The anglosphere and high-income concentration

Anthony Atkinson and Andrew Leigh have a working paper on the strikingly similar patterns in high-end income concentration in five English-speaking countries.

Here is their graph of the income shares of the top 1%:

and of those whose incomes put them between the 95th and 99th percentiles of the income distributions of their respective countries:

My understanding of this literature is that these trends are most pronounced in English-speaking countries (please correct me if I have this wrong), which begs the question of why this might be the case.

One story that suggests itself immediately is that of the 'brain drain'. If high-skilled workers in the various English-speaking countries could credibly threaten to move to the United States, then this outside option could be leveraged into ever-higher salaries for the people at the very top of the income distribution. Many people have their own anecdotes they can tell that would be consistent with this story.

A key element of this narrative is that the threat to leave is more credible for English speakers: moving to the US wouldn't mean abandoning their mother tongue. In particular, English-speaking Canadian migrants to the US have little difficulty adjusting to life in the US.

But there is a large sector of the Canadian population for whom the attractions of life in the US may be attenuated. Francophone Canadians know that moving to the US means abandoning their mother tongue and raising children in English. And for francophone Quebecers who grew up in a French-speaking environment, this may be a high cost to pay.

In their earlier study on Canadian top incomes, Mike Veall and Emmanuel Saez noted that the trends toward the concentration of incomes are markedly different for Quebecers who filed in French and for those who filed in English (an important caveat: some may have chosen not to file in their mother tongue). Here are the wage income shares for those who filed in French in Quebec, in English in Quebec, and for those who filed in the rest of Canada (source is the updated data set):

The incomes of those who filed in French in Quebec do not display the same trends as in the anglosphere. In fact, the trend toward the ever-higher concentration of income is most pronounced for English filers in Quebec. The brain drain story here would be that the threat to leave would be more credible for high-skilled anglophones than for francophones with equal skills.

Plausible as this story may be, it is not entirely convincing. When Mike Veall presented his updated numbers at the CEA meeting last weekend, he mentioned that he really didn't believe the brain drain story anymore. It turns out that Quebec isn't really that much of an outlier:

The concentration of high incomes in Canada is happening mainly in Ontario and especially in Alberta. We could presumably extend the narrative and claim that residents of (say) Atlantic Canada are greatly attached to their home provinces and cannot credibly threaten to leave, but this would greatly dilute the power of the original brain drain story.

Anyone got any better ideas?

Comments

[Luis Enrique]

[irrelevant pedantry: it does not beg the question]

[Nick Rowe]

Random thoughts:

My God those are big changes in the top graph! And was the Post-WW2 period the anomoly, and we are now merely returning to normal?

The "credible threat" to migrate might work in the case of bilateral monopoly-monopsony, where the high-income person's outside option determines the threat point, and this affects the equilibrium in the Nash bargaining solution. But each of these countries is relatively "large", so wages would not be determined in bilateral monopoly-monopsony. I think you would need *actual* migration, not just the credible *threat* of migration, to change supply, and incomes.

The "brain drain" hypothesis might explain why all these 5 countries move *together*, but it wouldn't explain why they *move*.

These are stunning graphs.

(And I've been carefully misusing "beg the question" in my previous posts. Damn! And to be corrected on this by a non-native English-speaker!)

[Nick Rowe]

Correction: Anthony *Atkinson*, not Anderson. Big name in income distribution economics. And still going strong, after all these years! (I think I used his text the one and only time I taught "Income Distribution and Poverty", while a grad student at UWO in 1979.)

[Stephen Gordon]

Gaah - fixed.

[Frances Woolley]

"The incomes of those who filed in French in Quebec do not display the same trends as in the anglosphere."

It's not the individual who files, but their accountant. Why might ultra-high income people be more likely to use an English-speaking accountant? Or, strictly speaking, an accountant who has English-language tax-filing software? Ultra-high income probably have really complicated income taxes with, e.g. income from multiple jurisdictions, all sorts of tax planning issues. There's only a limited number of firms in the country that provide this kind of advice - see, e.g. http://www.taxspecialistgroup.ca. This result could be driven by the software used by a single firm in Montreal.

[Stephen Gordon]

Yeah, that's why I added the caveat. Apparently, one of the things on Mike's 'to do' list is to match up the tax files with the census data to see what the overlap is between francophone/file in French and anglophone/file in English.

