Do higher marginal tax rates reduce income mobility?

by on February 15, 2016 at 12:12 am in Economics, Law | Permalink

It seems so, according to the job market paper from Mario Alloza (pdf):

The results obtained suggest that higher marginal tax rates reduce income mobility. Particularly, I find that an increase of one percentage point in the marginal rate is associated with declines of about 0.5-1.3% in the probability of changing deciles of income. A decrease of 7 percentage points in the marginal tax rate (slightly smaller than a standard deviation of non-zero changes in the rates) can account for about a tenth of the average income mobility in a year. The effect of taxes on mobility arises in specifications that consider income distributions both before and after taxes and transfers, suggesting that the impact of taxation on mobility goes beyond redistribution effects. The economic mechanism that induces this impact seems to be related to the labour market incentives created by changes in the tax schedule.

Interestingly, these effects are especially pronounced toward the bottom of the income distribution. If the implied labor supply response seems too high to you, then read some of the recent papers by Karel Mertens.  The idea that taxes matter is making a comeback in economics, though I am not sure you would get that impression from most of the economics blogosphere.

Do note the mobility results cover taxes only, and not how the money is spent.

For the pointer I thank Peter Isztin.

1 Rob February 15, 2016 at 12:27 am

What about the highest tax rate in a progressive income tax?

2 Nathan W February 15, 2016 at 1:10 am

Is the effect of raising from 1% to 2% similar to the effect of raising from 49% to 50%?

Or raising from 5% to 10% as compared to raising from 45% to 50%?

Recall that since lower income people are often not in formal professions, and I think are then more likely to be active in informal markets. Thus, whatever effect is found should be considered as an upper bound because there will be uptake in informal markets.

After all, the data does not include informal market income. For example, 1% of the bottom decline might actually belong in the second decline after accounting for informal market activities, and this will basically never show up in the data.

This is one of many reasons that the consumption approach can be better than the income approach in evaluating changes in well-being in a country over time.

3 Roy LC February 15, 2016 at 3:06 am

How much of income mobility is related to accumulated capital?

Increseases in marginal rates hit right there. I suspect the last thing that most households allocate money to is accumulating capital, not counting house payments, thus an increase in marginal rates would particularily interfere at this point.

4 rayward February 15, 2016 at 7:14 am

The period covered by the study (1967 – 1996) is marked by a sharp drop in the top marginal income tax rates, with relatively small increases and decreases in marginal income tax rates within the period, and a sharp and steady increase in inequality, and drop in income mobility, throughout. If decreases in marginal income rates correlate with increases in income mobility, why didn’t income mobility increase sharply during the period? Indeed, income mobility shows a steady decline (and inequality shows a steady increase) during the period even as (or because) marginal income tax rates declined sharply. But that’s not to suggest that tax rates aren’t relevant to income mobility (and inequality). I’d point out that the vast majority of taxpayers pay more in payroll taxes than in income taxes, and payroll taxes were subject to large and sustained increases beginning in the 1980s even as top marginal income tax rates dropped significantly – and payroll taxes aren’t assessed against most of the income of high earners. This study only considers income tax rates. A better design would include payroll taxes.

5 Mark February 15, 2016 at 1:25 pm

If you read the paper, he has a section in which he incorporates payroll taxes. See pages 22-23 and Table 9. The thesis appears to remain intact: ” The estimated coefficients are lower by about 0.25 percentage points when compared to those in Table 3, but still significant across all specifications considered at confidence levels of at least 90%. When considering pretax income, household are 0.565% (standard error of 0.20) less likely to move to a different income decile when the marginal tax rate goes up by one percentage point. Specifications considering transition across income quantiles report estimated coefficients ranging from -0.38 to -0.33 (with standard errors of 0.19 and 0.20).”

6 Nathan W February 17, 2016 at 9:38 pm

” If decreases in marginal income rates correlate with increases in income mobility, why didn’t income mobility increase sharply during the period?”

Perhaps broad- based public education, in particular massively expanded access to post-secondary education, improved the ability of formerly lower classes to access higher quality employment opportunities.

7 anon February 15, 2016 at 9:05 am

In terms of what I saw at the circus, a lot of option holders in startups turned them into shares, hoping for long term capital gains. A few executed one day sales, which were treated as regular income.

The stock rose, it fell, the tax avoiders ended up as income avoiders as well.

Even reasonable policy (treating long term income differently) can confuse people, make them miss potential (real, if they’d just sell) gains.

8 Ray Lopez February 15, 2016 at 10:45 am

Yes, my impression is that this study is simply empirical validation of the Ronald Reagan anecdote that actors avoided working more when the found they would be bumped to a higher bracket where more of their income would be taxed. This is as “old as the hills” (or as old as the US income tax code).

9 spencer February 15, 2016 at 11:32 am

The Reagan story on marginal taxes and actors not working is interesting.

