Scott Sumner  

Do low taxes explain inequality?

Soonish Success!... ICE's Flawed Cost-Benefit Anal...

Matt Yglesias suggests that the answer is yes:

Three or four decades later, scholars are able to look at the fruits of those policies and draw some conclusions. The same main technologies that exist in the United States and United Kingdom are also in use in Germany and Sweden. Those countries are also exposed to the forces of global trade and immigration. But inequality has grown much more sharply in the US and UK than it has in Germany and Sweden. And the main reason seems to be taxes.

Looking across the world in a series of papers, a line of academic research originally pioneered by Piketty and Emanuel Saez joined over time by a growing list of other economists finds that falling marginal tax rates on the rich are strongly correlated with higher levels of income concentration but don't lead to faster overall economic growth.

Yglesias also suggests a possible reason for this correlation:

In the old days of 70 or even 90 percent marginal tax rates, it wouldn't make much sense for executives to expend enormous amounts of time and energy trying to maximize the amount of money they can personally extract from a company in the form of salary.
There may be some truth to this claim, but we need to be very careful in not overselling this result. Consider the top income tax rates in the US, Germany and the UK:

US: Top federal income tax rate = 43.4%, including state and local income taxes the top rate is probably averages 45% to 50%, even more in states like California and New York, which have many high earners.

Germany: Top rate is 47.5%

Britain: Top rate is 45%

There is very little difference between the top rates of various countries, and hence it seems unlikely that taxes have much to do with the US and UK having more inequality than Germany. I'd add that Germany didn't even have a capital gains tax until 2009, and Sweden has no inheritance tax.

PS. During his campaign, Trump promised to cut the top income tax rate from 43.4% to 25%. Now he's hinting that the top rate will remain little changed. Trump seems to be endorsing the Obama "tax the rich" policies that the GOP referred to as "socialist" just a few years ago. Congress will probably go along. And yet the media spin is exactly the opposite; they claim that he ran as a friend of the workingman and is now implementing tax cuts for the rich. Very odd.

Always be skeptical of what you read in the media. And speaking of "fake news", the press will tell you that the top federal income tax rate is 39.6%, which is false. It's 43.4%.

HT: TravisV

PS. It's also possible that this post is "fake news". Please let me know if my tax data is incorrect. It's difficult to find accurate tax data.

Comments and Sharing

COMMENTS (25 to date)
Boris writes:

I'm actually curious how you're determining the top marginal rate. It looks like you're including the "regular" income tax (39.6%) and the 3.8% net investment income tax to get your 43.4%, right? But you're not including either Medicare (1.45%) or the additional Medicare tax (0.9%)?

(Also, you have "43.8%" in one place where I think you mean "43.4%".)

sean writes:

Even if you accept we have lower taxes on the rich.

Don't we also have the highest per capita income. By 20% over all other large diverse non oil countries?

It might be a case where the figures fit both narratives. Lower taxes increase inequality but they also boost (or are atleast neutral) to lower incomes. So it comes down to what is better for society - less inequality but overall less wealth (if you view income over a certain level as status competition more than utility then plausible) or higher absolute and wealth as a goal.

Also a bit of a free-rider issue for the rest of the world. If our model is better for growth then they can piggyback our technological achievement. Letting some people getting extremely wealthy and win status games pushes them to achieve more. It gives someone motivation to go from earning a nice 500k a year to earning a billion a year.

Andrew_FL writes:

Sure the top rates aren't that different, but what percentage of median/mean income is the income threshold for those rates in each country?

Aaron McNay writes:

The top marginal tax data used by the Piketty, Saez and Stantcheva paper referenced by Yglesias comes from this Tax Policy Center report:

The numbers in that report are pretty much the same as reported here. They should include central and sub-central government top statutory tax rates. For the US is has 46.3% in 2013. However, like Scott says, this is likely low for high tax states like California and New York.

When I take the tax rates reported in the Tax Policy Center document and match them to the World Wealth and Income Database for the countries that are available (16 out of 34) for 2013, I do get a slight negative relationship between the two, but a low R squared of 0.1187. What's more, the relationship is highly sensitive to the countries that are included. For example, dropping Turkey lowers the relationship by more than a half and reduces the R squared to 0.0291.

Overall, I agree with Scott that it seems unlikely that the 10 percentage point difference in the top 1 percent income share in the United States relative to Australia is due to a top income tax rate of 46.5 percent instead of the 46.3 percent rate we have in the United States.

Max writes:

Are we more equal than the US?
I am from Germany and I find that we are not more equal, but in general worse of than the same person in the US.

Also, besides income tax, we have social taxes (unemployment "insurance", solidarity tax, retirement tax etc.). We have to pay 18 € per month for public TV/Internet and we have for most items a 7% or 19% VAT. So total Tax burden is even higher and I dont count indirect taxes like Gas Taxes or Energy Taxes.

I believe the only limiting factor is that Germany has a weaker Start-Up Sector & more family-owned mid-size companies that retain a tech leadership in production compared to the US. This guarantees high wages for people without a higher education. Also trade schools are more widely regarded as a good option.

Brandon Berg writes:

I would assume he's including both halves of the Medicare tax, i.e. 1.45 + 1.45 + 0.9, which is also 3.8%. I believe the Pease limitation adds a bit more than one percentage point, too.

J Mann writes:

I think Scott gets to a 43.4% top rate by adding income tax, the 1.45% Medicare payroll tax twice (once for employee side, once for employer side), and the 0.9% "Additional Medicare Tax."

Scott Sumner writes:

Boris, Thanks, I fixed it. The 43.4% comes from adding the 39.6% regular income tax, with the 3.8% income tax surcharge added to fund Medicare and Obamacare.

Andrew, I believe the threshold is much lower in the UK (150,000 pounds) than in the US and Germany (over $420,000 in the US, even higher in Germany.)

Boris writes:

Ah, ok, if you include both halves of medicare that makes a lot more sense than assuming the income is investment income. ;)

The Pease phaseout bit is a good point. Assuming you're itemizing at all (and anyone in the top tax bracket is), you'd certainly hit that, because the phaseout isn't done by the time you get to the top bracket...

jamie writes:

In New York City for the highest bracket, you have to pay an additional 8.82% State income tax and an additional 3.876% City income tax (excluding SALT deductions). Mind you, if, like many do, commute from CT or NJ, get ready to pay the highest property taxes in the nation.

TravisV writes:

Great post overall! However, re: this:

"Now he's hinting that the top rate will remain little changed.........Congress will probably go along."

Hmmm, it seems highly unlikely to me that a tax bill without significant cuts for the wealthy would get enough Republican votes to pass in Congress. It also seems plausible that the GOP will narrowly pass a tax bill with large cuts for the rich. If so, I think Trump would sign it just to show that he got something significant done.

Maybe I'm not following closely enough, though. We'll see..........

Thomas Sewell writes:


It's very difficult to cut Federal income taxes on people who don't pay income taxes (leaving aside income grants, like refundable credits to people who don't otherwise pay).

As virtually everyone who pays income taxes in the United States is rich, how do you propose Congress cut income taxes for only non-rich people?

What would be your definition of rich?

Also, what do you see as the point of cutting income taxes?

Matthew Waters writes:

Mr. Sewell,

If the rich is defined as significant wealth (say >$1M income or >$10M wealth), then the tax cuts proposed in 2017 cut their taxes more than the proportion they pay in taxes.

Many of the plans floated from GOP have a few things in common:

1. Lower top tax bracket.
2. Elimination of Obamacare surcharge.
3. Lower rate on pass-through income, down to 25%.
4. Elimination of estate tax.
5. Elimination of AMT.
6. Lower corporate tax rates. (IMO, corporate tax is mostly incident on shareholders)

With #1-#3 combined, the top tax rate for Trump with, let's assume, $50M in pass-through income would go from 43% to 25%.

To try to make it revenue-neutral, things like lower 401k caps and eliminating state/local tax deductions are being floated. To actually be revenue-neutral and keep these tax cuts for the very wealthy, the middle and upper-middle class taxes would have to be increased considerably.

Maybe you could make an ideological argument that the tax system should become less progressive. But it's not like the tax cuts floated simply cut taxes on the rich because the rich pay more taxes. The proposed tax cuts in practice have been targeted to the rich.

Matthew Waters writes:

The Piketty paper linked by Yglesias seems to have a fundamental flaw. It hinges on the correlation between:

1. Pre-tax share of 1% income.
2. Drop in top marginal rates since 1960.

The correlation works out well enough. But it's merely because the US and UK had extremely high top marginal rates in 1960, while other countries didn't.

After 1960, the US and UK merely adopted similar tax rates as other countries. It doesn't make much sense why *changes* in marginal tax rates, rather than the marginal tax rates themselves, cause income inequality.

The causal theory gets tougher to accept the more you think about it. If marginal dollars are taxed at 80 cents vs. 30 cents, then sure the causal theory shows a big effect. But 40 cents versus 50 cents?

Income inequality, at least in US, has a few big causes:

1. Bad financial regulation in US (and probably UK). Not deregulation per se, but a lot of moral hazard scenarios which were abused. For example, hedge funds made a lot of fees from pension funds, which have government backing.

2. Less cultural restraints on executive pay. Krugman called this the "Monday Night Football" effect, where CEO's saw athletes get millions and they rose their salaries accordingly.

3. The income tax rates could be similar, but the US wastes more of its government money on overpriced medicine, military and criminal justice spending.

Scott Sumner writes:

Matthew, Excellent point about levels vs. changes in MTRs.

Thomas Boyle writes:

I get frustrated by the frequent comparison of the Federal U.S. tax rate to the tax rate of particular European countries.

In California the top marginal tax rate on individual earned income is well north of 50% (although it's almost impossible to figure out the actual rate). That's also true of Sweden (56% top rate). It's not true of Germany or the U.K., nor of Florida.

If we're going to compare particular European countries, let's compare them to particular U.S. states. For a Manhattan taxpayer, the federal rate is irrelevant.

Mark writes:

Matthew Waters:

"Income inequality, at least in US, has a few big causes:

1. Bad financial regulation in US (and probably UK). Not deregulation per se, but a lot of moral hazard scenarios which were abused. For example, hedge funds made a lot of fees from pension funds, which have government backing.

2. Less cultural restraints on executive pay. Krugman called this the "Monday Night Football" effect, where CEO's saw athletes get millions and they rose their salaries accordingly."

I don't see how either of these two are remotely tenable explanations for rising income inequality. Hedge fund managers and corporate CEOs are fare too small a population to account for this. And Krugman's claim is preposterous. Corporate CEO pay is actually well-explained by performance (

IMO, the biggest explanation for rising income inequality is likely just growing disparities in cost of living between cities and regions. E.g., the income of the average San Franciscan or New Yorker growing faster than the average Houstonian or Detroiter, due to corresponding increases in cost of living (mainly housing).

William White writes:

The observation about Monday Night Football may not apply. There are a few soccer players who make more than the top paid NFL player.

Matthew Waters writes:


It depends on whether "income inequality" means 20%/80% or the top 1% or top 0.1%. Managerial or financial professionals account for 70% of the income growth within the top 0.1%.

On CEO or managerial pay generally, for the sake of brevity I simplified my thoughts. There is a simple reason culture plays a big role: the divergence between American CEO's and other countries.

Either American CEOs truly earn their pay and culture depresses pay in other countries, or culture overcompensates CEOs in America. It's not easy to ascertain what a CEO is truly worth. It's far easier for either a line employee or Lebron James. The system where the board/CEO pays compensation consultants to justify compensation will of course tilt towards more compensation.

The financial piece is also a big deal, which involves a lot of subtle legal and principal-agent issues. PE funds buy a company and only hold a tiny slice of equity, leveraging the rest with junk bonds. Tax law subsidizes interest expense, but also PE funds pay themselves hefty management fees.

Floccina writes:
In the old days of 70 or even 90 percent marginal tax rates, it wouldn't make much sense for executives to expend enormous amounts of time and energy trying to maximize the amount of money they can personally extract from a company in the form of salary.

The above seems contradict the Democrat ideas that people care much more about relative income and there is a sharp fall off in the marginal utility of money.

I think you need to look at consumption rather than income. If in a full employment economy you tax Warren Buffet and give the revenue to a low income person who consumes more, who consumes less, not Warren Buffet.

Luke J writes:

Even if wealth distribution could be framed in Wins and Losses, it is still better to focus on equitable rules of the game, not .500 records.

How about 20% flat rate on everybody? The bottom rate "in the old days" was 10-20% anyway. Time for some people to start paying their fair share.

Mark writes:


The smaller the percentile, the less relevant it really is. The top .1% may be the object of a great deal of enmity, but redistributing the sum of their ‘excess income’ wouldn’t amount to much more than a drop in the bucket. Top 1% or top 10% amount to more statistically, but increasingly we are no longer talking about hedge fund managers and are instead talking about successful surgeons and small business owners.

Anyway, I don’t think ‘culture’ is the only way (or the best way) to explain it. High bracket tax rates have a significant effect on pre-tax salaries, for one thing. But it’s also possible that there’s simply more competition for high quality executives in one country than another, which likely has to do with what industries are dominant in each country. How common monopolies or oligopolies are also seems like something that would be relevant.

