Issues Affecting Economic Growth – Gored Oxen Edition
I write about issues I believe affect economic growth. For example, over the years, I have written a lot about taxes. And here’s a simple graph showing why:
What we see is that tax rates at any given time seem to be related to the growth rate of real GDP per capita over the next decade. What is more, the correlation is positive. That is to say, growth tends to be faster when tax rates are higher, and not lower. This of course contradicts popular belief, particularly among Republicans. However, since economic growth is important for the quality of life of all Americans, getting this right matters. Unfortunately, over the past few decades, government policy has gradually moved us in a direction that inhibits growth.
Of course, it could be the relationship between tax rates and future growth shown in the graph is a spurious correlation. But that is unlikely, since it is very easy to explain why (up to a certain point) higher tax rates would lead to faster economic growth. Additionally, even people who get the direction of the correlation wrong are certain a correlation is there. But… if it ever does turn out that the relationship is spurious, we won’t find that out by keeping our head in the sand.
Another topic I have been writing on a lot lately is immigration. Here’s what a graph looking at the foreign born population in certain years and the growth rate of real GDP per capita over the next ten years:
The correlation between the share of the population that is foreign born and the growth rate is negative, which indicates that as the foreign born share rises, growth falls. The correlation between these two variables, at least in the post WW2 era, is stronger than the correlation between tax rates and growth. This of course contradicts popular belief, particularly among Democrats. However, since economic growth is important for the quality of life of all Americans, getting this right matters. Unfortunately, over the past few decades, government policy has gradually moved us in a direction that inhibits growth.
Of course, it could be the relationship between the percentage of the population that is foreign born and future growth shown in the graph is a spurious correlation. But that is unlikely, since it is very easy to explain why (up to a certain point) having less immigration would lead to faster economic growth. Additionally, even people who get the direction of the correlation wrong are certain a correlation is there. But… if it ever does turn out that the relationship is spurious, we won’t find that out by keeping our head in the sand.
If it seems to you that I have written almost exactly the same thing about taxation and immigration, it isn’t your imagination. I did a copy and paste of a big chunk of the first half of the post to the second half and changed a few words. The fact is, the analysis is very similar. The only difference is whose ox is getting gored. A grown up is willing to look the data in the eyes and follow it where it goes.
Although I am on the conservative side, I must say I am more likely to believe the tax relationship than the foreign-born relationship. Per capita growth has been essentially a “bell curve” since WWI, not just here, but globally — particularly in the developed countries. Our immigration policies have changed in a similar pattern, so there are two different causes (economic conditions and immigration policy) causing similar patterns (growth rate and percentage of foreign-born).
I implore everyone here to read “Capital in the Twenty-First Century”. (I got it from the library, and liked it so much I bought the Kindle version and re-read it.) Picketty argues there for higher taxes and redistribution — even a wealth tax. And while I disagree with him on many points, his work there of gathering and presenting the relevant data is extremely valuable. Even arguing for higher taxes, he does not show a relationship between top marginal income tax rates and per-capita growth rates. (He does, however, show correlations to the growth of income inequality and wealth inequality.)
To further support your assertion on tax rates and growth, I recommend three things. First, put the dependent variable on the x-axis. Second, do it for every year. (After all, the Kennedy, Reagan, and Bush tax cuts came in the first few years of those decades.) Third, do it for three-, five-, and ten-year timelines.
Another interesting view might be to look at not the top marginal rate, but the marginal rate paid by those in the 90th percentile, and at the top capital gains rate.
Warren,
I have written each of the posts you have requested here at Angry Bear at some point between 2006 and the present. There is a reason I am confident in the relationship between taxes and growth, and a half a dozen other things. Data on Immigration is more sparse and I haven’t covered the topic in the same depth, but it isn’t difficult to see similarity in the patterns.
Benefits of immigration:
Would you please provide your data sources?
Of course, correlation is causation. None of these differences could possibly have other causes.
merde
Mike Kimel,
First the 1940+t high growth in GDP is easily explained. During the run-up to World War II industrial production grew rapidly to supply the growing US military. And that growth continued throughout the war. Anyone could get a job and rationing caused the personal savings rate to exceed 25%. After the war ended those personal savings were used to resurrect a consumer based economy. So the 1940+t data is just adding to the noise in graph.
