period. In each graph, recoveries are divided into three groups based on what happened to the federal government’s spending as a share of GDP from the start of the recession to the period one year after the end of the recession.
Before I get started, let me describe the data I’m going to use… Data on the starting and ending dates for recessions comes from the NBER, the folks who call the start and end. Real GDP per capita comes from the Bureau of Economic Analysis’ National Income and Product Accounts (“NIPA”) Table 7.1, updated on April of 2010. Real GDP per capita is available annually from 1929 to 1946, and quarterly thereafter. Data on federal government spending comes from NIPA Table 3.2, and GDP figures come from NIPA Table 1.1.5.
As I did last week, I am going to assume that the real GDP per capita in any month is equal to the real GDP per capita for the quarter (or if prior to 1947, the year) in which it fits. In other words, the real GDP per capita (in 2005 dollars) for the first quarter of 2008 is $43,997, and I am assuming that the real GDP per capita in any of the three months in that quarter (i.e., January, February, or March of 2008) is equal to $43,997. Government spending and GDP are treated the same way. That assumption shouldn’t cause any major changes in the results and it will keep me from having to go off on tangents about how the data was smoothed.
With that, here we go. The first graph shows the length of each expansion, in months.
There aren’t a lot of recessions during which spending was cut, but on average, they tended to produce the shortest expansions.
The next figure shows the annualized growth rate during each expansion.
Once again, on average, the recessions during which federal government spending shrunk as a percentage of GDP tended to producer slower economic growth. Two out of the three recessions for which the government cut spending were among the three that produced the slowest economic growth. The third one actually produced rapid growth, but as the first graph showed, that expansion didn’t last all that long either. Which leads us to the third graph, which shows the total increase in real GDP per capita during each expansion.
To summarize – while there were weren’t all that many recessions during which federal government spending as a share of GDP fell, those recessions tended to produce shorter, slower expansions than other recessions. And btw, we get similar results if we use total government spending (i.e., federal, state & local) as opposed to just federal government spending.
Now… consider last week’s post, which showed that recessions during which marginal tax rates were followed by underperforming expansions. The two findings seem to suggest that when it comes to getting the economy moving again during and just after a recession, government spending seems to be more important than private sector spending. One reason this might true – during recessions most private sector players companies hunker down and cut spending, and they usually don’t start investing and hiring people until they’re reasonably sure there’s going to be demand for their products and services. Meanwhile, individuals cut back too, fearful they might lose their jobs.
With everyone waiting until the other guy moves first, there isn’t much of a foundation set down for future growth. But if the government steps in and acts when nobody else is willing to do so, it could create that more stable environment the private sector needs in order to get off the ground.
I think the impressive thing is the fraction of thought about the economy based entirely on the 75-80 expansion. It is clearly untypical, yet it is all Republicans ever talk about.
Real GDP includes government spending, and that clearly influences the results.
If you look at private real GDP — GDP excluding government — how does that change the results?
Another good post.
Pardon if my response seems a little off the topic of your post but you’ve MADE me think of these things
What I have increasingly found to be true and odd about many of the people who talk about these matters on the conservative side of the spectrum is that they speak as though govt spending and private spending are different things. More specifically they talk as if a dollar spent by the govt behaves differently than a dollar spent through the private sector, that its more inflationary somehow. Dollars are dollars and GDP measures the flow of them…..period. There is no circumstance where cutting federal spending will improve the metric of GDP measurement by simply “allowing” for more private sector spending.
Theres no mechanism for the private sector to “know” that the govt is not spending and thusly the private sector simply fills the gap with all their excess savings going to new consumption left by the govt void. The govt DOES have a mechanism to “know” what it needs to spend………. unemployment measurements. The private sector buys what it wants at all times and what it wants is usually tied to credit conditions and savings desires. When the private sector wants less, spending drops savings increases and unemployment rises. This should function as a signal to govt to fill the demand gap and keep employment from dropping too precipitously, negatively affecting credit conditions and possibly leading to bank runs.
Some might say the deficit level is what the private sector looks at but I think the private sectors response to deficit levels becomes counter intuitive and counterproductive. The deficit levels rise initially because of losses in private spending and increases in the automatic stabilizers like unemployment payments and food stamps. When the private sector “sees” these increased deficits they are more prone to act countercyclically by saving (thinking that the govt is spending too much) which worsens the downturn and increases unemployment payments. The private sector does not have a good thermostat to look at.
That’s because Jimmy Carter caused the 1970s.
spencer,
Grumble. There’s always someone out there creating more work…
OK. Reran it with just private GDP. (I left out the per capita… hopefully that isn’t a problem.)
