April 23, 2007
A Case of the Mondays: Jane Jacobs
Jane Jacobs attained a certain degree of fame after she published her iconoclastic approach to city planning in The Death and Life of American Cities in 1961. Subsequently she wrote two more books about cities that eclipsed The Death and Life in their level of iconoclasm; sadly, those books are still exceedingly off-mainstream. The first of the two, The Economy of Cities (1970), introduces the ideas that the basic unit of macroeconomics is not the nation but the city, and that economic development always begins in cities. The second, Cities and the Wealth of Nations (1984), takes the idea to the next level and talks about broader city regions, inter-city trade, the rise and fall of cities' economies, and cities in relation to nations.
Jacobs' ideas are sufficiently unknown that I am going to spend half this article just summarizing the two books. If you've read them, feel free to skip to the second half, though the first half might still give you a glimpse into how I mentally organize those points before I critique them.
The starting point of The Economy of Cities is that development comes from cities. It begins with a (weak) archeological argument that even the agricultural revolution was an urban renovation, with small trading posts functioning as cities in the preagricultural age. From there, Jacobs builds her points into a considerably stronger thesis that a certain level of population density is necessary to sustain economic development. It ranges from very difficult to impossible to put a factory on the ground without a network of urban suppliers. When Henry Ford tried producing every part of his automobiles at one site, he failed. Only when he changed his operation to assembling car parts produced in other factories, which formed part of Detroit's manufacturing network, did he succeed.
At the same time, Jacobs makes it clear that development can only happen in one way: import replacement. A city develops by having entrepreneurs and inventors take apart imported goods and learn how they work and how to produce parts for them until they can produce them more cheaply than they can be imported. The example she keeps referring to is Tokyo's bicycle industry, which replaced American imports with local production. Once a city replaces an import, it can use the extra money it gets to import other, typically more expensive goods, triggering further import replacement. This process is coupled to the full cycle of division of labor, in which division of labor involves adding new work, which in turn triggers more division of labor, and so on.
Finally, Jacobs warns, these processes can never work outside cities. Programs meant to develop agricultural countries by creating jobs in rural areas, most spectacularly the Great Leap Forward, invariably flop. This includes many softer programs for rural development, notably the green revolution and birth control. The green revolution's productivity increases displaced the rural poor without creating city jobs to compensate, Jacobs says. And birth control does not matter so much given that Japan, Western Europe, and the eastern US have high population densities without widespread poverty.
In Cities and the Wealth of Nations, Jacobs broadens her theory. First, she defines a city region, a penumbra of a city that appears to reach somewhat beyond its metropolitan area. Within that city region, the city's economic development spills over without wrecking society too much. This is based on five factors: the city's thirst for supplies of primary goods, job creation, productivity increases, transplants, and capital formation.
The most interesting factor, supplies, contrasts cities with supply regions—for example, oil-producing states—which are temporarily wealthy due to their richness of primary goods, and then crash once the goods run out or are replaced with alternatives. That, Jacobs says, is what places Saudi Arabia in the third world. It is economically passive, while the US is not.
More in general, she divides the world into productive and passive regions. Productive regions, i.e. vibrant cities, are in the first world; passive ones, including all rural areas as well as economically dead cities such as Pittsburgh, are in the third world, and only appear comfortable because of subsidies from Jacobs' first world. In that third world, stagflation, defined as high prices and not enough work, is endemic. She gives the example of Portugal, where unemployment is high and the prices, while low by American standards, are out of the population's reach. The problem, then, is that mainstream economists mistook the boom of the 19th century and much of the 20th century for a constant economic condition, not realizing that stagflation was perfectly normal.
Much of the rest of the book is devoted to fleshing out her earlier ideas more in full. She talks more about how productivity increases can hurt rural areas by making too many people redundant. She integrates trade into city development, showing how cities grow by not only replacing imports but also exporting goods. She continues her historical narrative, skipping from the agricultural revolution to medieval Europe; her main argument is that Venice developed by trading not only with Constantinople but also with other cities in then backward Europe, and in general cities should not become colonies to bigger cities by only trading with them but also create their own mini-networks of cities.