At least the way I heard it Reagan says he did not work and consequently fewer movies were made.

But he did not document that fewer movies were made, leaving open the question that Reagan not working actually lead to fewer movies being made. The studios could very easily turn to another actor to make the same of more movies. So the Reagan story is interesting folklore but not interesting economics.

10 JWatts February 15, 2016 at 11:51 am

“But he did not document that fewer movies were made…”

Ok, do you have any evidence that more movies were made do to higher taxes?

Essentially you have three scenarios, 1: higher taxes slow economic growth down, 2: higher taxes have no effect on economic growth and 3: higher taxes increase economic growth. Which case is supported by the evidence?

In all cases, you have to look at effective marginal tax rates. Just looking at the marginal rate alone, doesn’t tell you much, because deductions have had dramatic impacts on the effective rate.

11 Mike Kimel February 15, 2016 at 12:30 pm

JWatts,

But think of it from the perspective of the folks bankrolling movies. At the end of the year, they are sitting on $X in profits. They can either pull the money out then, pay the taxes on their income, and engage in consumption, or they can re-invest the income (thus making more movies) with the thought that they can take their profits later when the profits are larger. Would they be more likely to pull their money out and engage in consumption or would they be more likely to reinvest it if the marginal tax rate is 90%? How about at 90%?

For what little its worth, I am a salaried employee, and my wife runs our small family business. Reinvesting every dime that my wife’s business makes is both our primary tax avoidance (or postponement, if you prefer) strategy and how we intend to fund our retirement. However, the lower tax rates fall, the less of an incentive we have to postpone our consumption and maximize the amount we invest.

A thirty second google search isn’t turning up number of movies released every year, but it seems (http://www.the-numbers.com/market/) that ticket sales peaked in 2002, which is to say, with movies greenlighted under Bill Clinton’s marginal tax rates. Yes, there were technological shocks going on at the same time, but this really doesn’t fit the Reagan story.

12 JWatts February 15, 2016 at 2:20 pm

“…but this really doesn’t fit the Reagan story.”

I don’t see that.

First, from a studio’s point of view, operating in the 1950-60’s, it might make sense to give everyone catered alcohol heavy lunch because that meal can be expensed vs paid out in wages and taxed at the high marginal rates of the time. Or for the studio to create some labyrinthine partnership scheme that effectively tax shelters the money. It seems highly likely that in a high marginal tax rate environment a good deal of productive labor will be spent on minimizing the taxes paid. That labor is always a societal net loss.

Also your conflating Reagan’s story, as a paid employee, versus the behavior of the employer. From Reagan’s point of view, there was no easy way (at the time) to avoid paying taxes on the money at the highest marginal rate. There’s little incentive to work if you are paying 90% in taxes.

And in any case, there have been several studies that support the Laffer Curve, which is effectively what we are referring to here:

“The New Palgrave Dictionary of Economics reports that estimates of revenue-maximizing tax rates have varied widely, with a mid-range of around 70%.”

https://en.wikipedia.org/wiki/Laffer_curve

13 Mike Kimel February 15, 2016 at 4:13 pm

“First, from a studio’s point of view, operating in the 1950-60’s, it might make sense to give everyone catered alcohol heavy lunch because that meal can be expensed vs paid out in wages and taxed at the high marginal rates of the time.”

I’ve noticed that in general, the better a company is doing financially, the more leniently it treats its employees. In the 1960s, just about any studio in the US was a world beater, just because it was a studio in the US. That hasn’t been true in a long time, so the margins are tighter.

“Or for the studio to create some labyrinthine partnership scheme that effectively tax shelters the money”

Do you believe that studios employ fewer and less labyrinthine tax avoidance schemes today than in the 1960s? If the top marginal tax rate dropped to 10% tomorrow, would you expect the majority of tax accountants and tax attorneys to lose their jobs? I grant you some of the tax schemes would be different and take a different focus (particularly when it comes to international transfer pricing), but I cannot see the overall complication going away unless tax rates drop very close to zero.

“lso your conflating Reagan’s story, as a paid employee, versus the behavior of the employer. From Reagan’s point of view, there was no easy way (at the time) to avoid paying taxes on the money at the highest marginal rate. There’s little incentive to work if you are paying 90% in taxes.”

I phrased my statement poorly, but I was responding to your question to spencer about whether fewer were made. Sure, Reagan may have felt less incentive to appear in a movie at a 90% tax rate, but as I noted, the owners of the studio had more incentive to reinvest their profits into making more movies at that rate than at lower rates. In other words, either fewer movies made with Reagan in them, or Reagan got paid more than he otherwise would have in order to get the movies made.

“And in any case, there have been several studies that support the Laffer Curve, which is effectively what we are referring to here:”

I respectfully disagree. We are talking about the relationship between tax rates and economic activity. The Laffer curve indicates that there is a relationship between government revenue and tax rates, and I have yet to come across a proponent of the Laffer curve that doesn’t start with the implicit assumption that output is maximized when tax rates are zero (which also happens to be one of the points where the government collects nothing).

