Bryan Caplan  

Civilization Bets Considered

Sticky wage model test: The re... Henderson on Trump...
My civilization bet with Garett Jones fizzled out.  Now it's time to review all my other offers.
Richard writes:

The hardest, fairest measures I can think of are low birth rates plus increasing wealth. Might want to use polity scores too. Then again, Japan has all these things, and many don't see it as "Westernized."

All reasonable, even though I see low birth rates as a bad thing.  Per-capita GDP will keep rising.  The Total Fertility Rate will decline in the Third World and remain below replacement in the First World.  Polity Scores will rise.  I'll bet on any or all of these for the next fifty years.
Maxim writes:

Because "western civilization" is so broad, I wonder if the least bad way to do this would be to track Pew's polls of people in other countries and "do you have a favorable view of... [The US, France, etc.]" That certainly seems to be the most literal way to do it, though I suppose the drawback is when western civilization acts contrary to the *ideal* of western civilization, such as but launching a war or something. Which are we trying to measure here, western civ as it actually exists or as our ideal vision of it?

If the ideal vision, you may want to measure with some basket of values (like Western movie sales + sales of The Wealth of Nations + non-religiosity in surveys, or something like that...

As written, these are too vague to bet.  But suitably refined versions seem reasonable.
ChacoKevy writes:

How about tracking the S&P vs Hang Seng in ability to generate revenue in neutral global markets?

I see China's entry into the capitalist world as a great triumph of Western civ.  So comparing S&P vs. Hang Seng seems off-point.
Dangerman writes:

I don't think defining "reasonable" is the hard part of this bet... I think defining "Western" civilization is the difficult part.

What about metrics that used to commonly be associated with "Western Civilization" like: (1) percentage of the population that identifies as Christian; (2) the pervasiveness of monogamy as the fundamental social structure underlying family formation; or (3) stable to above-replacement fertility rates.

I predict that over any given moderate timeline (2-10 years) at least two out of three of the above metric will decline, and so "Western Civilization" is NOT in fact winning.

Will pre-pay $100.

"Western" is indeed the rub.  I see religiosity, monogamy, and high fertility as three ways that the pre-20th-century West resembled other pre-modern societies.  Not only are these traits not distinctively Western; they're now distinctively non-Western. 
Andrew S writes:

I'd go for measures that emphasize individual liberty to quantify western civilization (especially versus places like Beijing or Tehran). Measures I would use:

1) Percentage of articles published in a country's major newspapers (or online news sites) critical of its government.

2) Number of officially banned books

3) Percent of country in jail for victimless crimes (drugs, politics, speech, etc.)

4) Correlation between probability of being in jail and personal wealth prior to being charged

5) Number of an enumerated list of individual rights that are not prohibited or restricted by law (practice religion x, travel within country, dance in street, ...)

(1), (2), (3), and (5) all seem good as long as we've got solid measures.  (4) could, for all I know, be higher in Western countries, so I wouldn't want to bet on it.
mico writes:

Observant Islamic population of the globe will grow faster than the non-islamic population of western countries.

Most western countries will have both immigrant chauvinist and native nationalist parties polling over 20% of the vote consistently within 30 years, except Asian countries whose populations will collapse.

The current western culture set doesn't even manage population replacement so in the absence of genocidal treatment of others (again not western in the sense you mean) the question is when not if it dies.

Since the non-Islamic population of Western countries is shrinking, I'd lose the first bet unless there's a high threshold for "observant."  The chauvinist/nationalist party bet is interesting, but too vague to bet.  Is the U.S. Republican Party already "native nationalist"?

Aside: A culture that makes converts doesn't need biological population replacement to survive.

Eric Rall writes:

1. Percentage of population proficient in English as a first or second language. Proxy for Anglo-American cultural influence and for the value placed on business, educational, and cultural contact with the West. Might also be worthwhile to include French and German proficiency, if you don't want to just use Americanization as a proxy for Westernization.

