Could Employing Tax Cuts Tax Employment?

Repub­li­cans have been inces­sant in their demands for the top income tax rate to remain at 35%, and are will­ing to allow the rate reduc­tions on the mid­dle class to expire rather than con­cede the top rate reset. Since the days of Ronald Rea­gan, the right has con­sis­tently main­tained that reduc­tions in tax rates, par­tic­u­larly on the wealth­i­est Amer­i­cans, ben­e­fits us all. Even today, we hear from the right that a return of the top income tax rate to 39.6% will fur­ther increase unem­ploy­ment. So I thought it would be inter­est­ing to look at cor­re­la­tions between the top income tax rate and employment.

I looked at the US Bureau of Labor Sta­tis­tics employ­ment data dat­ing back to 1960, cov­er­ing both pri­vate and pub­lic sec­tor jobs; the over­all results were the same whether I included or excluded the pub­lic sec­tor. I then cal­cu­lated the per­cent­age growth in jobs year-​​over-​​year. One issue that cre­ates noise in the per­cent­age growth is the impact of pop­u­la­tion growth. So, to com­pen­sate for this, I cal­cu­lated the mean job growth over the past 50 years (1.83%), and then cal­cu­lated how much the year-​​over-​​year job growth dif­fered from that mean. The results appear below:

The last col­umn is a point diver­gence; that is, in 1961, job growth was 2.18 points below the mean.

I also plugged in the top tax rates dur­ing those same years. The top tax rate was raised four times, and low­ered 10 times, in the past 50 years, rang­ing from a peak of 91% in the early 1960s to a low of 28% dur­ing the last years of the 1980s. Of par­tic­u­lar inter­est to me was the years imme­di­ately fol­low­ing a tax rate change, since the claim of the right is that these changes have an inverse rela­tion­ship to employ­ment. First, I looked at the same year as the tax rate change, and estab­lished a cor­re­la­tion between the change in the top rate and the diver­gence of employ­ment growth from the mean. A cor­re­la­tion of 1 would mean that a 1% decrease in the top income tax rate would cor­re­spond to employ­ment growth one point higher than the mean.

The results don’t point to much of any­thing. But, then again, employ­ment tends to be a lag­ging indi­ca­tor, so it makes sense to shift the employ­ment rates by a year or two. I tried a one-​​year shift and didn’t see any bet­ter results, but then I tried two years:

Now this looks inter­est­ing. From 1964 to 1987, decreases in the top tax rate cor­re­sponded to better-​​than-​​average employ­ment growth two years later, while increases in the top tax rate cor­re­sponded to worse-​​than-​​average employ­ment growth two years later. Every sin­gle time. But the impact was very small in 1987’s tax rate change.

What’s even more intrigu­ing is what hap­pened after 1987. From 1988 to 2003 (the last year of a change in the top tax rate), decreases in the top tax rate cor­re­sponded to worse-​​than-​​average employ­ment growth two years later, while increases cor­re­sponded to better-​​than-​​average employ­ment growth. Again, every sin­gle time.

Why might this be? One ele­ment that jumps out is the points at which the cor­re­la­tion change from pos­i­tive to neg­a­tive. Rates above 50% cor­re­sponded to pos­i­tive cor­re­la­tions, while rates below 50% cor­re­sponded to neg­a­tive cor­re­la­tions. I don’t have any good expla­na­tion for this, but it is cer­tainly wor­thy of contemplation.

I don’t claim based on these data to have proven that return­ing the top tax rate to 39.6% will nec­es­sar­ily lead to more jobs. How­ever, I do claim that there is rea­son to believe that allow­ing the Bush tax cuts to sun­set won’t be dam­ag­ing to employ­ment, and may in fact improve it.