Pritchett’s Postulates and Urbanization

After promoting women’s groups in West Bengal as a route to development a West Bengali woman asked Lant Pritchett:

You all are from countries that are much richer and doing much better than our country so your country’s women’s self-help groups must also be much better, tell us how women’s self-help groups work in your country.

Pritchett’s inability to answer the question led him to what I call Pritchett’s postulates of development, four criteria to decide whether factor X is an important determinant of development.

  1. More developed countries must have more X than less developed countries.
  2. The developed countries must have more X than when they were less developed.
  3. Recent development successes must have more X than development failures.
  4. Countries that are developing rapidly must have more rapid growth of X than those that are developing slowly.

Since more developed countries don’t have noticeably more women’s self-help groups, this idea fails Pritchett’s postulates. Indeed, so do many fashionable development ideas being tested by RCTs which is one reason why Pritchett’s postulates are controversial in the development community.

Paul Romer, however, (whose important blog post led me to Pritchett’s postulates) has a different approach. Instead of dismissing ideas that fail the Pritchett postulates let’s look for ideas that pass them.

Romer provides evidence that urbanization passes all of Pritchett’s postulates. I think he is correct and that suggests that policies to increase the rate of urbanization could have a very big payoff for development.

We are used to thinking about urbanization as a consequence of development but it is surely also a cause. Consider, for example, the micro evidence. It’s not that rich people move to cities, it’s poor people who move to cities to become rich. We also know that cities are engines of innovation.

We can have too much urbanization or too much in one place as when we get a bloated capital city. Nevertheless, it seems that we could speed the rate of urbanization by reducing the cost of urban development – both the obvious costs like improving land allocation in say India but also improving sanitation and air quality in order to lower the health costs of urbanization. Similarly, well planned, efficient, even beautiful cities increase the benefits of urbanization. Urbanization policy in general becomes growth policy.

How else can we increase the rate of urbanization in developing countries?


"Since more developed countries don’t have noticeably more women’s self-help groups"

They don't? Have you never been to a "Lean In" book club?

Seems like a lot of theorizing on account of not being personally familiar with the facts.

My mother called her self-help groups where ladies got together to give each other advice "bridge" and "bowling."

Do you know what factors X and X' fit Pritchett's Postulates best?

Fraction of country's population derived from pre-Industrial-Revolution Western European and Eastern Asian forebears.

Slightly less predictive but very powerful: fraction of country's population with high intelligence (g, IQ).

Some other factors which test well: low rate of inbreeding (endogamy, cousin marriage); chiefly monogamous marriage pattern; and economic/social system organized around widespread production rather than tournament for control of key natural resources or central despotic power (central Asian model, where "key natural resource" is just 90% of population held in bondage).

Shhhhh ...

Urbanization seems to do a pretty good job of reducing all of those things. People in Mumbai are much more sophisticated than people in Rural India. I think we all agree that's a good thing. What's your point?

Cargo Cult logic in both cases. There isn't necessarily just one path to development, and rapid urbanization without infrastructure keeping up can be counter-productive.

@FM - right you are. And also these Postulates ignore than sometimes being first is best for historic reasons, but won't work the second or third time. For example, both the USA and Australia got rich from having farmers or convicts enter a virgin territory and kill off the natives. But if China tries that with say Tibet today, they are likely to fail. Sometimes it's just luck that makes a country strong.

Infrastructure *is* urbanization. If you have a dozen villages each spaced a few miles apart and you build expressways connecting them together, then you've created a city in economic terms. People and goods from one village can travel to the other villages to exchange goods and services: that's what a city is.

An extreme example is Silicon Valley. There's no single city which dominates the Valley and it's mostly low-density housing. But thanks to lots of fast wide roads, workers can commute from one end to the other. (Ok the traffic is hardly smooth-flowing at rush hour, but compared to Manila or Jakarta it's pretty amazing.)

Well I see a fly in the ointment. If urbanization is so great, then greater Manila, with 30M+ people, or greater Tokyo with 40M, with the same (roughly 1/3rd of the population), or NYC, or London, or greater Karachi (32M to 60M out of 200M would all equally be powerhouses. But they are not, only the first world countries are. In fact the only way a city can be a powerhouse is if it is linked to a dynamic country. Same with Indian big cities, none compare to Shanghai or Beijing, despite being about as big. Now you can try the 'fudge factor' of saying (as AlexT hints) that it's because these non-powerhouse cities are 'too much in one place' but this is just an artifice. How do you explain the success of Beijing and Shanghai (both crowded messes) and Tokyo (id) compared to the un-successes of big cities like Manila and Karachi and in India? Clearly then cites are parasitic to the host country: when the host is a powerhouse, then cities are powerhouses too, if not, not. So without a dynamic USA then Silicon Valley, whether 'spread out' or concentrated, would be a failure, regardless of infrastructure. Chalk up another bogus development remedy dreamed up by humanities majors masquerading as scientists (aka 'economists').