In the case of Germany, until not long ago, German CEOs tended to come from within the company and were often not professional managers (e.g., Siemens had Chemists or Engineers in managerial positions). Professional management came to Germany (and I suspect other European countries) much later than in the US. Also German in German companies, the employees (per regulation) get a much bigger say in corporate policy (and I wouldn’t be surprised if similar ‘stakeholder’ rules apply in other European countries). So the role of management (and who determines their salaries) is different from the US. These factors also likely play a role in difference in managerial compensation.

Lastly, inasmuch as managers do ‘overpay themselves’, would this not depress profit margins? I guess one could argue that investors, by and large, are so passive that they don’t notice or care much about their CEOs being overpayed, or that the amount of overpayment has only an insignificant effect on profits (in which case I’d be inclined to rule insignificant in general). In any case, if managers are robbing their investors, it seems like an issue for the investors, not the tax payers.

Fred_PA_2000 writes:

Re: Mark's observation on manager's potentially robbing their investors.

The Fortune Global 500 companies had $1.5 Trillion in total profits in 2015, or $3 Billion each (on average).
Forbes says the average Fortune 500 CEO had total compensation of $10.5 million. Or 0.35% of the profits. (Assuming we do no great violence by assuming the domestic 500 and the global 500 are similar.)
Assuming the CEO of a company you've invested in is only worth half what he/she is getting paid, their "robbery" costs you 0.18% of what you would otherwise have been entitled to.
I suspect most consider that insignificant / not worth trifling over -- especially given that evaluating the CEO's contribution is very difficult.

John Major writes:

I think tax is only a part of the phenomenon, since the UK and US are also the places in the OECD where pre-tax income for the 1% rose the most.

However, you forget that what matters for the 1%, 0.1%, 0.01% is the capital gains tax rate, which is relatively low with a top rate of 19.1% in the USA, versus 28% in the UK, 31% in France and 25% in Germany (2011, Ernst & Young).

One thing we need to remember is that tax rates have an effect on pre-tax income if said income is mainly derived from capital gains; if we reinvest our income, then lower taxes or evaded taxes in year Y mean more pre-tax income in year Y+1 !

Reasons I haven’t seen in the comments for this rise in income for the richest 1% :

1) Financiarization : the UK and US are heavily financiarized economies: the correlation between with the share of finance in the economy and the share of income of the richest 1 percent is strong. This is historical, but was accentuated by a series of deregulations since the 1970s. It is an important factor in increasing inequality, since pay inequality is extreme in finance, and since larger financial sectors bring with them the too-big-to-fail mechanism, whereby taxpayer’s money serves to insure large financial institutions and the state is forced to bail them out in times of crisis. Also, the presence of a financial industry brings lighter regulations on capital movements and brings in know-how in tax planning and tax evasion.
See graph (and text) :

2) Immigration (both legal and illegal) : obviously, expanding the supply of labor keeps wages low. The US and UK have a high net migration rates, and for the US a very high number of undocumented migrants who keep wages extremely low in cleaning, construction and agriculture. The number, between 10 and 15 million, means about 3-4% of the total population, when it’s under 1% for other Western countries (or 2% in the UK if you take the extreme estimate of 1.3 million, based on a 1.1 million number in 2010 according to anti-immigration group MigrationWatchUK).

3) The executive pay competition, which is probably cultural in some part.

4) Weakness and decline of unions. This one is a bit less certain. But here’s the gist: in the US, only 20% of workers were unionised in 1983, a proportion which went down to only 10% in 2016. The UK, which had a high trade union density in 1975, at 43.7%, went down to 25% in 2013. But the Germany, France and the Netherlands also went down by half of more (to rates close to that of the US), while Spain stagnated at 10-15% and the Nordics remained at 50-70%.

5) Lack of effective minimum wage laws. In many European countries, there are minimum wage laws which automatically raise minimum wages with inflation. Employees also negotiate wages at the company or sector level. This means that growth doesn't necessarily translate into higher wages in the US for low-wage sectors and jobs.

George writes:

My personal theory is that most if not all the measures of inequality correlate with social trust and social capital as Robert Putnam defines them. They have positive externalities beyond the individual and his or her family. So as social capital/trust declines inequality will increase. I’ll wager that due to the migrant crisis in Europe and for example Sweden the social capital and trust will decline due to many factors but namely diversity as a result I would expect to see increasing inequality in the next 10 to 20 years in these countries.


Return to top