[Nick Rowe]

One thing that strikes me as very very weird about these figures, if I can just get personal for a minute:

OK, my income is somewhere around the top 1 percentile in Quebec. How come most cars on the road around here are newer and more expensive than mine? How come lots of people live in more expensive homes than mine? How come there are lots of expensive restaurants around that I don't eat in? How come most people I see are dressed in more expensive clothes than me? And how come somebody must be flying off to expensive holiday destinations. Etc.

OK, I'm a cheapskate, but nevertheless.....

In short, how the hell can everybody else afford to spend what they seem to be spending? It just doesn't seem to add up.

And I can't believe the answer is: "They went into debt". Sure you can buy an expensive car on credit, but you still have to pay eventually, and you pay more in the long run. This has been going on for decades. We must be in the long run by now.

[Frances Woolley]

I'll ask my cousin who does high-end tax accounting (though not in Montreal). This is the kind of thing that Alan Macnaughton (editor of the Canadian Tax Journal and accounting prof at U Waterloo) would know about.

[Patrick]

How much do you save/invest? Lots of people spend every cent they make.

[Patrick]

What are those rich people doing to earn those wages? Apart from the hockey players and the Raptors, I mean. My guess is that in Ontario most of them are in finance. And in AB ... oil & gas? Maybe some real estate too (until recently anyway).

[ClaudeB]

Here's what the Saez and Veall (2005) said about filing language in Quebec (p. 841, note 14):

Francophones are defined in the data as those who complete their income tax returns in French. Hence an alternative explanation for the smaller rise in francophone top income shares could involve high-income francophones choosing to file in English. Without independent information on language status, we cannot rule this out. We do note, however, that accounting firms in Quebec almost invariably allow a client to choose filing language. We also note that our tax-based data do not contain information about education or occupation, so we cannot determine whether those who file in French are different in other respects.

[Simon van Norden]

Looking at the first graph, I find that the various national troughs (late 1970s and early-mid-80s) correlate well with specific political events (Thatcher's election in Britian, Reagan in US, wholesale reforms in NZ) in several countries. A key element in all these reforms was cutting marginal tax rates, which meant large net tax savings for those at the top of income distribution.

I'd be curious to see how well those graphs line up with top marginal tax rates on earned income in each country.

I'd also be curious to know whether a media/political mechanism might be part of the explanation. Given a common language, there is considerable spillover from media outlets in the US and UK into other english-speaking markets. Given a change in the political climate in the UK and US, I'd expect this to propagate more effectively to other english-language markets than to non-english markets (e.g. Italy, Greece, Germany, Spain, Japan.)

[RSJ]

"The "brain drain" hypothesis might explain why all these 5 countries move *together*, but it wouldn't explain why they *move*."

That seems right, but the argument flew right over my head. I never felt comfortable with game theory, and don't follow the reasoning :(

If I can throw my own 2 cents in, I would say that reducing the marginal tax rates, as well as capital gains taxes, was the main culprit. Union busting as well. Cold war solidarity was also important.

While the social construction of monetary policy posts are also over my head, I do believe that the marginal product of labor is a social construction.

Again, if there was only a single laborer, this measure would make sense. But, HPQ has 300,000 employees and revenues of 120 Billion. How can you determine that the marginal product of one of these workers is X while that of another is Y? Are they really all bricklayers, laying X bricks per day?

Businesses that we have today are not generating real output growth at a faster rate, and yet the ratio of CEO compensation to average worker compensation has grown about 8 fold from the 1970s to the present. From the 1930s to the 1970s, this ratio shrank by half.

The whole superstar earner phenomena *seems* like a social construction, one which was successfully suppressed in a different tax/industrial relations regime.

As an amusing example of what our superstar earners are actually doing, let's take a look at Avis rent-a-car ("We try harder"):

"
IN 1946, Warren E. Avis (who died last month at the age of 92) had an idea: rental cars should be available at airports. So he founded Avis Airlines Rent-a-Car. In 1954, he sold the company to another businessman, Richard Robie. Two years later, in 1956, Robie sold Avis to an investment group led by a company called Amoskeag. In 1962, the investment banking firm Lazard Frères bought Avis. In 1965, Lazard sold Avis to the giant conglomerate ITT Corporation.