So, Singapore is a powerhouse city because it is tied to a dynamic country?? Which 'one' would that be? As far as Karachi is concerned, it is the country that is the parasite.

Mumbai, Manilla and Karachi are all success stories relative to the rest of India, Phillipines and Afghanistan. Pakistan and Afghanistan are essentially the same people and have similar cultures, but the former is much more prosperous because it is more urbanized than the latter. Annual incomes in Mumbai are three times as high as in the rest of India, including for the poor. Yes, there are decreasing returns to scale and India is better of developing dozens of small cities rather than herding everyone into Mumbai. This is obvious. But what is also obvious is that people become wealthier as cities grow.

Ray you are stupid. You are being disagreeable for the sake of being disagreeable. Nobody said urbanization would make cities equally prosperous or equally successful. You idiotically use the word "powerhouse" five times even though that term is never used by Paul Romer or Alex Tabbarrok.

Romer provides evidence that urbanization passes all of Pritchett’s postulates. I think he is correct and that suggests that policies to increase the rate of urbanization could have a very big payoff for development.

Yeah; does anyone think a Soviet model of forced urbanization ("just build the cheapest, quickest, biggest city and make everyone move there") would be good or create actual growth?

I submit that urbanization is an indicator of the things that cause development, not a first-level cause in itself.

(See "owning a house" and "being middle class" for a parallel.)

The poor logic is on your part. Rapid urbanization without infrastructure may or may not be unproductive, but it has been demonstrated that urbanization is almost universally crucial for a country's development.

Now, the appropriate response isn't that "rapid urbanization without infrastructure keeping up can be counter-productive." Everyone freaking knows that. You added zero knowledge with that point. The appropriate response is how can we mold policy and institutions (whether through free markets or through government control) to promote infrastructure growth that can sustain urbanization.

The "there isn't just one path" to development is such a worthless statement as well. Nobody says there was only one. But we do know that development comes more rapidly with urbanization than without. So urbanization is a policy worth pursuing, and we should consider what is the best way of going about it. You and Ray both provided useless commentary.

Are you being deliberately dense, or just a natural? I will explain slowly.

First: urbanization vs. development

Obviously development is going to be accompanied by urbanization. If everyone stayed in the countryside, you would have an economy based on high-labor/low-productivity agriculture, which is not a path to development (I suppose people could stay in the countryside and focus on labor-intensive high-value crops like asparagus or saffron, but that's not widely applicable). The question is whether urbanization is a consequence or a cause of development. Places like Karachi, Mexico City, Lagos or Cairo would seem to disprove a causal link between urbanization alone and development, something else is needed, whether infrastructure or institutions (see the difference between Mumbai and Karachi).

The South Pacific cargo cults drew the inference that towers with people wearing headphone-like gear on their ears (US military air traffic controllers) caused planes to appear from the blue, carrying goodies. When the US Army left, they built straw towers and manned them with islanders wearing straw headphones to conjure up the goodies again, and failed. Leaping from correlation to causation in urbanization vs. development is another instance of magical thinking.

Second: the postulates

Telephone landlines satisfy all the requirements for X. Does that mean developing countries should squander capital on building a copper landline infrastructure? Of course not, they are all leapfrogging directly to wireless networks. Thus the path to development in the 19-20th Century West involved landlines, whatever path developing countries taket in the 21st will be different in that respect at least. This simple example shows the uselessness of the alleged postulates as a guide for evaluating policy.

Now the question is not whether urbanization should be encouraged, but what kind, with what infrastructure planning, culture and institutions, so the path is closer to Singapore or Shanghai than Lagos or Karachi. Unfortunately, the method and postulates in the article, in addition to being demonstrably false, are useless at explaining what is required and what policies are required. Some candidates:
- stable political institutions (not necessarily just or corruption-free)
- law and order
- adequate, not gold-plated transportation, electric, water, sanitation, medical and educational infrastructure
- good transportation and communications links to the outside world

Most underdeveloped countries have policies in place that have the net effect of destroying the ability of rural workers to move to urban areas or to prevent entrepreneurs and businesses -- especially foreign businesses -- from developing rural land and turning it into productive urban/commercial property. Inefficient agriculture is often subsidized and protected at the same time its prices are depressed (thus hurting farmers) to help the cities that do exist. Various other complicated tax and regulatory structures both lead to the inefficient mega cities and mega slums with low performing rural areas.

