Wednesday, May 18, 2011

Deficits Don’t Hurt the Economy – But Taxes Might

I was recently involved in a Facebook thread about the relationship between deficits, taxes, and the economy.

One of my friends wrote, “if you look at our past four economic recoveries (Reagan, Clinton, Bush, and Obama) and compare them with the deficit, a very clear pattern emerges: the higher the deficit, the weaker the recovery. It seems to me that if you compare them with level of taxation, another pattern emerges: the higher the taxes, the stronger the recovery. I'm not claiming that higher taxes result in economic recovery, but it's clearly not the death-knell of the economy.”

In true nerd fashion, I reacted by making charts. You can take the Black Belt out of Six Sigma but you can’t take Six Sigma out of the Black Belt.

GDP Growth vs. Previous Year Federal Surplus (Deficit) 1948-2010

The first chart shows growth in Gross Domestic Product (GDP) as a function of the previous years’ Federal budget surplus or deficit. The technical details are below. As you can see, there is no clear correlation between the size of the deficit and the strength of a recovery. Although the Clinton recovery, during a time of budget surplus was strong, and the Bush recovery, during a time of above average deficits was weak, the Reagan recovery, during a time of even bigger deficits, was the strongest of all.

When you look at the entire data set, rather than just the three peak recovery years, no correlation between deficits and GDP emerges. Within the margin of error, the trend line is flat.

GDP Growth vs. Previous Year Taxes 1948-2010

Taxes tell a different story. Again, there is no clear pattern in the peak recovery years. For the data set as a whole, there is a statistically significant trend: higher taxes equals lower growth. Every one percent increase in Federal revenues as a percentage of GDP corresponds to a 0.55% percent decrease in GDP growth. Taxes aren’t a death knell to the economy, but they do damage.

Although there is a correlation between taxes and growth, it is a weak one. Combined with the absence of correlation with deficits, the data suggests that Milton Friedman was right: other things drive the economy far more than fiscal policy. Monetary policy is one driver. As Friedman said in a 1996 interview, “One of the things I have tried to do over the years is to find cases where fiscal policy is going in one direction and monetary policy is going in the opposite. In every case the actual course of events follows monetary policy. I have never found a case in which fiscal policy dominated monetary policy and I suggest to you as a test to find a counter-example.”

Technology is another driver. The invention of the World Wide Web was clearly a factor in the strength of the Clinton recovery.

It should also be noted that the data falls into a narrow range. For almost the entire post-war period, the Federal account balance varied between a deficit of five percent of GDP and a surplus of two percent. But one data point sticks out dramatically - 2010. With deficits in excess of 10%, President Obama is taking us into new territory. The economic consequences of that remain to be seen.

Technical Details: The Federal tax and deficit data used in the charts comes from the “Fiscal Year 2012 Budget, Table 1.1—Summary of Receipts, Outlays, and Surpluses or Deficits (-): 1789–2016”, and is available from the Office of Management and Budget. The GDP data used comes from the “Current-Dollar and ‘Real’ Gross Domestic Product”, April 28, 2011, and is available from the Bureau of Economic Analysis.

The slopes of the trend lines and the uncertainty in the slopes were calculated using the Excel LINEST function which uses the least-squares method for fitting a line to the data. My claim above that there is no correlation between deficits and GDP growth is based on the slope of the trend line being less than the deviation. In the case of taxes, the slope is between two and three deviations.

My friend pointed out that, “There's going to be a lag between policy changes and their effect.” The charts are based on a one-year lag, e.g. the 2010 GDP is plotted against 2009 deficit and receipts. I found the one-year lag gave the strongest correlation (highest R-value).

I identified the various economic recoveries with the years in which economic growth peaked. I make no claim that this is the best way, or even a good way to measure the strength of a recovery.


  1. My friend Jim says I should change this to "Deficits don't matter as long as someone will lend to you."

  2. Found an error in one of the deficit chart and corrected it, along with the corresponding text.

  3. Interesting analysis, Michael. I still stand by my observation for the past three presidential terms. In other words, if you order them by deficit, taxes, and economic growth, you get (lowest to highest):
    Deficit: Clinton/Bush/Obama
    Taxes: Obama/Bush/Clinton
    Economic Growth: Obama/Bush/Clinton

    It looks like the Reagan period had a high deficit and a better economic growth than I had remembered. (During most of his presidency I was making $12K/year in an industry that fared much better under Carter than Reagan, but apparently I was not typical.) Still, I was surprised to see the Reagan recovery as so much higher than Clinton's. Of course, it depends on how you measure it. For example, if you look at the length of period of economic growth, Bush/Clinton is higher than Reagan. (See

    I have learned a lesson here, and that is that it may simply not make sense to play armchair economist and take a complex system such as the US economy and boil it down to a small number of variables. It's just too easy to play with the numbers until you get the answer you want. Should your data be weighted so that more recent information is worth more than the past, where economic rules may have played by different rules? Should we be looking at deficit or debt? Should we be dividing numbers by GDP to normalize them?

    I personally believe that the US economy is too large a beast to respond in just a year's time, but, hey, that's just me playing armchair economist again. If I'm right, then we should just now be starting to see the effects of the decisions of the Obama administration.

  4. Peter: Thanks for posting.

    Unfortunately we all need to be armchair economists. The need to evaluate the various candidates' proposals when we go to the voting booth, along with disagreements among the experts (even in cases where there shouldn't be any), turns us all into students of the dismal science. One more argument for separation of economics and state.

    The time needed for the economy to respond to a given policy depends on the type of policy. After World War II, the German economy responded to the removal of price controls literally overnight. On the other hand, it took several years of high interest rates in the late 70s/early 80s to bring inflation under control in the U.S. In the case of fiscal policy, keep in mind that the one-year lag that I used is the delay between a change in tax receipts and a change in GDP growth. In general, that would be after a year's delay between a change in policy and a change in tax receipts - so really we're looking at a 2-year lag. That's consistent with what we saw in the Reagan administration when there was a tax cut in 1981 and it took until 1983 before the economy started to turn around.

    Regarding a couple of your other points: because of the long term growth in the economy, it is necessary to normalize by GDP for any meaningful year-to-year comparison. Also the economy does seem to have been playing by a different set of rules from 1983 to 2007 than either before or since. Recessions were both less frequent and less severe during that period. Economists refer to it as "The Great Moderation". I find it interesting that as a % of GDP, government was smaller during The Great Moderation than either before or since.