Convergence, Big Time

In his influential 1997 paper, Divergence, Big Time, Lant Pritchett estimated:

…that from 1870 to 1990 the ratio of per capita incomes between the richest and the poorest countries increased by roughly a factor of five and that the difference in income between the richest country and all others has increased by an order of magnitude.

Pritchett was correct but Patel, Sandeful and Subramanian show that just where Pritchett’s study ended, convergence began!

While unconditional convergence was singularly absent in the past, there has been unconditional convergence, beginning (weakly) around 1990 and emphatically for the last two decades.

The figure above plots the coefficient (“beta”) from the plain vanilla unconditional convergence regression (relating average growth of real per capita GDP over the long run to its initial level). A statistically significant negative beta denotes convergence and divergence otherwise. Since we know from Johnson et al. (2013) that growth rates vary widely across datasets, we plot the annual betas for three such sets: the Penn World Tables (PWT), the World Development Indicators, and the Maddison Project (Bolt et al. 2014).[1] While the point estimates vary across datasets, the consistent pattern across them all is a statistically significant negative beta since around 1995 (unconditional convergence) and its lack prior to that (see also Roy, Kessler and Subramanian, 2016).

Our basic point doesn’t require regressions. Looking at the 43 countries the World Bank classified as “low income” in 1990, 65 percent have grown faster than the high-income average since 1990. The same is true for 82 percent of the 62 middle-income countries circa 1990.

Neo-liberalism has been incredibly successful, essentially delivering on all of its promises of economic growth, declines in poverty, and peace. Yet, the ideas behind what Andrei Shleifer called The Age of Milton Friedman are now under attack and in retreat.


Vive les Koch Brothers!!! Allez!

Before we all get too carried away, the slight negative correlation between initial income and subsequent growth documented by Patel, Sandefur, and Subramanian for 1995 to 2010 doesn't imply that catch-up will occur or even that convergence has begun. Even if it did, the implied rate of convergence would be very slow – much lower than the “iron-law” rate of 2% per year. The data span that they use is very short (for a formal convergence analysis) and contains important high and medium-frequency influences on GDP such as the dot com bubble and its collapse and then the largest global economic slowdown since the 1930s and its immediate aftermath. The headwinds that they mention have become reality in many of the countries which helped deliver the negative coefficient. The β-convergence approach is flawed and can have low power against non-convergent alternatives such as club convergence so finding a statistically significant negative coefficient does not mean that convergence is occurring.

Wait, we're supposed to ignore nearly 150 years data for a mere 28 years data?

What I find most convincing about inequality is, besides the fact as Piketty observed it increases in peacetime, is how constant it is across all ages and societies: roughly the top 1% control 50% of wealth (not 'income', but wealth is the relevant measure), scheduled to go to 67% by the year 2050 in the UK.

Full disclosure: my family is in the 1%, minimum net worth $10M in the USA to qualify. Sorry I don't mean to come across as a "Sloan Ranger" for you UK readers.

Maybe so, but most human inequality is between countries rather than within countries, and inequality between countries is extremely variable historically: Therefore we can still decrease inequality overall by ensuring that the trends of the last 28 years continue.

Someone posted this interesting link recently. The last 30 years has seen something extraordinary from humanity.$state$time$value=2018;;&chart-type=bubbles

Pritchett's quote seems to be about sigma-convergence whereas Patel, Sandeful and Subramanian write about unconditional beta-convergence. It would be interesting to see if we also observe sigma convergence since 1990. The two will not coincide if growth is heteroskedastic over time, which is the case.

China: achieves double-digit growth for many years, catches up to US.

Tyler: it's a triumph of neo-liberalism!

This is an AlexT post, you can tell by the prose, he has a distinct style.

Isn't it, though?

I don't know. I'd hardly call China a paragon of liberalism, neo- or otherwise. China has taken advantage of much freer trade in goods, but has not opened its goods markets nearly to the extent that Western states have. Capital and labor are still tightly controlled. And let's not forget that much (most?) of the growth China has experienced over the last decade is due to a massive unfunded stimulus program that has seen indebtedness skyrocket. Those are hardly elements of the neo-liberal consensus.

Spent five minutes reading the paper, and that's about five hours for lesser minds, and conclude that China is why the statistics seem to support the OP's argument. Remove China, which essentially is simply an outsourcing center for the West, and the "convergence is happening" and "there's no middle income trap" arguments disappear. Once again, lies, damn lies, and statistics (programs, the authors possibly can't code by hand whatever statistic they're trying to measure, but know what keyboard buttons to push).

India has also had pretty impressive growth since 1990, and has been consistently growing even faster than China in recent years. India and China each have more people than all of Africa, so they cannot be discounted (conversely, if one discounted Africa, which continues to converge slowly, developing country convergence would look even more obvious and impressive). There has also been very impressive growth in Southern and especially Eastern Europe as those countries converge to Western Europe.

India's average GDP per capita growth rate from 2007 to 2017: 5.7%
China's average GDP per capita growth rate from 2007 to 2017: 7.7%

Last five years... India 4.8% ; China 7.6%

Is Eastern Europe converging to Western Europe? Yes and no.