Are you sure it is possible to have a "well planned" city that also "[reduces] the cost of urban development"? The most carefully planned cities may also be the most expensive to live in while unplanned cities the greatest engines of mobility.

That depends on how you define well-planned.

In California one of the most affordable cities near the coast is Irvine which was entirely planned by a real-estate company to maximize property values. It's clean, safe, business-friendly and preserved a lot of open spaces around the city for healthy recreation.

Los Angeles on the other hand. . .Well, they have a very active and powerful planning department- but not one that considers costs as a negative (more expensive permits and projects = more campaign contributions from unions and construction companies).

In Alan Riding's 1986 bestseller about Mexico, Distant Neighbors, it's assumed that Mexico City will grow to 30 million people. But that didn't happen. Instead, Mexican peasants moved to the United States.

"How else can we increase the rate of urbanization in developing countries?"

Two things (USA and Europe can do): Close the borders and for the illegals that previously absconded into: absolutely no free stuff of any kind. When Pancho slams Pedro in the head with a lead pipe or Khalid shoots Barack in the ass: don't call an ambulance, put them on a bus to the border.

The greater Mexico City metro area is well over 20 million, if not all the way to 30 million.

The other reason "it didn't happen" (yet) is that Mexico's fertility rate declined pretty fast and now approaches replacement.

Yeah, Steve. Wise up.

When Barkley Rosser does the maths, 20 million = 30 million.

Gosh, mbcb, are you trying to make up for your complete flop in the Obamacare discussion? Did you pick up that by saying "maths" you might look smart? Sorry, that simply makes you look British.

As it is, JonFraz is very relevant: fertility rate in Mexico has fallen sharply, and the growth rate of Mexico City declined after 1980. Sailer is at best only partly right, and not at all for recent years.

For the record, current population of Greater Mexico City is 21.2 million. I grant that this might not be "well over 20 million," but, mmcb, you decided that I said it was 20 million, exactly. When are you going to get that dental appointment to get your fillings replaced, or maybe you just need a brain transplant. Getting one from a gorilla would probably constitute and improvement.

You didn't do so well in the Obamacare thread either, Barkley.


Let history judge!!!

I feel comfortable with my share of the Obamacare discussion, thank you. I am not sure your own contributions will hold up so well in the cold light of morning. Once you sober up, I suspect you may agree with me but perhaps you never will.


Please get with the program, you are only partly right. At best!!!

Barkley's analysis demonstrates conclusively that 21.2 million = 30 million and that his critics need to have their filling replaced. Or something. You can't argue with analysis like that, you can't duplicate it, you can only shake your head in awe.

By the way, Steve, I can't believe you haven't already apologized to Barkley for making him look up a fact. Barkley is too busy for that kind of foolishness. Do you want him to start calling you names? Does he need to unveil his analysis of your dental issues?

Don't worry, he's a professor at JMU so no one is getting educated whether he's there or not. If Barkley Rosser is taking 20 million stiffies, but his wife is subsidizing his stiffie intake by 10%, how many stiffies will he have to take when his wife's cooch dries up? Barkley Rosser maths says 20 million, because increasing the percentage of something does now actually increase the number!

Wow, you guys who fell flat on your faces in the ACA discussion are really desperate to prove something or get a dig in one way or another, aren't you all?

Here is something all of you should deal with. On that one, the facts simply are not on your side, but you all just want to keep on keeping on with the same old propagandistic drivel. At some point you simply have to face that the world is not flat, but I know that it is really really really hared to do, so we get these sorts of infantile displays out of you all. Maybe you all can get into Holocaust denial as an upgrade on the quality of your discussion.

You're just stuck in the echo chamber of academia and have no interest in freedom. Let me guess: dad was a union man, grandparents came from southern or eastern Europe, mom was a 50's house wife with a fat clit big titties. You read the cliff notes to "Das Kapital", tried to trudge through the New Yorker weekly, and found a home with the wannabe hippies/fascists of your day in Wisconsin. You couldn't decide on a career because you're stupid and mommy encouraged you to follow your dreams, so you ended up a third tier professor at a third tier school, fawning over pseudo intellectuals who preached your message and never thinking about anything besides how wide you can spread those cheeks. Did I get it right professor worthless? Oh, and you love the French because they are " so sophisticated and advanced."

