Saturday, June 25, 2011

I don't think that number means what you think it means

We all have our foibles - mine is that I take pride in thinking carefully about things and trying to base my opinions on real data rather than knee-jerk emotion.  So when a conservative friend pointed out that the Bush tax cuts resulted in an "unprecedented" increase in tax revenues of $785 billion between 2004 and 2007, I decided that I needed to pause and re-examine my thinking.  Here's what I found.

From the Washington Times, it was reported that
"From 2004 to 2007, federal tax revenues increased by $785 billion, the largest four-year increase in American history."[1]
I decided to see if these numbers were correct.  Indeed, if you go to the IRS website and download tax revenue spreadsheets[2], you can find the tax revenue data.  A graph of revenues between 2003 and 2007 looks like...
If you take the 2007 receipts ($2396 billion) and subtract the 2003 receipts ($1650 billion) the results is $746 billion, which is close to the $785 billion claim.  So now let's see if this is indeed "unprecedented."  If we get the IRS data going back to 1995 - which is readily available online at the same site, we can graph it as...

The increase from 1997-2000 with the Clinton tax increases was $524 billion - a massive 42% increase, so it looks like bravo for Bush!

But do these numbers really mean what conservatives claim?  Do these numbers really show that tax cuts increase tax revenue?  Let's revisit the basic argument that was made to support the Bush tax cuts - when taxes are decreased, the GDP will grow, and a smaller tax rate of a larger number will provide more taxes. That is, the GDP pie grows so much larger that the government can take a thinner slice and actually have more.

Before we proceed, let's ask a question: what constitutes meaningful growth in the GDP?  If inflation is 10% and the GDP grows by 5% has the economy grown?  If the population grows by 10% and the GDP grows by 5% is our society increasing its productivity?

Let's apply two very simple rules - the first I call the Weimar Republic Rule, the second is the Pie Rule. The Weimar Republic Rule is that government doesn't get economic growth credit for devaluing the currency.  You can't compare money numbers from two different years without adjusting for inflation.  Otherwise, you could say that the Weimar Republic collected 100 times more taxes in 1923 than in 1922 - which may be true, but says nothing about real growth of tax revenues.

The Pie Rule is best illustrated by example - let's say your company provides free pie at lunch and your manager tells you: "Great news, our pies will now be 25% bigger." That sounds like a win for the employees - until you find out that  the number of  employees has been increased by 50%.  So the Pie Rule is that government doesn't get economic growth credit for simply increasing the population.

By applying these two rules, we can decide whether or not the raw IRS data for 2004-2007 provides evidence that the tax cuts increased revenue.

First, let's apply the Weimar Republic Rule . Adjusting for inflation using data from [3], the revenue figure in 1995 dollars[4] looks like
During 2004-2007, inflation adjusted tax receipts increased by $394 billion while the Clinton increase from 1997-2000 was $349 billion.  Applying the Weimar Republic Rule the Bush tax cuts still seem to outperform the Clinton tax increase, but by just 12% rather than the 42% we got with the raw data.

Of course, our population increased between 1997 and 2007, so to apply the Pie Rule we can look at the tax revenue per person.  The US population by year can be obtained from the Census Bureau[5], which gives us...

For 2004-2007, the tax revenue per person increased by $933, but in 1997-2000 the revenue per person increased by $1133.  That is, after the Bush tax cuts we saw an increase in tax revenues that was 17% less per person than the increase in revenues after the Clinton tax increase.  The GDP pie may have gotten larger, but the government didn't get its promised bigger piece.

Of course, the big difference is that by 2000, Clinton's tax increases produced a budget surplus.  By 2007, Bush and the GOP turned the surplus into a massive deficit.

At the present tax rates, tax cuts do not produce more revenue.  If the Laffer Curve is really something more than a figment of the supply-side imagination, then we are demonstrably on the uphill side where tax cuts reduce revenue and tax increases provide more revenue.





[4] For example, the inflation rate of 2.81% in 1995 means that the 1996 raw net taxes is multiplied by 0.9719 to get 1995 dollars.  With an inflation rate of 2.93% in 1996, the 1997 raw net taxes are multiplied by 0.9719 x 0.9707 = 0.9434 to get the 1997 tax in 1995 dollars - that is, you have to decrement each year for all the previous years to get back to 1995 dollars.


No comments: