Republicans - All Wrong, All the Time, Pt 7.1

To further illustrate the idiocy of Ed Lazear's claims in his WSJ article, I've done some more playing with the numbers.  I also got a new data set for GDP growth, so values through 2009 are included, based on chained 2005 dollars. I also found estimates of  Taxes/GDP for 2009 of 15.1 and 15.4 %.  I used 15.4%, since the higher number favors Lazear's case, though the difference is decimal dust.




The first graph shows how GDP growth, year over year, and Taxes as a percentage of GDP vary over time.  Taxes/GDP is in blue, GDP YoY is in pink.    Also included are period averages and best fit lines (least squares method) for each data set.  Taxes/GDP average is 17.97%; GDP YoY average is 3.34%.

For Taxes/GDP, the period average and best fit lines are close.  Slope of the best fit line is slightly positive at 0.017% per year.  For GDP YoY, the slope of the best fit line is negative at -0.04% per year.  This corresponds to a decline of just over 2% in 50 years.  This also explains the negative correlation Lazear emphasizes so strongly.  While the trend of Taxes/GDP has been almost level, the trend of GDP has been down.  This imposes a negative bias on a comparison of the two data sets.

The really striking thing, though, is the comparative motions of the data gyrations over time. From 1950 through the mid 80's, motion was typically contrary: peaks in one corresponded to troughs in the other.  Since the mid 80's, motion has been generally quite similar.  Over the most recent 25 years of data, this is the exact opposite of what Lazear would have us believe.

So, what does this mean?  Absolutely nothing.  Big GDP Whoop!  Each value is a resultant, not a determining factor that can rationally be used as the independent variable in some psuedo analysis.

But, just for kicks, let's play Lazear's game.  Here's the cloud plot again, with the new data set.  The worst number since 1950 now belongs to Barack Hoover Obama.  Note this also corresponds to a very low tax receipt value.  GDP is down because business activity is down.  Taxes are down because profits are down and people are out of work.  As I said, both numbers are resultants.  This is not rocket science. 

The green line is again the best fit straight line.  Correlation is now -.28, deeply into the weak range.  All by itself, the 2009 result utterly destroyed the tattered remains of Lazear's ridiculous hypothesis.

But why don't we take it another step.  The Blue and red lines are the averages of GDP growth for Democratic and Republican administrations.  Democrats achieve 4.19% growth, Republicans achieve 2.77%.  The difference is almost Lazear's 1.5%.

So let's say this:

Republican administrations are clearly bad for the U.S. economy.  For example, were we to elect Republican presidents over the next 25 years, GDP per capita in 2035 would be about 30% less than if we were to elect Democrats.  A 30% per capita GDP differential is about as large as the difference between the U.S. and Greece today.

This might seem as silly as the Lazear quote I've manipulated.  But there's a difference.  Lazear's conclusion was based on the assumption that a meaningless data artifact is determinative.  Mine is based on the assumptions that policy matters, and that Democratic and Republican policies are different.   Since 1968 we have had a Republican president 68% of the time.  Lets take another look at GDP growth over the period.   I've calculated a 13 year exponential moving average over the period.  Segments in blue represent Democratic administrations, segments in red represent Republican administrations. 

Republican administrations are clearly bad for the U.S. economy.
.