Wow. I'm not sure if that's the most offensive blog comment ever or the best Lou Reed song that never was.

The greater Mexico City metro area is well over 20 million, if not all the way to 30 million.

Mexico City population in 1986, when the book was written: 8.5 million. Mexico City population now: 8.8 million.

that is one badly failed prediction. Stop digging just because Sailer wrote something.

How about bring your total fertility rate below 3.0? That seems to be a rule in the developed world in the 21st Century (Israel of course being the exception).

But that seems more like effect than cause. Once it becomes more cost-effective to invest heavily in few children than to invest broadly in many children, fertility rates drop naturally.

I don't think most people are making cost/benefit decisions about the ROI of children.

The decline in fertility for developed countries is probably due to a number of factors - more years of education for women, women entering the workforce, availability and social acceptability of birth control... I doubt you can draw an easy cause & effect conclusion.

Steve, urbanization is a pretty effective way of reducing fertility. You've noted it on at least two hundred occasions about New York and California. Are you forgetting to Notice Things?

Someguy - all these things are effects of urbanization.

Your ideological predecessors already tried this - urbanization was the key to development theory for a couple of decades - and it failed. But, the new generation is always free to use the same logic to repeat mistakes.

Planned/accelerated urbanization seems to be working extremely well in China.

Seems being operative - there's excellent reason to believe much Chinese "progress" is hollow Potemkin growth.

(And much of it isn't, of course - but I'd bet a dollar the real growth rate of Chinese wealth is far lower than the Government of China would have us believe.)


The balance of evidence, and my own experiences talking with Chinese policymakers, suggests that, on net, Chinese growth has been under rather than overstated.

If things are looking up in China, why are millions of people willing to spend over $100,000 on a degree from an Anglosphere university raises the probability of getting residency?

*which raises...

Self contradictory question.

If things are NOT looking up in China, where do millions of people get the $100,000 ...

Not to go all phenomenologist on this, but that's looking backward. I'm wondering why people are making credible commitments to a life of what could reasonably be called self-imposed exile,

Runaway urbanization of the capital city was a well-known problem in the 3rd World a generation ago: e.g., Mexico City, Lagos, Manila, Calcutta, Jakarta, Karachi, Dhaka, etc.

Neither Calcutta nor Karachi is a national capital city, although Calcutta was the capital of the British Raj for a period of time.

Pshaww. Why are you bothering Steve Sailer with facts?

Facts, irrelevant to the discussion, but facts nonetheless.

Calcutta is the capital of the west bengal state, which has a population of atleast 90 million which would make it around the size of the phillipines or bigger than Vietnam or Germany. Steve's point holds. For apples to apples comparison many Indian states can be compared to countries.

England used child labor as part of its industrialization strategy. Therefore India should use child labor as part of its industrialization strategy.

England had an industrialisation strategy?

These sound like necessary conditions, but maybe they aren't also sufficient conditions for development? Also, identification seems to be an issue: Is urbanization the key, or is there some factor that's driving both urbanization and development? Maybe it's industrialization. Build some factories (under the right conditions,) and the cities will come of their own accord along with other development.

It's a useful way to look at things, but you need to look very carefully and not get distracted by apparent silver bullets. Maybe a literature search is also a good idea? It seems kind of obvious that someone else might have tried something similar already.

Sanitation passes Pritchett's postulates handily, and as you observe above, sanitation is important to foment urbanization. I can only imagine that sanitation improvements would, by themselves, pay off and make cities more attractive.

Actually, the sanitation is a clue that this "urbanization" argument and probably these postulates are largely nonsense of the cargo cult variety.


Across human populations urbanization occurs naturally in the sense that cities are typically a net drain on total population. The level of urbanization at any point in time is the result of factors besides urbanization policy itself.

Nope, you're wrong, sanitation actually doesn't don't do anything for development.

See bottom of page 5.

argh. Leave out "don't" in that comment.

The obvious response to the paper Hanson is citing is "endogeneity". In this case, it is likely that the cities with the most advanced sewage systems are responding to...well sewage. All it takes is that and a lag in the effect of sewage on development indicators xyz for no effect to appear in the cross-section.

I'd like to read more about the Lee, Rosenzweig, and Pitt paper before I could comment further, but I'm not going to just buy the article. But the possibility of a reduced effect from sanitation is noted.

The physician John Snow was able to pretty convincingly demonstrate in 19th century London that there was a link between sewage getting into the water supply and cholera deaths. Snow's research was a major breakthrough for its time and it is surprising Hanson does not mention it.

"Similarly, well planned, efficient, even beautiful cities increase the benefits of urbanization."

The U.S. has become less rural in the last century but also less urban in the sense that core cities are less densely populated than they used to be. Manhattan, for example, is about a third less populated than it was 100 years ago. And this is not a peculiarly American phenomenon -- the city of Paris has shrunk by a similar proportion over the same period (and now accounts for only about 2 of the 12 million living in the metro area). While the core cities have shrunk, the surrounding suburbs and exurbs and grown dramatically. So wouldn't it be more accurate to say that in the most developed countries, what is associated with modern growth is not 'urbanization' but rather with 'suburbanization'? Suburbanization creates fragmented local governments that must compete for both businesses and residents with clear benefits for government efficiency--the threat of exit is much more powerful when all you have to do to escape is move a few miles across the border to an adjoining suburb.

The US has something like 60 cities with over 1 million in the metro area. Maybe 100 with over 750,000. For some reason, every commentator thinks that urbanization means the creation of Mumbais. What the Chinese government (and the Indian government) has is actually doing is promote the growth of second and third tier cities - the Kansas City's and Birminghams of China and India rather than the Chicago's.

Suburbinization reflects better transportation, greater demand for space, better communication etc. All suburbs are surrounded around a city (even Silicon Valley). You need to have a city to start with. That city does not have to be 20MM strong. India would be a lot better off with a bunch of Gandhinagars rather than building Mumbais. Gandahinagars are still examples of urbanization.

Am I missing something, or does this pass anything that correlates with development? So, like, X=Playstations passes the criteria?

Is "must" supposed to carry a lot of weight? This is a genuine question. Is someone supposed to have figured out causality in addition to conducting this analysis? Is it ever possible that something is required in the development phase but not later on? For example, and I'm totally making this up, what if nations had to pass through a period of harsh rule of law on the way to development but that once developed you could relax that considerably?

Getting back to the original issue raised to Pritchett, womens' self-help groups are hard to measure, as Steve Sailer pointed out. However, one X factor that fully fits regarding women is the level of education of women, with the percent having secondary education. This is strongly associated with lower fertility rates and also with urbanization, with urbanization also being independently linked to lower fertility rates.


The postulates ignore what X is supposed to accomplish. If the goals are achieved otherwise then who needs X?

Take the self-help groups. I'm not sure what they are expected to accomplish, but assume it is something along the lines of increasing women's independence, economic power, and general participation in political, economic, and social matters. If so, then what we need to look for in advanced societies is not the self-help groups themselves, but these hoped-for benefits.

Yup. When people start riding a bicycle, they usually use training wheels. But a good rider doesn't add additional training wheels. She takes them off.

My mother participated in many women's self-help groups where much of the the socializing consisted of the ladies giving advice and support to each other, but they typically had nominal goals of either recreation (bridge, bowling) or charity work. And the women typically got some benefit in social standing by being asked to join. A women's self-helf group organized around the premise You Need Help would not have been very appealing.

Attending Church itself sounds like a women's self help group. 100 years ago in USA that might have been the biggest women's self help community.

There is some good work by an economist at your fellow school of George Washington, Remi Jedwab who works on the question "Why has Africa's urbanization not resulted in growth". Many African countries had urbanization rates on par with East/South Asia, but have not had similar levels of growth. A good example paper is here: The primary takeaway is that urbanization increases from *migration* are pro growth, as they are focused on employment, innovation, etc., while urbanization from increased fertility in cities is generally growth reducing (perhaps its a proxy for high levels of patronage/aid/welfare dependence?)

"high levels of patronage/aid/welfare dependence"

tough to see this is most African cities

Yeah, and thus the problem; if you adopt urbanization as a policy goal, and use aid to do it (like we sortof have in Africa as an inevitable consequences of how NGO's work), then you get the outcome without the benefits.

Waugh's "Scoop" (1938) discusses in passing how a fictionalized Addis-Ababa rapidly grew in size:

"It had been found expedient to merge the functions of national defence and inland revenue in an office then held in the capable hands of General Gollancz Jackson: his forces were in two main companies, the Ishmaelite Mule Taxgathering Force and the Rifle Excisemen with a small Artillery Death Duties Corps for use against the heirs of powerful noblemen. ... Towards the end of each financial year the General's flying columns would lumber out into the surrounding country on the heels of the fugitive population and returned in time for budget day laden with the spoils of the less nimble ...