The period of 1930+t is another exception since it is coincident with the Great Depression. That period was before tax rates were raised to support the build up of the US military and GDP growth was being supported by FDR’s massive government spending programs.
That leaves all of the ‘normal’ GDP growth at tax rates on 70% or more. (Defining ‘normal’ as a little less than 2% and up to 3% growth)
That leaves you with only 2000+t and 2010+t to support your theory. But taxes had been reduced in the early 1980s so the 1980+t should have caused low GDP growth too. I do not see any external explanation for the 1980+t to run high growth in the face of lower taxes.
So I believe that your theory about tax rates vs GDP growth is flawed.
Moving on to your theory about the % of foreign born persons vs GDP growth, the 1930+1 and the 1940+t data points are also just noise here.
That leaves all the ‘normal’ GDP growth at rates of foreign born persons at 8% or less. (Defining ‘normal’ as a little less than 2% and up to 3% growth)
So you have only 2000+t and 2010+t to support your theory about a higher percentage of foreign born persons adversely affecting GDP growth. And those periods are explainable by the run up in free trade treaties, which enabled US manufacturers to move production to foreign countries, which adversely affected the employment and wages of American workers. The most dramatic affects of those actions only showed themselves after the housing and debt bubbles popped. And the aftermath of those actions are with us to this day.
So I believe that your theory about the percentage of foreign born persons vs GDP growth is also flawed.
I would end by noting that I have serious reservations about including foreign persons who are here legally, into a discussion about the problems in our economy. I do not believe that legal immigrants to the US have caused our current problems and they should not be blamed for them. I do believe that legally and morally we could limit legal immigration in the future. But once admitted, legal immigrants have to be allowed to integrate into our society. Legal immigrants have enough problems without us making their situation worse.
I have no such reservations about blaming illegal immigrants for problems caused to American workers with a high school education or less. The vast majority of illegal immigrants are decent hardworking people but we have to give a preference to decent hardworking Americans.
OK — I pulled real (2009 dollars) per-capita GDP numbers from https://www.measuringworth.com/usgdp
I did trailing 3-year, 5-year, and 10-year growth rates, starting in 1945.
This is a slight upward trend, but the residuals are huge. For the 3-year data, 4.5%; for the 5-year data, 4.3%, and for the 10-year data, 1.8%. The range of the 3-year data spanned less than 7%, less than 5% for the 5-year and ten-year data. To have residuals half the range of the data indicates practicality no correlation at all.
Warren,
When I get home I will send links. I am usually better about including sources in the post and I am genuinely sorry I left that out. But real GDP per cap is from nipa table 7.1, tax rates from the IRS and foreign born come from the Census.
If you include all years be aware that the relationship is quadratic. Try this approach:
http://angrybearblog.strategydemo.com/2016/10/37672.html
JimH,
You must have noticed the graphs show correlation both for the entire sample and for 1959+ to avoid both being accused of cherry picking and to show just a post war period.
Re tax rates and growth rates, see here:
http://angrybearblog.strategydemo.com/2007/11/tax-rates-and-growth-rates-some-graphs.html
I discuss the Reagan tax cuts by pointing them out on a graph.
As to immigration… with taxes you could make some sort of unit root argument if you try hard enough. I have dealt with that issue in a few posts. But the foreign born share series is different. It is v shaped in the sample reaching a nadir around 1970. Off of memory (I don’t have my spreadsheet here) we are now back where we were around 1930.
I used FRED’s GDP per Capita, with their Annual Average (instead of quarterly data):
https://fred.stlouisfed.org/series/A939RX0Q048SBEA
Using the rolling 3, 5, & 10 year growth rates I get -0.04%/year, -0.1%/year, & -0.2%/year respectively.
The residuals, as Warren pointed out already, are huge. Rsquared for the linear regression are 0.058, 0.136, & 0.231 for 3, 5, & 10 Year growth rate correlation respectively. I didn’t do the std dev (1s) for the residuals, but eyeballing it for the 10 year growth rate 2s = ~ +/-10%, and for the 3 year growth rate 2s is ~ +/- 5%… but Rsquare tells the story better than using the variance.