2. McDonald's franchises relative to the population. McDonald's is widely seen as a symbol of Westernization in general and Americanization in particular, so franchises flourishing signals both cultural acceptance of Westernization and institutional access to western businesses.

3. International revenue for Hollywood movies. Same logic as #2, with emphasis on cultural acceptance.

4. Percentage of Western-educated people in powerful or influential roles (political office-holders, CEOs, college professors, etc). Potentially problematic, since a decline could indicate an improvement in the perceived quality of domestic educational institutions rather than a decline in value placed on a Western education.

I'll bet on more specific versions of (1), (3), and (4) .  (2) is too narrow, since McDonald's is an inferior good.  Multinational franchising in general though is bet-worthy. 
Grant Gould writes:

These are suggestions, not offers to bet, as I doubt our views diverge enough to find odds between them.

* Fraction of earth's population who are legally permitted to blaspheme every god, prophet, avatar, and messiah
* Number of countries in which a double-digit fraction of the population has seen one of the highest-grossing American films of the past ten years (note: Measure must include estimate of people watching via piracy!)
* Fraction of the world population that believes that fraud and embezzlement are more serious crimes than divorce or homosexuality
* Fraction of the world population where the expected (severity times likelihood) punishment for tax evasion exceeds the expected punishment for complaining about tax rates

The "legally permitted to blaspheme every..." requirement is so stringent, and population growth in the First World so low, that I wouldn't want to bet at even odds.  I would bet that a continuous global ranking of de facto religious freedom will rise.  (2) and (3) seem good, as does (4), assuming low expected punishment for complaining is what counts as "Western."
Peter H writes:

So we want a measure that's concrete, tied to participation in the global community, and hard to game.

How about change in % of GDP that constitutes foreign trade? It's imperfect as a measure of westernization, but not terrible as it tells us about engagement with the world at some level. First derivative of the foreign trade percentage gives us the answer to "getting more or less western" and has a nice zero point, as well as controlling for current patterns - we're just looking for direction of the trend.

So what say you Bryan?

I'll bet on this over the medium-term (10 or more years), even though there are signs of a cyclical reversal right now.
phil writes:

I think total share of world economy held by western countries (US, Canada, EU, Australia) is likely (maybe almost certain) to fall

you could probably find a lot of takers at a certain ratio of that question

I suspect you largely attribute that to a 'Westernization of the rest' effect and wouldn't find that a particularly attractive bet


Daniel Fountain writes:

I would classify "Westernization" as "expanding individualism". That is to say a respect for the choices of the individual. Modern Western Civilization is, at its heart, a society of live and let live. Obviously there are grotesque exceptions but this is the creed by which the west has taken over in the last 200 years or so.

For legal institutions there is one clear metric by which to gauge: Percent of individuals in jail or fined for a non-externality inducing actions. Also along these lines is the percent of laws dedicated to banning non-externality inducing actions.

Since virtually everything creates externalities, this isn't a good bet.  Betting on fraction of the population in jail for canonically "victimless crimes" is okay, though.

For social institutions the biggest innovation western civ has to offer is the education of all groups. So a country could be said to be westernizing if the mean education level of women is closer to the mean education level of men over time.

If you rephrase to "average female education minus average male education is rising," I'll sign.

Similarly the lack of racial crimes is a core tenant of modern western civ, so the less racially based crime there is the more the country is westernizing.

Interesting, but too vague to bet.

Brad writes:

How about the number of visa applications from China,Russia,and other non-western Nations. Measure where the world's people are moving.

If they move to Western countries, does this show Westernization or the opposite?

Bottom line: If I said a metric is good, I'm ready to bet on it at even odds once the metrics and stakes have been clearly specified.  For the rest, I'm open to revisions.

COMMENTS (9 to date)
E. Harding writes:


-China? Polygamous? Arabia? Polygamous. Africa? Polygamous.

Who am I missing?

Dangerman writes:

"I see religiosity, monogamy, and high fertility as three ways that the pre-20th-century West resembled other pre-modern societies. Not only are these traits not distinctively Western; they're now distinctively non-Western."

Two additional points:

1. I didn't want to bet on "religiosity" - I said "Christian" specifically for a reason. I don't think there's too much debate about the connection between Christianity and Western Civilization.

2. On the other hand, I get the sense that there IS debate (these days) about the connection between monogamy and western civ. Likely because monogamy and western civ HAD won over so much of the world. I would cite the following as a mere sampling of how people thought before this was taken for granted:

A. "Polygamy has always been odious among the northern and western nations of Europe, and, until the establishment of the Mormon Church, was almost exclusively a feature of the life of Asiatic and of African people." - SCOTUS Reynolds v. United States (1878)

B. The work of William Tucker:

C. The work of Joseph Henrich - "The anthropologically peculiar institutions of imposed monogamous marriage may be one of the foundations of Western civilisation."


So, with respect to at least the above two points... if you agree that "they're now distinctively non-Western" then we're agreeing that the West is in decline.

Bob writes:

Bryan, I appreciate that the issue here is defining "Western Civ." I just finished a book called In Search of Civilization by John Armstrong and am now watching the old BBC series "Civilisation" hosted by Kenneth Clark.

What I've taken away from both is that Civilization is indeed quite hard to define. So much so that the parallel is Potter Stevens "I'll know it when I see it" quote.

Not that any of that helps define your bet, but it does help validate why it's so tough to do so.

Pietro Poggi-Corradini writes:

This might be measurable: "how closely does a country or zone's central bank comes to targeting NGDP."

It seems to me from reading Sumner, that anglophone countries such as England, Australia, Canada, and the US have marginally better monetary stances. Will this continue? Will other areas be better?

Nathan W writes:

I think the measures proposed by Andrew S are generally consistent with what I see as the worthwhile parts of "modern" Western civilization.

Is there such a thing as a "non-Western" thinker who advocates for individual rights or freedom of conscience? I am not aware of any examples.

Miguel Madeira writes:


-China? Polygamous? Arabia? Polygamous. Africa? Polygamous.

Who am I missing?»

"Monogamy" compared with "polygyny" is Western; compared with "promiscuity", perhaps not.

And note that, in most "polygamous" societies, usually most marriages are monogamous; in the end, perhaps the share of persons living in monogamous marriages could be higher in non-western societies.

Probably the better indicator of "western values" in marriage is the age of marriage - more higher, more "western"

mico writes:

"Since the non-Islamic population of Western countries is shrinking, I'd lose the first bet unless there's a high threshold for "observant.""

You wouldn't win it in that case either. Islam is not moderating and secularising into oblivion like Christianity. The ultra-observant fraction exemplified by people like Al-Qaeda and Islamic State is much larger, not smaller, than it was in 1950 or 1980. The growth rate of Islam probably increases with observance, albeit the ultra-observant are still only a small minority of the total Islamic population.

"The chauvinist/nationalist party bet is interesting, but too vague to bet. Is the U.S. Republican Party already "native nationalist"?"

In my opinion no, but it will be if it selects Trump as its candidate. The Democrats however are already an immigrant chauvinist party, not necessarily because they are more extreme but because they can express their beliefs openly with much less scrutiny or criticism. I agree it's too vague to bet, but I doubt many will disagree with the qualitative description in hindsight.

"Aside: A culture that makes converts doesn't need biological population replacement to survive."

It is possible for a culture to survive purely by conversion but it is not stable. One example of a culture with a persistently high birth rate, a relatively high rate of deconversion, and almost zero conversion is the Amish, who have consistently grown considerably faster than the host culture in which they live. In the long term, it is a question of when and not if such cultures arise and eventually displace the host culture.

DD writes:

I still like my Night Clubs per Capita metric (which was ignored by Caplan). Look at places like the Middle East which were much more Westernized in the 70's than they are today. One thing they had back then which they certainly don't have now? Places to drink alcohol and socialize.

Could divide the metric by GDP as well so that you're not just picking up countries becoming richer.

Dangerman writes:

Relevant today:

Comments for this entry have been closed
Return to top