"Under this liberal and progressive regime, the Republic may be said, in some way, to have prospered. It is true that the capital city of Jacksonburg became unduly large, its alleys and cabins thronged with landless men of native and alien blood, while the country immediately surrounding it became depopulated, so that General Gollancz Jackson was obliged to start earlier and march further in search of the taxes ..."

Throughout the history of civilization there's always been a flood of rural migrants into cities: the countryside was healthier than the city (less danger from crowd plagues and safer water supplies) so more people survived childhood-- only to find that they aren't making any more land so there was no place for them in the country if they were not an older child. And cities survived only because of this steady migration: urban death rates exceeded urban birth rates pretty much everywhere right into the 19th century. Yet until the 1700s (with a few spurts earlier, e.g., the Renaissance and the Hellenistic era) there was very little growth from all this migration.

Transportation infrastructure passes all four tests and also supports urbanization. Erie Canal multiplies economic importance of NYC, creates Buffalo. Railroads create Chicago, Atlanta.

Also, services? Haven't most nations gotten wealthier by adding services to their output?

Other X factors might be, nutrition, education.

How to make a country increase urbanization? It appears China still has the plan. Enforce a one baby per woman policy and push all the young poor woman into the cities to work for nothing all the while increase the size of current farms.

These are not such great clues.

Bed nets or deworming tablets might stop people from contracting malaria or intestinal parasites in the developing world, but at some point in development they stop being so necessary because the underlying problem is solved by other means (draining swamps and better sanitation respectively). They could be valuable in the poorest places, but will fail 1 and 2 looking at the richest countries.

Arguably women’s self-help groups are useful in a place with the most sexist traditional values, but no longer useful (or at least the most effective approach) in more cosmopolitan countries.

Macroeconomist that doesn't do development, but is up on the growth literature here ...

And what I've been thinking a lot about lately is the breakdown in the urban systems in the western parts of the Roman Empire from the 3rd century onwards.

This dovetails nicely with Romer's insight.

Maybe there's a paper a few years down the road about whether the growth of the villa system was a cause or a result of declining urbanization, and whether the decline of Rome in the west is even more of an economic growth story than we recognize.

Another, not exclusive, possibility is a strategic retreat by aristocrats who decided it was better to rule in their latifundia than risk death or exile in the coups and civil wars.

Does the concept of "self-help groups" strike anyone else as contradictory?

So we can go full Soviet on this: The developed countries had more steel mills. They had more steel mills than they did when they weren't developed. Therefore, we communists need to build more steel mills.

Look how far that got them. The whole thing confuses cause & correlation.

Are you seriously suggesting that the Soviet strategy failed to industrialize the country thereby generally raising living standards?
I doubt you'd find any serious study agreeing with this.

Also greater steel production is strongly correlated with industrialization - see England, US, Germany, Japan etc

"I think he is correct and that suggests that policies to increase the rate of urbanization could have a very big payoff for development.

We are used to thinking about urbanization as a consequence of development but it is surely also a cause... It’s not that rich people move to cities, it’s poor people who move to cities to become rich... Urbanization policy in general becomes growth policy."

I wonder what the farmers forcibly relocated to Chongqing would say about the idea of policies encouraging people to move into cities before some minimum level of growth in agricultural productivity has occurred, because "urbanization policy in general becomes growth policy"...

Basically the postulates just boil down to "is there a correlation between actual development and your clever idea for improving development". Now, correlation is not causation, but it's still useful to look at. But causation without correlation is quite rare, while correlation without causation is fairly common. (Essentially you'd need an unrelated countervailing trend to get causation without correlation, whereas there are related causal structures that give correlation without causation.) This means that false negatives, where the postulates say an idea won't help but it will, will be rare, while false positives, where the postulates say an idea will help but it won't, will be more common. So I think it's correct to just use the postulates to filter ideas, and not to generate them as you're doing.

I decided to do a Pritchett-Romer “smell test” on education, just to see how it all works out. I was pretty pleased with the results. And, it looks like education passes the smell test.

Stop helping rural farmers in under developed countries. When i worked in Honduras at the institute of agriculture one of the Ag PHDs (from the university of Kansas) working there told me that they had grain subsidies to keep the farmers from moving into the city.

Comments for this post are closed