Average GDP per capita growth rate from 1990:

Hungary : 1.6% - no
Czech Republic: 1.8% - no
Romania: 2.6% - slowly
Bulgaria: 2.7% - slowly
Poland: 3.7% - yes

Germany: 1.4%
U.K.: 1.3%
Spain: 1.3%
France: 1.0%
Italy: 0.8%

I'm confused. Shouldn't all the Eastern European countries be a yes?

Hungary and the Czech Republic have a growth lead that is so small over Western Europe that I put 'no' but 'very, very slowly' might be better.

Bonus: GDP per capita growth over the past 10 years:

US 0.7%
Japan 0.7%
EU 0.6%

Hungary 1.2%
Czech Republic 1.0%
Bulgaria 1.9%
Poland 2.9%

Perhaps reading for more than 5 minutes would have led you to a different conclusion. For instance: "Our basic point doesn’t require regressions. Looking at the 43 countries the World Bank classified as “low income” in 1990, 65 percent have grown faster than the high-income average since 1990. The same is true for 82 percent of the 62 middle-income countries circa 1990."

Romain Wacziarg 1
Ray Lopez 0

But Ray is the sharpest tool in the shed - he just said so.

The main cause of convergence beginning in the 80s-90s was the end of communism in China and Eastern Europe and the end of socialism in India. After that, the ex-communist world rapidly began catching up with the West. Because most human inequality is inequality between rather than within countries, that means that ironically communism in developing countries significantly increased human inequality!

@Zaua - "Because most human inequality is inequality between rather than within countries"...I think that's a typo, or you are mistaken, since usually the opposite is true. No poor in the USA cares that their plight, relative to Zambia, is not so bad. Keeping up with the Jones. But you're right that Communism skewed the statistics.

No, about 60% of overall inequality is due to between-country inequality and only 40% is within-country. Even a poverty level income in the US would put you in the global top 10%: Of course, a poor person in the US will subjectively compare himself to a middle-class person in the US and not a poor person in Zambia (just like upper-middle-class people in the US compare themselves to rich people). But quantitively, the story of human inequality is primarily a story of between-country inequality.

Aha! I’ve got it solved. If everyone moves into America, there won’t be any between country inequality anymore.

@Zaua - that's a nonsense comparison. If nobody in the real world is comparing inter-country (as opposed to intra-country) inequality, what's the point of the exorcise? It's of interest to academics only, not real people? As Beliasarius Rex says, everybody should move to the USA so we don't have to worry about inter-country inequality (as well as pick up trillion dollar bills on the sidewalk).

The amusing thing is that hundreds of millions of people would run from the idea of living in America.

The shine is off the apple over the last generation for many people.

"The main cause of convergence beginning in the 80s-90s was.."

There are many moving parts, but I think the information age has done something to make progress more portable. Disintermediation?

Yes, inequality between rich and poor countries is declining, but inequality within both rich and poor countries is increasing. People in poor countries are definitely better off than they were, but I suppose not relative to the rich within the poor countries. This is a distinction often overlooked.

@rayward - it's not overlooked, it's a consequence of the normal distribution and the law of large numbers. All things being equal, as N--> infinity, the pile of sand gets taller and wider. Do you want a WWIII so we can have more equality? That's what Piketty says happens when there's equality, a natural or man made catastrophe causes it. That said, I'm for a minimum guaranteed income, as long as it's very low ($5k a person a year) which will encourage lazy Americans to live in the Third World (like I am, but out of choice, remember, I'm in the 1%).

"That said, I'm for a minimum guaranteed income, as long as it's very low ($5k a person a year) "

That's an awful idea.

"which will encourage lazy Americans to live in the Third World"

Yes, and that's one of the reasons why it's an awful idea.

A much better/expanded EITC would have far less deleterious results for America. It would still protect us from a rapidly automating market that was putting people out of work. The expanded EITC would allow people to gracefully transition to productive roles where they had a comparative advantage over automation.

That came across as far to harsh Ray.

""That said, I'm for a minimum guaranteed income, as long as it's very low ($5k a person a year) ""

Kudos on not labeling it a "Universal" income (because those plans never are universal) and for setting it at a feasible level. Your idea is plausible and would probably even be beneficial. Though not as beneficial nor as flexible as some variant of an expanded EITC.

The EITC is the way to go. Though the employment situation is fubar - too many low paying demeaning service jobs catering to horrible rich people. Everybody can and should work. Too many parasites is unhealthy. We can all work kissing the asses of the rich while subverting them at every opportunity. It will be especially satisfying to be a nurse or home health aide.

Or you could try treating people the way you'd want to be treated. Crazy, I know, but it just might help you get ahead in life.

@JWatts - thanks, no problem, call it what you want, but even a fellow at the Hoover Institute (!) was in favor of a minimum income for people if it was low enough, I believe he mentioned $5k, it's where I got the figure from, as it's actually cheaper than welfare and possibly equivalent to your proposal.

This is not the ideal place to discuss this, but I'm interested why you feel shipping US poor to the Third World: ME: "which will encourage lazy Americans to live in the Third World" YOU: Yes, and that's one of the reasons why it's an awful idea." is so bad? You possibly feel the USA would be constantly having to rescue people from kidnapping situations, but all the USA would have to do is say: "once you leave the USA, you're on your own, no Marines will come rescue you". That's the way it is in most countries. The USA should not be 'world policeman' for anybody, not even their own citizens.

"why you feel shipping US poor to the Third World is so bad?"

It's not as specific as kidnapping. It creates a terrible incentive problem. Those people would be able to vote per current US law and there incentives would be drastically different than that of those living in the US.

First, they would have a direct incentive to support any candidate who would increase the amount they receive per year, since the costs are entirely born by others.

Second, they would have a direct incentive to support active intervention by the US (both militarily and politically) in any country they were living in. This would lead to something similar to the US interventions in Latin America from a century or more ago.

Third, this would lead to a drain on low skilled later and an inevitable push to replace the poor Americans with hard working immigrants. Which would effectively create a two class society. I would foresee a sizable resentment from Mexicans (or any third worlder) immigrating to the US and working hard to pay taxes to support the lazy Gringoes, living in Mexico on the taxes they were paying.

Effectively the idea strike me as an Ivory Tower proposal, that might sound good in theory, but would collapse under the pressure of normal human behavior.

"an Ivory Tower proposal, that might sound good in theory, but would collapse under the pressure of normal human behavior."

-so, like communism and anarcho-libertarianism

A flip side of the Global Information Age is that it does allow cumulative advantage on that global scale.

Zuckerberg is as rich as he is because he can serve a product to billions. The gulf between a Bangladeshi rice farmer (with a cell phone and Facebook) and Zuckerberg himself must be the most dramatic on the planet.

I think you have a missing comma. "Decline in poverty and peace" suggests both poverty and peace have declined.

On a less pedantic note, Robert Lucas discussed these issues in an issue of the JEP at the start of the new millennium. He argued that poorer countries will learn from richer countries and reform, and then converge. His model suggested that the world was going through peak inequality (national level, not personal level) in the late-1990s.

In any event, I am not sure why we care about inequality at the national level. Let's focus on convergence at the personal level. Do we see global personal convergence? I suspect yes, but slowly.

If monkeys care about inequality, why not us?

But sure, if we are dispassionate utilitarians we should focus more on who is in need.

Because we shouldn’t design our economy or political system based on the ethics of chimps.

HIV meds and mosquito nets for Africa : Yes.

Tearing down Stuyvesant and keeping “the yellows” out of Harvard : No.

We might agree about mosquito nets and Harvard, but I don't agree that we can ignore our monkey roots.

It's behind everything, including the idea that you need a Big Man (or lately Woman) to run a country.

It was a small, dirty thing, but when Donald loomed physically over Hillary in the debates he was playing to the monkey brains in the audience.

We should strive to overcome irrational things, even if we cannot fully expunge them. Chimps also kill male chimps from other tribes, but that doesn’t mean we should beat immigrants to death and then eat their corpses.

The loathing of freedoms of choice, association, and commerce may be a part of our genetic heritage, but it’s still idiotic.

That's a very selective list. We are also (from the economists perspective) unreasonably cooperative and share way to much of our wealth.

Probably just because we are a "get along" social species, we are very geared toward "niceness."

Your view of the “economist perspective” is both hilariously naïve and 40 years out of date.

Humans follow their own utility function. If it involves giving money to charity or telling the offerer of an unfair deal in a game theory experiment “f you” then so be it.

Offering to share wealth or income is fine. Using violence to steal property in the name of social engineering is not.

Did you actually say my view is out of date and then immediately follow that with "utility function?"

We are biological animals, wet and squishy. No utility function captures our often conflicted biological drives.

“Neoliberalism has brought peace”

I was wondering what was going to cause my morning coffee spit take.

That bit of magical wishfull revisionist hand waving nonsense was truly exceptional

In what way?

Curios, it is almost as though the policies that allowed rapid growth in the rich companies (light effective regulation, low compliance costs, abundant population, and low distortation taxation) are being implemented in high growth countries. In contrast the high effective compliance costs and low population growth that previously marked low growth areas is being implemented in high wealth areas.

How much of this convergence "success" is due to good policies in poor areas and how much is due to poor policies in wealthy areas?

Ultimately, global GDP growth was ~5% back in in the diverging era. Now it is around 3.5%. I would much rather live in an unequal world expanding 30% faster. Certainly the frontiers were pushing further faster back then.

I think Tyler makes a similar but slightly different argument. He says sure the Chinese are polluting their rivers, and we would never do that, but it is worth it to them because of the rapid progress.

Personally I feel rich enough that I do want clean rivers, salmon and trout.

I wonder if we would've found a cure for malaria and a cheap solution for climate change in that missing 30% of growth. I guess we will find out in 20 or 30 years.

Comments for this post are closed