Bryan Caplan  

Multiplicative Growth: A Just-So Story that Deserves a Hearing

PRINT
Ozimek's Challenge to Paternal... Should the Cost of War Include...
Here's Greg Clark on growth over at Cato Unbound:
I fully agree with McCloskey about the surprisingly poor ability of incentives alone to account for growth. In order to hold on to the central idea that the 10,000-year delay in the Industrial Revolution from the first appearance of settled agriculture was created by a lack of incentives, economists have to maintain the collective fiction that all societies before 1800 were run along the lines of Kim Jong-Il's North Korea. Yet, in case after case, we find, deep in the 10,000 years of economic stagnation, fully incentivized market societies.
Clark is basically expanding his position in A Farewell to Alms: Since England had good policies for centuries before the Industrial Revolution arrived, good policies can't explain modern economic growth.  Other economic historians I consulted told me that Clark neglected many important English policy changes between 1500-1800, such as the enclosure movement.  But suppose that English policy were utterly stagnant between 1200 and 1800.  Would that imply that policy was irrelevant?

No; it would merely show that policy is not enough.  But the facts are quite consistent with a multiplicative model.  Maybe growth requires policy, population, and science; when Policy*Population*Science hits a critical value, growth takes off.  Clark might dismiss the multiplicative model as a just-so story, but he shouldn't be so hasty.  Low scores on any one of these variables really do seem to choke off growth, and high scores on all three really are a near-guarantee of growth.

1. Policy.  The natural experiments of North and South Korea, West and East Germany, mainland China and Hong Kong/Taiwan/Singapore, show that policy makes a huge difference.  Awful policy is enough to overpower every other advantage combined.

2. Population. As Simon, Diamond, Kremer explain, small isolated populations are extremely backward.  Slow discovery of knowledge is a big part of the problem, but inability to take advantage of scale economies is a huge handicap too.  This is why people can't escape the consequences of bad policy by moving to a desert island. 

3. Science. Modern production depends heavily on modern scientific knowledge.  Indeed, there was never a truly rich society before the rise of modern science.  So science looks like science another crucial ingredient for growth.  Without it, good policy and large population aren't enough.

Clark's right to argue that if policy were the sole key to growth, England would have enjoyed an Industrial Revolution much earlier.  But he's wrong to neglect an equally show-stopping point: The Industrial Revolution began in a country with exceptionally good policies.  A multiplicative model can explain both patterns.  Growth didn't start earlier in England because other crucial ingredients - like population and science - weren't strong enough.  But one important reason why England grew first is that its policies were exceptionally good, giving it an edge over other countries with comparable advantages in population and science.  (In fact, you could say that England's openness to trade vastly swelled its effective population).

Now you could ask: If growth is really multiplicative, why is it so important for the Third World to focus on policy?  For science, the answer is simple: Most of the critical knowledge is already free for the taking.  For population, the answer is only slightly more complex: As long as your economy is open to the international economy, it's the world's population that matters, not your country's. 

The upshot: Third World countries face much more favorable circumstances than England did in 1800.  England had to line up three ducks.  But two of the Third World's ducks are already lined up.  If the world's poor countries adopt tolerably good policies, they will grow.


Comments and Sharing





COMMENTS (15 to date)
gabriel rossman writes:

Also see Kremer's 1993 "O-Ring Theory of Economic Development" from QJE for a similar multiplicative theory of development

Hyena writes:

Are we sure England's policies were particularly good or that science was particularly important? China seems to have had better policies as early as the 13th century, a much larger population, far more technical expertise, large numbers of mechanized designs (in agriculture and industry) and a massive urban population.

China, in fact, should have been the site of the industrial revolution, even on your account. As Needham has made clear, it had every element but the revolution itself.

Andrew writes:

"In fact, you could say that England's openness to trade vastly swelled its effective population"

And surely its overseas empire as well?

John Thacker writes:

Arguably growth started in the Netherlands in the 17th century during the Dutch Golden Age before shifting to the UK in the 18th century. Population surely played a role in the latter.

Evan writes:

Hyena, I don't have it on hand, but I seem to remember Thomas Sowell discussing medieval Chinese policy in Knowledge and Decisions and addressing your questions. Sowell argued that while China did have excellent policies for a while, eventually dogmatic control-freak scholar officials instituted policies that cut off and stifled growth.

Jared Diamond makes a related point that the reason Europe eventually "beat" China was that China was always more united than Europe, so one bad government could topple the whole growth process. Europe, by contrast, was united enough to enjoy the benefits of trade, but not united enough that one government controlled all policy.

mdc writes:

btw, how do we even know that growth was "stagnant" before the industrial revolution? For sure it was less than we see now, but growth is an exponential function - which always do look 'approximately flat' far enough back in time. Since we examine history generally in vague and barely justified stereotypes rather than mathematically, we could easily be interpreting a growth rate as great as today in % terms as 'stagnation' simply because the base economy was so long in comparison with ours today.

If you think this is unlikely, then look at an area of history where most people are at least vaguely familiar with advances in technology: warfare. The development of gunpowder weapons and ships in particular is a quite continuous progression from the late middle ages, not a sudden occurrence starting around 1700 or 1800 as we are led to believe by our stereotypes about agricultural and industrial progress.

Institutional factors then fit the growth evidence - just like they do in the present.

mdc writes:

*base economy was so low

Jnye writes:

I'm with Bryan on this one. The fact that later industrializers -- like China and the Asian Tigers -- have seen rates of growth that are vastly superior to those of any European leader in the nineteenth century indicates that the factors are both multiplicative AND often cumulative. Policy then leverages your ability to reach the technological frontier in a brief period of time and to push it forward. Conversely, a world of high transactions costs and low tech is a world where the opportunity costs of bad institutions are lower. As a consequence, societies may "choose" (or if you prefer "elites" choose) social arrangements that don't necessarily maximize wealth relative to order, stability, status hierarchy, etc. because the benefits of doing otherwise were small -- at least in the short run.

Even today, many countries stay poor, because those in charge prefer the benefits of existing social arrangements to a quintupling of per capita income. Imagine how much easier it would be to sustain bad govt in a world where change might buy you a mere doubling of income and only after many decades of unrest and instability.

I made a brief argument of this sort in my 1997 essay "Thinking About the State."

liberty writes:

Wow - I agree with this post so much I was sure it was Arnold and not Bryan who wrote it :)

You pretty much address these points anyway, but I will note my minor quibbles. The only thing I would change from the above is the shorthand for #2, rather than "population" I would say "scope of society" or something--size of land and population provides this within a given country, but a tiny island like the UK or a small country like Poland or Luxembourg can still reach the intense growth than a large country like the US or China can reach, so long as they have free trade with other countries.

So, "policy" can sort of do #1 and #2, but there is a distinction. The "policy" of #1 is about how easy and worthwhile it is to do any kind of production and trade within or outside the borders of the country; #2 is about resources - are they in the country? Can they be imported? Are their customers outside the country if not within it? etc.

#3 of course is endogenous to a large extent. The initial spark of modern science probably came after trade and good policy allowed for a moderate rise in the living standard of some part of the population (its more likely that someone with a bit of free time on their hands invents something than someone living a Gulag-like lifestyle); and once the economy really takes off, many inventions follow in support of new industry.

fundamentalist writes:

What we know about the economic growth numbers in Europe in the early modern period comes mostly from Angus Madison. In one of his books, Madison stated that the first nation to create secure property rights was the Dutch Republic. That correlates well with Douglass North's ideas about the importance of institutions to development: private property is the most important institution.

As John Thacker mentions, the industrial revolution began in the Dutch Republic, not in England. See Jan de Vries "The First Modern Nation" and Jonathan Israel's "The Dutch Republic." All the Dutch did was to create institutions that protected property from theft by the state or nobility. England imported Dutch institutions with the "Glorious Revolution" of 1688.

Why did the Dutch protect property so well? I asked Israel that in an email once and he said he didn't know. But I suspect is had to do with the fact that Lessius had transmitted the thinking of the Late Scholastics at Salamanca, Spain to the Netherlands and the Dutch simply decided to implement that thinking.

As for science, it didn't begin to contribute to economic development until the 20th century. Before, tech progress came from mechanics and tinkerers. And as others have pointed out, steam power was known to the ancient Greeks; the Chinese and Ottomans were much more advanced scientifically than Europe, were much wealthier, had greater populations and larger cities and, in short, had all of the advantages that most writers on development claim spawned the industrial revolution, but they never did.

Craig Fratrik writes:

In The Rational Optimist, Matt Ridley posits that industrial innovation begets technology, which begets science, not the other way around. I'd be interested to hear the Mr. Caplan's thoughts on Mr. Ridley's arguments.

Tracy W writes:

One argument is that Britain had coal and iron ore close to each other, so you could dig out coal and use it to extract the iron from the iron ore without having to pay a vast cost in moving the coal by land. Then as you had coal and iron ore right next to each other, you could invent a railway, which was the first cheap way of moving heavy goods by land.

After that, any country wanting to get rich can get the coal and iron sent in by ship, and start building the railway from the port out.

One thing on my mental list of things to do is to get a world map of coal and iron deposits and check this story, including seeing if there were other countries that could have done this.

Left Outside writes:

Hey Bryan,

Very interesting, but I have a problem with this sentence "But one important reason why England grew first is that its policies were exceptionally good, giving it an edge over other countries with comparable advantages in population and science."

The thing is, England didn't grow first, extensive (i.e. total but not per capita growth) has existed always, and pretty much for all human societies.

The first clear cut example of intensive growth wasn't England, it was Song China. In the 11th century iron production in China hit levels not seen until the late 1700s in Europe. Likewise, the agricultural sector was seeing a turn to production for the market not seen in Europe for some centuries. Science was very advanced in Song China even compared with early modern Europe.

Your question can't just be "why England?" but "why not China?"

Pandaemoni writes:

It seems that the widespread access to education that books provided also was a spur that was necessary. Education make it possible for the huge numbers of scientists and technologists that we currently have, and the more you have, the more likely one will stumble upon some breakthrough. The better the means of disseminating the discovery, the better able other scientists and technologists are to make use of it, and build on it.

The ancients had many is not all the needed pieces, sophisticated machinery and machining knowledge (see, for example the Antikythera mechanism) and steam engines (see Heron of Alexandria), but the limits of ancient communication and education made it less likely those pieces were going to be improved and combined.

Add to that the importance of laws that allow one to seize the profits of one's inventions. In ancient times, the only way to capture surplus was by first mover advantage and that seems to have fostered secrecy rather than openness when military or commercial advancements were made.

WhiskeyJim writes:

I believe I agree with this post.

But is it not ironic that Bryan links to a paper by Jeffrey Sachs, who has just spent 5x income per capita in a number of cities throughout Africa, a grand experiment by the economist king so to speak.

And unfortunately for him, and the argument in your post, is that it is very difficult to measure the positive effect when compared to similar but untouched villages. But surely the return must be hugely significant, should it not? We should not have to argue about the significance of the regression, which is what he is doing to 'prove' success.

Creativity is anything but quantifiable. There must be a cultural component to advancement. Nature says that is a complex system which exhibits self-organization and autonomy, power laws and scale free networks, socially and economically and spiritually.

Societies that do not, will struggle and mimic, or even fail.

Comments for this entry have been closed
Return to top