If I didn’t screw up the data, the big changes I’m seeing on annualized growth.
1. ’38 – ’45 looks the worst. (Not surprising… from 41 to 45 the private sector shrunk to nothing.)
2. 45 to 48 is by far the best performance. Essentially that’s the flip side of ’38 to ’45, with the war over and the government winding down.
Thanks for what I’ll take as a complement.
And I agree with the rest of what you wrote, but there’s one piece you’re missing. The private sector also cannot act countercyclically else they go bankrupt. Of course, the government can also go bankrupt if it spends like Reagan or Bush in good times and then tries to spend again in bad times.
It looks like the two most significant unambiguous effects following changes in the marginal tax rate was the great depression following the Hoover tax hike and the economic expansion in the 1960s following the Kennedy tax cuts. Also you had recovery following the Reagan tax cut which caused a big jump up in after tax personal income. Dubya used tax cuts to fight recession and today Obama is doing the same thing.
So far economic history is showing that nobody wants to be Hoover all over again. With Bush it was the tax high among other such as the first Gulf war that put us in recession. Bush new that raising the rates were going to slow the economy so it hardly makes any sense him to have lowered marginal taxes right after he raised them accepting the eventual slowdown and risk to his reelection.
It looks like the two most significant unambiguous effects following changes in the marginal tax rate was the great depression following the Hoover tax hike and the economic expansion in the 1960s following the Kennedy tax cuts. Also you had recovery following the Reagan tax cut which caused a big jump up in after tax personal income. Dubya used tax cuts to fight recession and today Obama is doing the same thing.
So far economic history is showing that nobody wants to be Hoover all over again. With Bush #41 it was the tax high among other things such as the first Gulf war that put us in recession. Bush knew that raising the rates were going to slow the economy so it hardly makes any sense him to have lowered marginal taxes right after he raised them accepting the eventual slowdown and risk to his reelection.
Mike,
Carter’s policy and the FED policy caused inflation. Economic stagnation and inflation showed that the Keynesian models did not work very well — which set to economic research agenda in the 1980s to explain why.
Too bad there isn’t a way to measure and compare the degree of human suffering vs gov’t spending during recessions. One would hope there is evidence that human suffering declines when there is a higher level of gov’t spending during a recession.
Mike, out of curiosity, how does 33 -38 look/rate?
Mike said: “Of course, the government can also go bankrupt if it spends like Reagan or Bush in good times and then tries to spend again in bad times.” May I conclude you think Obama’s policies are incorrect?
RG said: “One would hope there is evidence that human suffering declines when there is a higher level of gov’t spending during a recession.” Could be true as long as the Govt spending (policy) isn’t prolonging the recession.
I think you would find many conservatives hold this view as a likely possibility.
Using wikipedia as a quick source the average length of the 13 recessions is 12.9 months and the average peak UE is 11%. The average of the 3 recessions highligted here is 9.7 months and the average peak UE rate is 6.4%.
The recessions highlighted were both shorter and less severe than the average to a significant degree.
“but there’s one piece you’re missing. The private sector also cannot act countercyclically else they go bankrupt.”
Well, I didnt mention that specifically but I do understand that. The private sector,in the described situation of rising deficits due to initially falling spending and increases in the automatic stabilizers, can only fill the gap by going into more debt, if the govt side does not run deficits large enough to fund the saving desires of the private sector. The other way to fill the gap is to increase exports. The increased debt of the private sector does have limits as we’ve seen.
One place I disagree though…………………”the government can also go bankrupt”…….
Not ours. We can ALWAYS make a payment in $US. Greece can go bankrupt, California can go bankrupt but not the US of A. Many other things can result from our spending but not running out of money. Impossible
“Could be true as long as the Govt spending (policy) isn’t prolonging the recession.”
Spending can NEVER prolong the recession. Recession is described as a decrease in GDP (spending). ANY spending will increase the GDP and not be recessionary. Austerity packages on the other hand will ALWAYS worsen a recession.
By recessions that were highlighted, I guess you mean the ones where average government spending increases per year exceded 1.5%. Your findings regarding the length and severity of the recessionary periods of the same business cycles with the strongest and longest recoveries seem to confirm Mike Kimel’s thesis if I understand you correctly. Here is an interesting article I found to confirm these findings. It argues that,
“…if the government borrows to put people to work creating long-term investments that increase the productivity of the U.S. economy, like infrastructure and education, then it is in a much better situation. The income generated by the more productive economy, as well as by the newly employed workers, can help to provide the tax revenue to service the debt…cutting taxes for the rich…do not lead to the kind of productive economic growth that generates strong tax revenue.”
The case for Keynes is quite strong, indeed.
http://www.dollarsandsense.org/archives/2010/0510wolfson.html
By recessions that were highlighted, I guess you mean the ones where average government spending increases per year exceded 1.5%. Your findings regarding the length and severity of the recessionary periods of the same business cycles with the strongest and longest recoveries seem to confirm Mike Kimel’s thesis if I understand you correctly. Here is an interesting article I found to confirm these findings. It argues that,
“…if the government borrows to put people to work creating long-term investments that increase the productivity of the U.S. economy, like infrastructure and education, then it is in a much better situation. The income generated by the more productive economy, as well as by the newly employed workers, can help to provide the tax revenue to service the debt…cutting taxes for the rich…do not lead to the kind of productive economic growth that generates strong tax revenue.”
The case for Keynes is quite strong, indeed.
http://www.dollarsandsense.org/archives/2010/0510wolfson.html
Greg said: “Spending can NEVER prolong the recession.” Of course it can when the spending policy is to delay implementation or spend on long term items with back-end versus front-end loading.
There are bad, good and best spending policies. The full continuum is possible.
“when the spending policy is to delay implementation or spend on long term items with back-end versus front-end loading”
I’ll just say that a promise to spend 10 billion next year is NOT spending.
GDP measures current spending, spending of any type, spending to dig and fill holes and spending to buy a car. Spending to fix broken bones or spending to upgrade your power grid. The GDP metric is BLIND. It cannot see where the money comes from and doesnt care. It only cares about money moving through the system. SO govt spending can never prolong a recession. Other govt actions can prolong a recession but not spending. Again, NOT SPENDING always prolongs a recession! You cannot save your way out of a recession. Even “bad” spending policies add to GDP. I’m not advocating waste I’m simply pointing out what GDP (how we measure recessions) measures. Personally I think using GDP as a thermometer of our economic health is only marginally better than following the Dow.
“By recessions that were highlighted, I guess you mean the ones where average government spending increases per year exceded 1.5%”
No, I actually meant the 3 recessions with decreasing government spending, sorry for the confusion.
You may be correct about the average length and severity of the recessions where there was less government spending during the entire duration of the business cycle in each of the three cases you examined. But Mike’s point, which I believe is crucial, is that low and decreasing government spending during these cycles made for very weak recoveries. Consider the recession of March 2001 to November 2001. This was by all accounts a relatively short and shallow recession. But by most accounts it was followed by the weakest post-WWII expansion phase. Josh Bivens and John Irons of the Economic Policy Institute conclude:
“Of the 10 expansions since 1949, as measured from the end of the recession (trough) to the end of the expansion (peak), the expansion from 2001 through last year ranks last in average growth of GDP, investment, employment growth, and employee compensation.”
Read the whole study here;
http://www.epi.org/publications/entry/bp214/
But Mike’s point, which I believe is crucial, is that low and decreasing government spending during these cycles made for very weak recoveries.
I understand that this is (likely) the point of the post, but there is not even close (not even remotely close) to enough evidence presented here to support that point. It is generally widely accepted that shallower recessions lead to less energetic recoveries (Krugman has pointed to this himself several times in the recent past on his blog) due to the relative differences in unused productivity that is generated by the recession. There are probably dozens of other issues that would crop up (governments may be more likely to react to more sever recessions) but to ignore one of the best known correlations about the severity of recessoins and the strength of the rebounds is really shoddy.
I see what you’re saying. It could just be simple logic that a recovery following a shallow recession is likely to be weaker because of the narrow relative differences in capacity utilization in both phases of the business cycle. That must be taken into account. But I also think that some of the data Mike presented was interesting and relevant. It does seem that government spending does help. Of course, when talking about the 2001 recession during the Bush years, it is important to note that this recovery conincided with incredible deficit spending which ultimately doubled the national debt yet it was a very weak recovery by all accounts.
Perhaps the “jobless recovery” as it is known, had to do with an increasing amount of investment going overseas and the boom in outsourcing. Most of Bush’s spending was on overseas wars which didn’t stimulate much production at home like other wars did. Military spending was still only 4% of the US GDP during Bush as opposed to 14% during the Korean War and 10% during the Vietnam War.
Also, the financial sector, which creates relatively few jobs, grew much quicker than the real sector. The $2.2 trillion in total tax cuts from Bush mostly went into financial markets and didn’t stimulate much output or job growth. Financialization and globalization were the culprits in the weak recovery of 2001 to 2007 despite massive deficit spending.