Since 1946, Avis has been sold or reorganized 17 or 18 times, depending on how you count. Each time Avis changed hands or structure, there have been fees for bankers and fees for lawyers, bonuses for the top executives and theories about why this was exactly what the company needed.

In 1972, ITT spun off Avis as a publicly traded company. Then, in 1977, the company was bought by another giant conglomerate, Norton Simon. In 1983, a company called Esmark (formerly Swift & Co.) bought Norton Simon. In 1984, Esmark was bought by Beatrice Foods, and in 1986, Beatrice was bought by the leveraged buyout firm Kohlberg Kravis Roberts & Company.

Kohlberg Kravis Roberts immediately sold Avis to an investment group called Wesray. Wesray sold Avis’s fleet leasing business to a company called PHH Group. Then it spun off Avis’s foreign operations and took them public as a company called Avis Europe P.L.C. And then, in 1987, Wesray sold Avis to its employees under an employee stock ownership program. Wesray more than tripled its money in 14 months.

Two years after the stock ownership deal, the company sold General Motors a complicated security that effectively gave it a 26 percent stake in Avis. Apart from that, Avis’s employee ownership experiment lasted nine years, until 1996, when Avis sold itself to a company called HFS. Employees got an average of $26,000 each. Eighty or 90 current and former Avis executives got an average of $1.75 million each.

A year later, in 1997, HFS took Avis public. (The initial public offering raised just over $330 million. The banker Bear Stearns charged $15 million for its services.) In 1999, Avis bought PHH. Remember PHH? That was the company Avis sold its fleet leasing operation to in 1987. PHH was owned by Cendant, a company that had been formed in 1997 by the merger of HFS — right, the company that had spun off Avis in 1997 — and another company called CUC. HFS had retained 19 percent of the company’s stock when it took Avis public. With the stock portion of Avis’s purchase price for PHH, Cendant now owned 34 percent of Avis.

A couple of years later, Cendant bought the roughly two-thirds of Avis that it didn’t already own and made Avis a wholly owned subsidiary.

In 2006, Cendant split itself into four independent companies, one of which was the Avis Budget Group. (Somewhere along the line, Cendant had also acquired Budget Rent a Car.) The Avis Budget Group became the parent company of Avis Budget Car Rental...

Last September, a week after the Avis Budget Group began trading on the New York Stock Exchange, The Wall Street Journal reported that the new company was “ripe for the picking.” Carl Icahn, another wily financier from the 1980s, had acquired a $100 million stake in the company and would not comment about his intentions...
"

[Determinant]

I've seen this argument before. In the 1950's the role of CEO was that of a manager. The corporation was permanent, it's business fixed. The CEO was expected to be and paid as a senior administrator. But the shift began in 1954 with Robert Young's proxy battle for the New York Central Railroad. By the 1980's CEO's were expected to be corporate buccaneers, for lack of a better term. They were paid correspondingly. It's not so much a wage as prize money.

[Phil Koop]

I think Patrick is in the right of it. A credible "threat" requires not only that consequences of denial be credible, but also that demand acquiescence be viable. It's not just that a Toronto investment banker can credibly threaten to move to NY; his bank can also plausibly afford to pay him more. Similar remarks apply to a Calgary oilman. For a PEI farmer, not so much. He can only leave, not negotiate.

[Matthew]

"But there is a large sector of the Canadian population for whom the attractions of life in the US may be attenuated."

Makes sense. For example, doctor's salaries in Quebec are considerably lower than in the rest of Canada.

[Frances Woolley]

"I find that the various national troughs (late 1970s and early-mid-80s) correlate well with specific political events (Thatcher's election in Britian, Reagan in US, wholesale reforms in NZ) in several countries. A key element in all these reforms was cutting marginal tax rates, which meant large net tax savings for those at the top of income distribution."

which a straight forward tax incidence analysis predicts would cause before tax wages to fall

"But there is a large sector of the Canadian population for whom the attractions of life in the US may be attenuated."

Which reminds me of the papers that have found hours worked in Quebec are lower, on average, than in the rest of Canada.

On cheapskates - some people need to buy stuff to signal their social status, and some people don't.

[Jack]

@Nick: I agree, but the long run can be a long ways away... I assume they are heavily in debt (net debt, they don't save anything either except what is mandatory, if applicable). Some people don't understanding saving, others don't trust it.