The only point here is that the rolling GDP per Capita Growth Rate is negative and simply follows the cyclical recession cycles, so effectively if you’re going to use GDP or GDP per Capita to correlate with anything over time you’re actually correlating with the business cycles…. so trying to make a case for causation with anything has to deal entirely with the business cycle causes. Good luck.
Economics assumes that individuals want to maximize their after-tax income.
The standard analysis is that if you raise taxes on profits it will lead to investors reducing their investments because they have less available to invest.
But it could be just as realistic that an individual wishing to maximize their after-tax income after a tax hike would increase their investments so as to regain and perhaps increasse their after-tax income.
No trained economist looks at the world this way, but why not?
It seems just as realistic, and maybe more so than the standard line of analysis. This seems to be in line with the findings that people tend to value loses more than gains. So if they lose something because of a tax increase,. they would be highly motivated to get that lost income back.
Spencer:
ditto on the last sentence.
My father used to point out that back in the days of a 90% upper tax bracket, he never heard of anyone refusing a raise.
I just posted a several-paragraph response to Kimel’s comments to me last night in the thread to his last post, at http://angrybearblog.strategydemo.com/2017/01/i-still-think-thomas-is-a-hack.html. I began that response with:
“Wow, Mike. I don’t think I know of anyone who so habitually conflates such an array of things that are so obviously distinct from one another and irrelevant to one another.”
I invite you to read it.
The two hallmarks of his interminable series of racial and anti-immigrant posts are the habitual ridiculous conflations and the correlation-between-these-two-statistics-as-opposed-to-the-correlation-between-these-two-statistics-equals-causation-of-the-difference-between-the-two thing. As EMichael says above, none of these differences could possibly have other causes.
“A grown up is willing to look the data in the eyes and follow it where it goes.”
Presumably, you mean ALL the relevant data, Mike. Including data from this country in the 1920s and late 1940s and ‘50s. Lots of foreign-born types lurking around here in those eras.
You should grow up.
But you’ll certainly hear of people declining overtime.
It’s not just about maximizing after-tax income. If if were, even people with full-time jobs would be looking for work on the side a lot more than they do.
Ah. Guess your father didn’t know anyone offered a raise who cared about freedom, Jack.
“But you’ll certainly hear of people declining overtime.”
I cannot believe how cheap technology has become.
To think someone so incredibly stupid has access to the internet i mind boggling…
Spencer,
My argument is different. Based on my business experience, I would say that when tax rates are higher, a business owner is less likely to take money out of the business and more likely to reinvest all profits into the business. In other words, the higher the tax rates, the more likely people try to avoid those tax rates and the easiest way to avoid paying them is to reinvest which leads to more growth.
Warren,
Tax rates: https://www.irs.gov/uac/soi-tax-stats-historical-table-23
Foreign born pop: https://www.census.gov/newsroom/pdf/cspan_fb_slides.pdf
They knew about freedom but money was the way they kept score. As for overtime, Warren, all I observe is that despite our “confiscatory” tax rates, people jockey for position to get the opportunities for overtime. Guess they haven’t heard about your theories so should be forgiven for their irrational behavior.
As for reasons for owners investing in businesses, I’d guess it had more to do with demand for their product/service than anything else.
JackD,
Of course the implicit assumption in my statement was that the business has growth potential. Most businesses do.
Maybe some people who aren’t under severe financial stress prefer time off to overtime?
Actually, when consumers don’t have money, most businesses don’t have growth potential.
Thanks, Mike.
I’m glad we changed the rules about where our immigrants come from, but normalizing massive immigration beyond the levels set by law was a big mistake. It basically gave employers a big lever to use against workers. It hurt worst at the low end. There are way too many businesses in the US structured around the assumption that labor is cheap and can’t complain, and a lot of it is because there has been a growing pool of immigrants ready to take any job on any terms. We need to get rid of businesses like that, even if we have an ugly transition stage. If nothing else, maybe it will force some businesses into the 20th century and get them automating.
@Kaleberg,
Basically, yes. As I’ve posted elsewhere, if we ever wanted to be serious about undocumented workers we would (1) impose prison time on employers who hire undocumented workers and (2) fine anyone who benefits from undocumented labor (restaurants, hotels, golf courses, etc). As you say, there would be “an ugly transition stage,” but that’s what treating the issue seriously looks like. Until then, business and consumers are locked into a conspiracy of silence.
Joel,
Ditto.