14 anon February 16, 2016 at 7:42 am

We should return to what these actors say. Michael Caine, net worth $185M, famously complained of a 50% tax burden on new income. His 2016 income was $58 million. At 50% tax, is $28M per year in take-home pay worth getting out of bed?

Was the complaint rational?

15 JWatts February 16, 2016 at 10:20 am

“I have yet to come across a proponent of the Laffer curve that doesn’t start with the implicit assumption that output is maximized when tax rates are zero (which also happens to be one of the points where the government collects nothing). ”

I highly doubt that. How many proponents of the Laffer Curve want the entire Federal government abolished? All of the US military disbanded, the Department of Treasury, etc? There might be a few people who believe that but there not common.

16 JWatts February 16, 2016 at 10:21 am

” ..their…” (once again I wish this site had a Preview or Edit feature)

17 anon February 15, 2016 at 9:07 am

Basically, when people erroneously avoid income to avoid tax, that is an education problem, not a tax problem.

18 Cliff February 15, 2016 at 11:42 am

Nope. Leisure/work tradeoff. See Europe.

19 anon February 15, 2016 at 12:14 pm

Again, what I saw at the circus.

And from what I saw, those that booked income, paid the tax, got the leisure.

20 Cliff February 15, 2016 at 9:38 pm

Just because something didn’t work out for someone doesn’t mean they made the wrong decision ex ante. As a former professional gambler I know that lesson very well.

One anecdote does not mean anything. The only person with an education problem is you.

21 anon February 16, 2016 at 7:33 am

You are now “proving” a leisure tradeoff in Europe with your gambling.

Genius.

22 anon February 16, 2016 at 7:47 am

But to put it clearer, these are Behavioral responses to the feelings of tax, rather than a shrewd assessment of risk adjusted return.

(The textbook recommendation on IPOs and options is to execute and diversify. It was a long shot to do otherwise, based on historic market data.)

23 Mark February 15, 2016 at 1:26 pm

Actually that has nothing remotely to do with this paper.

24 anon February 15, 2016 at 1:38 pm

“The economic mechanism that induces this impact seems to be related to the labour market incentives created by changes in the tax schedule.”

My point, or question really, is if the response to those policies is rational in the traditional sense? As I say, I believe I have seen otherwise. Seems Behavioral rather than rigorously Rational.

25 asdf February 15, 2016 at 9:30 am

By far the most useful tax tool I’ve ever seen. Its inflation adjusted effective (not marginal, all-in) tax rates over time by income level.

http://qz.com/74271/income-tax-rates-since-1913/

26 Mark February 15, 2016 at 1:33 pm

Even more striking than the excerpt Tyler quoted was this paragraph:

“To sum up, higher marginal tax rates increase mobility in both samples (less clearly in the case of college graduates). But the effects on upward and downward mobility are the opposite: non-college are, on average, more likely to move down in the income distribution, while college households are likely to move up (or, at least, less likely to move down) as a result of an increase in the marginal tax rates. These results, although should be taken with caution due to the increased uncertainty resulting from smaller samples, have important policy implications. Fiscal reforms that homogeneously reduce marginal tax rates seem to contribute to income mobility by making households with non-college education more likely to occupy relatively higher positions within the income distribution (and vice versa for college-graduated households).”

Trickle that down, Thomas Piketty!

I suspect the explanation for this is that the more educated worker was much more able to obtain raises or other income to offset to some extent the tax impact. Not just in the private sector, since the period he examines was one of considerable increases in public sector pay too. Whereas a tax cut stimulated, I would imagine, spending on goods and services provided by sectors of the labor force that are populated by less educated workers.

27 John Thacker February 15, 2016 at 6:16 pm

Interestingly, these effects are especially pronounced toward the bottom of the income distribution.

Rather bad news for the ACA, then, as it massively increases marginal tax rates on the lower end due to the subsidy phaseout.

28 Ricardo February 16, 2016 at 12:19 am

That’s true of every means-tested program. If you are in favor of means testing, you are also in favor of effective increasing marginal rates. Just like the right-leaning people who say they want to means test SS and Medicare are effectively calling for a wealth tax, whether they realize it or not.

Not surprinsgly, we live in a world of trade-offs and it isn’t much of an argument against a given program to say it has some perverse incentives. One needs to quantify the effects of those incentives and compare them to the cost of not having a program or having a program that is more expensive and/or more poorly targeted.

29 Jim B February 17, 2016 at 11:43 am

Summary: Taxes reduce incentives. Effects are very small. Soooooo shocking.

30 Clay B February 19, 2016 at 11:23 am

Does this mean higher marginal rates at the worker’s current bracket, or in the potential higher bracket?

Comments on this entry are closed.

Previous post:

Next post: