by Mike Kimel
How Tax Rates Affect Investment and Consumption – A Look at the Data
Cross posted at the Presimetrics blog
This post looks at how changes in the top marginal tax rates affect peoples’ decisions on how much to consume and invest. Ask a libertarian or conservative economist and the answer is obvious – raising tax rates on high income individuals dissuades them from doing productive things – that is to say, it causes them to cut back on working and investing. In the interest of avoiding strawmen at all costs, many libertarians and conservatives might assume that a hike in the tax rates on such individuals can also cause them to cut back on consumption. After all, if tax rates are raised high enough, perhaps people won’t have enough left over to consume as much as they otherwise would. However, the reduction in voluntary activities that take effort (i.e., work, investment) should easily swamp the reduction in consumption, some amount of which is needed for simple survival. Put another way, as tax rates rise the ratio of investment to consumption should fall.
That right there is what’s known as a testable implication, and there’s data aplenty for that purpose. But before we move on to testing this, let me note that the first paragraph actually provides a second, far more familiar testable implication, namely that higher tax rates will generally lead to slower economic growth. While that particular narrative seems to be widely believed even by non-economists, it certainly isn’t borne out by US data from the last eight decades or so:
(Figure 1 first appeared in this post.)
The graph shows that, at least until top marginal tax rates get somewhere above 50% (a bit more precision available here), increasing those rates does not correlate with slower economic growth, but rather with faster increases in real GDP. In fact, raising top marginal tax rates doesn’t have many of the effects many people seem to expect (And incidentally, its worth noting that state level data also produces results the Chicago school and most libertarians don’t expect.)
Now, the reason I mention the data’s irresponsible failure to abide by conservative and/or libertarian philosophies when it comes to tax rates and growth is because I think the relationship between tax rates and economic growth can be at least partly explained by the relationship between tax rates and investment. As I stated here, in my opinion, higher tax rates can lead to more investment. After all, one way a person who owns a business (large or small) can reduce the taxes they pay on profits is to reinvest the profits, which in turn leads to faster economic expansion. Furthermore, the incentive to avoid taxes and reinvest increases as tax rates increase. Of course, at some point, tax rates get high enough to encourage individuals to reinvest even though the “benefits” from more reinvestment, at the margin, are negative. Where that happens, I don’t know, but based on Figure 1, my guess is that it takes a top marginal tax rate above 50%.
So… here’s what libertarians and conservatives should expect to see: as the top marginal tax rises, the ratio of investment to consumption falls.
Here’s what I expect to see: the relationship between the top marginal tax rate and the ratio of investment to consumption is somewhat curved. For top marginal tax rates between 0 and some point X (where X > 50%), an increase in the top marginal rate leads to an increase in investment/consumption. After that, as the top marginal rate rises, we should investment/consumption level off, and for even higher marginal rates, investment/consumption should fall.
So… using top marginal rates from the IRS’ Statistics of Income historical table 23, and national investment and consumption figures from the Commerce Department’s Bureau of Economic Analysis’ National Income and Product Accounts table 1.1.5, we can construct this little graph:
It’s not a perfect quadratic curve, but it sure looks a lot more like what I had in mind than what any conservative or libertarian would expect. FYI, the correlation between the top marginal tax rate the ratio of investment to consumption for top marginal tax rates below 50% is 55%. That is to say, an increase in tax rates increases the ratio of investment to consumption when tax rates are below 50%. On the other hand, the correlation is a negative 11% when tax rates are above 50%. That is to say, increasing tax rates when they are above 50% has a (not particularly strong) negative effect on people’s “invest v. consume” decision.
Put another way – conservatives and libertarians have a very, very flawed theory of the world. At the very least it does not conform at all with historical US data. At all. Which of course has serious consequences; because that theory is somewhat dominant in the political sphere, and has been since the late 60s. The end result – slower economic growth for all of us since the late 60s. That has real consequences for real people – 310 million of us. That should have repercussions for the consciences of economists who peddle this garbage, though apparently it doesn’t.
But that’s for another post. Today, I want to remain clinical. So… what does Figure 2 mean, with respect to Figure 1? It means that, yes, at least until a certain point (somewhere above 50%), raising the top marginal rate both increases the ratio of investment to consumption and the real economic growth rate. Its not outlandish to assume that increasing investment is one of the factors that can increase real economic growth. (Worth exploring more in a post sometime in the future.) Other things matter, of course, but I think investment is up there among factors that matter.
It also means that since we’re keeping tax rates at the level they’ve been for the past decade or so, we shouldn’t expect sustained rapid investment or economic growth any time soon. Don’t expect the 60s or even the 70s again any time soon – we’re in end of the 80s and 00s level taxation, and over the long haul, we’re in for that sort of growth too.
A few closing remarks:
1. 1. This post was partly inspired by an interview I had with George Kenney of Electric Politics. It was a really useful and interesting conversation for me, in part because he pressed me a bit past my comfort level. I understand he’ll have that interview up in a few weeks.
2. 2. I’ve noticed that a number of the graphs I’ve been putting up are not so much quadratic as a bit bimodal. I have been thinking about that for the past few days, and I think that will be my next post on this “kimel curve” approach to world I’ve been following lately.
3. 3. Please, please, please, please, if you object to something in this post and are planning to bring up Romer and Romer, read this first to save me from being embarrassed on your behalf.
4. 4. As always, my spreadsheets are available to anyone who wants them.
5. 5. Between the time I finished my spreadsheet and the time I wrote this post, I read this blurb from Tyler Cowen’s next book. It might be uncharitable of me, but my first thought was to wonder about the odds a libertarian professor would add two and two together, and whether he could remain libertarian if he inadvertently did.