Tax Rates and Growth
by Benjamin Studebaker
Last week, I took a look at optimal tax rates, the top rates of income tax which economic research suggests would maximize revenue if implemented as policy. The research suggested that for every 1% higher the top rate of income tax rises, the rich report 0.25% less income. This suggested an optimal top rate of between 73 and 80%. Toward the end of that post, I suggested that it might be the case that even as the rich report lower earnings, the economy as a whole might operate more efficiently at those high rates, if the government is more effective at investment than the private sector. Today I’d like to look at some empirical data to see if there’s any historical basis for that claim.
So what I’m looking at here is the relationship between the top rate of income tax and the average growth rates that accompany them. I’ll look at US data, which is quite detailed. The earliest data I have on average annual US growth rates begins in 1948, so let’s start there. The top rate of income tax for the period 1948 to 1964 was about 90%, so this period can sensibly be called the “90% period”:
The 90% period shows an average annual growth rate of around 4%. In 1964, Lyndon Johnson enacted a major tax cut, with the goal of increasing investment and consumption. This tax cut reduced the top rate from 90% to around 70%. This “70% period” lasted until around 1982:
In the 70% period, the average rate declined to just over 3%, a small reduction–the large tax cut passed by Johnson appears not to have boosted long-term growth. Of course, this period contains the infamous 70’s, in which global oil prices knocked the US economy about. We need more data. In 1982, Ronald Reagan cut the tax rate again, this time to 50%. The 50% period was brief, and provides only a few years of data:
This tax cut is the famous “Reagan Recovery” we hear so much about. Of course, it was also accompanied by a relaxing of interest rates, with the after effects of the very bizarre 70’s still in the mix. The rate for the period as a whole is about equivalent to the 70% period, at between 3 and 4%. Then, in 1986, Reagan lowered the tax rate again, this time to around 30%. The 30% period prevailed until 1994:
With the 70’s far behind us, the noise of the monetary experimentation that went on in the background of the last two periods is finally removed, with a stark result–dropping the tax rate further in the mid-80’s produced average annual growth rates below 3% for the first time in the post-war period in the United States. In 1994, the Clinton tax hike kicked the top rate up to 40% for a brief period:
This saw the average annual growth rate kick back up to just over 3%, but as we now know, much of the growth during this period was ephemeral–the stock and tech bubbles, which burst at the end of the period. At this point, the Bush tax cuts kick in, lowering the top rate to 35%, inaugurating another period of very low rates:
During this period, the United States produced the worst average growth figures it has produced since World War II, and much of the growth during the early period was revealed to be the product of an unsustainable housing bubble. As you can see looking at the other periods, no period has come close to average below 2%.
What does all of this reveal? The popular narrative in the United States, that a bloated, inefficient, state-dominated system in the 70’s was replaced by a more market-driven system that more efficiently allocated resources is a false narrative.
Here’s what appears to have happened instead. In the 70’s, oil price spikes created by OPEC and the Iranian Revolution created very high inflation, which the government struggled to break the back of with high interest rates. In an effort to counterbalance the contractionary influence of high interest rates, the Reagan administration drastically reduced top tax rates, from 70% when Reagan took office to around 30% when he left. After inflation came down, interest rates came down, and the result was very loose monetary and fiscal policy. Investors had a great deal of money to play with. At the same time, the right sought to reduce and eliminate regulations that constrained how investors could choose to use that money. The expectation was a far more efficient economy than the one that prevailed in the 90% and 70% periods, but that isn’t what we got.
Instead, private investors seem to have been dominated by herd mentality–they went from sector to sector, vastly over-investing and creating large bubbles. This even happened internationally; in the 80’s and 90’s entire national economies were sunk by over-investment by predominately American investors (think East Asia, Latin America, and so on). The arena for speculation has moved throughout the recent period, from developing countries to stocks to tech companies to the housing market, but the same fundamental conditions have consistently been at the heart of these shifts–private investors have too much influence over the distribution of funds within the economy and they consistently distribute these funds disproportionately and in economically unhealthful ways.
What really happened is that we replaced a system in which the people, through their government, decided what they wanted or needed to a system in which a group of individuals, relatively small, controls the majority of the funds available for investment. These individuals, instead of considering what is needed by the community, what the people want or would benefit from, allocate the funds available to them on the basis of what they hear from others is profitable. They make decisions based on hearsay and mob mentality. In effect, the community has given over its power of investment to a group of gamblers. These gamblers have spectacular support among the population, which genuinely believes them to be supermen, more capable of allocating resources efficiently for the benefit of the group than the group’s own decision-making institutions are.
When the right says to us that the choice is between having the freedom to spend your money as you like and having the faceless bureaucracy take your money and waste it, the right is fundamentally misrepresenting the system. The system, as conceived by the right, puts most of the money in the hands of a small group of private investors whose only concern is short-term personal profit, regardless of whether or not those short-term gains produce long-term gains for the community as a whole. The average citizen does not control a large enough portion of these funds to make any substantive difference in their allocation. He is relegated to spectator status. While he is also a spectator when the government makes the choice, the government is at least bound by moral obligations to him, and he can participate in a wider societal debate over what the government ought to do which has influence. With private investors allocating society’s output under the protection of property rights, the average citizen has no means of influence at all, of even making his grievance heard. He can petition the state to change policies, but the state itself is weaker precisely because it has given away so much of its investment power to the private sector, a sector which is immune from obligations to the community. In this way, the economic output of our society is controlled and allocated by people who have no obligations to that society, no allegiance to it, and consequently the benefits of that output will inevitably go not to the citizenry as a whole, but to a privileged subsection of it, which will continue to invest inefficiently and irrationally at the cost of the future economic output of the state.
A very well written, cogent argument. Bravo! If only more people actually took the time to look at the raw data…
Thank you, glad you enjoyed it! It seems many voters make their political decisions the same way investors making their investment decisions–on hearsay.
The interpretation of time series data is quite a tricky topic though. You can say there is a positive correlation between growth and tax rates because the state allocates resources more efficiently than the private sector. But you can also say it is because all the time a recession occurs the government lowers the tax rate.
I don’t think that interpretation of the data meshes as well with what we see. The 90% period has 4 recessions in it, during none of which the government dropped the top rate of tax. The transition to the 70% period comes during a boom. The 70% period experiences 3 recessions, but only during the last one does the government lower the tax rate, and it does this to counterbalance contractionary monetary policy used to fight high inflation produced by the oil price spikes, an unusual situation. The 50% period includes no further recessions, the 30% situation includes 1 recession during which the tax rate was not lowered, the 40% period includes one recession, during which the rate was lowered, but the growth that followed was that mid-2000’s growth, it was not very sustainable. That 35% period includes one huge recession in which the top rate was not lowered further.
All told, I count 10 recessions, only two of which potentially follow the narrative of “tax rate was lowered due to the recession”. The former is the 80’s recession in response to stagflation, a special circumstance, and the latter the Bush tax cuts in the early 00’s, which failed to produce sustainable long-term growth in their aftermath.
This was just an example about the difficulty of separating between false and true causation. There are a lot of other variables which can possibly influence growth and taxes. To prove or falsify the hypothesis it would acutally be necessary to build a complicated regression testing for all the other variables which could potentially distort our finding.
Indeed, to make something like this publishable, a great deal of additional work would need to be done to separate out all other potential variables.
I agree completely. There’s also the issue of how long it takes to fully price in a tax change. Labor obviously feels the early effects, capital the latter. If there is increased spending in and interest rates/prices are relatively insensitive to government spending, there would be an immediate increase in GDP. If interest rates are sensitive, there wouldn’t be. Inflation gets the last laugh in this kind of thing, but not for a year or two in some cases.
Point being, a much more complex analysis is needed to even get at this.
For sure there are other variables and other factors involved, and one could spend many careers researching those factors, but the story we hear, that high tax rates strangled growth in the pre-Reagan era, and then Reagan lowered tax rates and all was good in the world, clearly does not reflect reality. At minimum, if lowering taxes helps growth, it helps much less than is commonly believed. The majority of cases we can point to in which economic resources have been wrongly allocated, spawning bubbles, have been the result of private sector investment and a lax state, not a blundering state crowding out the private sector with corrupt crony capitalism.
There is a simple solution to this…
Regardless of whether or not other factors play into the growth rate, it seems pretty obvious that tax rates play a role. We can discuss to what degree they affect it until we’re blue in the face, but saying that you need to perform a regression test for other variables seems disingenuous.
Well if you follow this link you see another “causation” which seems pretty obvious. I am just mentioning a common problem in statistics.
Sorry, I didn’t mean any disrespect. Often times I forget that it’s hard to infer tone through text.
He’s right that if I was to do this study to a publishable standard as research, rather than in this casual, theorizing way, I would need to make certain lots of other variables were not in play. That said, the fact that no variable sticks out, that there is no obvious problem, speaks to the strength of the conclusion.
Well, I guess what I’m trying to say is that when it comes to something like this, I don’t really see what the point of proving causation is. It would seem to me that a strong positive correlation would be enough for a reasonable person to say, “perhaps we should reevaluate our tax rate.” Especially since economics is not an exact science.
And that’s exactly what I criticise. Just a positive correlation doesn’t tell you anything about correlation and the underlying mechanisms. Otherwise we should import storks to fight declining birth rates.
please replace the second correlation with causation
Well, to that end, I don’t think it’s ever possible to prove causation. You can say that there’s a high degree of certainty that one things causes another, but you can never account for every factor, every confounding or lurking variable. I’ve always found the correlation vs causation argument to be sort of a cop out. It looks great on paper in a statistics book, but the principle is extremely hard to practically use.
[…] old sparring partner, Benjamin Studebaker, has a post up about the correlation between tax rates and growth. He says the American experience has been […]
The data I looked at did not look at growth/tax rate relationships in individual years, but looked at annual growth rate averages under different tax regimes, where the rate of tax I examined was the top rate of federal income tax. These periods were all longer than 3 years in length, and many of them were much longer than that. You’ve taken individual data points out of context, using many years that were pre-war and WWII-era. You haven’t looked at trends, you haven’t considered the numbers in their historical context, and you’ve included a lot of depression-era data that isn’t particularly relevant. I’m unmoved. Also, I’m fairly certain you aren’t using the same data. You have years in the 50’s that you label closer in tax rate to years in the 00’s than other years in the 50’s. Where did you come up with that?
[…] other week, I wrote a post in which I observed a connection in the United States between the rate of federal income tax on the […]