While she gives some examples of how cities can stab themselves in the back in The Economy of Cities, it is only in the later book that she develops that into a coherent idea, which, incidentally, is also where she is weakest. First, currency feedback is crucial in telling cities when to import and when to replace imports, so national currencies at best depress all cities but one—London in Britain, Paris in France, Milan in Italy, and so on—and at worst create a total disharmony of economic feedback, as in multi-city countries like the US. In developing countries, national currencies are pegged to rural goods or primary supplies, which tend to strengthen the currency beyond what the cities can take without deindustrializing. One major reason Singapore developed so fast is that it was kicked out of Malaysia for political reasons and subsequently used its own currency. The US and Japan needed explicit tariffs to protect local industry; in Singapore (and Hong Kong), this tariff was called the national currency.
Further, cities can stifle themselves by engaging in various forms of discrimination, including against small businesses. Caste systems and racial and gender inequality rob the city of needed talent. Regulations such as ground rules established in New York in the 1960s, wherein the city let firms bid on the redevelopment of 37 buildings in Harlem but only if they could bid on all 37 at once, ensure blighted neighborhoods cannot develop their own talent. In the US it is so egregious that in The Economy of Cities, Jacobs quotes a civil rights activist who says government interference with slum development causes so many problems it would be better if it left black neighborhoods alone entirely.
In Cities and the Wealth of Nations, Jacobs also introduces the idea of transactions of decline. These are forms of spending that on paper increase GDP but in practice never produce any innovation or further growth. She identifies three such transactions: military spending, subsidies to rural hinterlands, and trade with backward countries. Military spending can help the economy when it is temporary, but otherwise it is a drag since it produces nothing. The same applies to subsidies to the hinterland and loans to third world countries that can never pay them back. All three are preoccupations of great empires, which is why cities tend to have cycles of growth, followed by imperialism, followed by decline.
Before criticizing the specifics of Jacobs' argument, let me say that I think the basic notion that cities are the basic units of macroeconomics makes some sense. There is no special reason for nations to be the basic unit, especially not in the age of the European Union and the Euro. Urban areas and city regions are natural units defined economically and socially, independent of arbitrary political boundaries. Nations—and, incidentally, city limits—are not. In addition, let me note that there are many sub-issues I cannot address for space constraints. My above summary has 1,100 words; Mark Rosenfelder's has more than 6,000 and still misses some important points.
The weakest point in Jacobs' argument is the exact definition of import replacement and when it occurs. She peppers her writing with examples of when cities replace imports and when they do not, without a shred of evidence that this is in fact what happens. It is subtle and remote enough in the two books I am dealing with here, which is why I only noticed it upon reading Dark Age Ahead (2004). In that book, she talks about Canada's rapid economic growth in the early 2000s as an example of import replacement in Toronto. The sum total of the evidence she includes there is an anecdote of an office chair with "Made in Canada" printed on it. The evidence she gives that some city in the US or Britain underwent a surge of import replacement in the 1840s is even thinner.
In fact, her example of Toronto's import replacement shows how fragile her analysis is. One of the major factors behind Canada's recent growth is the fact that Alberta is sitting on more oil than is present in the entire Middle East, albeit in tar sand form, which is more expensive to produce than Saudi crude. Increasing development of tar sands is causing labor shortages in much of Alberta, which then translate to reduced unemployment in other provinces, which send migrant workers to tar sand mining operations.
Second, the three transactions of decline she identifies are not the only or even the most costly types of spending that do not produce wealth. Health spending, debt interest, infrastructure repair, policing and internal security, and even some forms of welfare that go to the urban poor are just as economically unproductive. Much of this is covered by the difference between gross and net domestic product. The rest boils down to how much the city spends versus what level it could theoretically lower its spending to, which is itself a function of its wage level, or its existing level of economic development.
In fact, health spending dwarfs Jacobs' three transactions of decline in almost every developed country. The US spends 4.5% of its GDP on defense, and New York's tax imbalance with the state and federal governments, which significantly overlaps military spending, totals 5% of its gross city product. Aid to other countries, including loans, totals 0.2%. That compares with 15% of American GDP spent on health care. Although other developed countries spend closer to 9-10% of their GDP on health, they also spend closer to 1.5-2% of their GDP on the military. The only developed country that spends more on the military than on health is Singapore, which has no hinterland to subsidize (though Israel comes close, and also massively subsidizes settlers above and beyond IDF protection).
There are a few more areas in which Jacobs' theory is fuzzy. It says nothing of how subsidies to poor regions can in fact produce innovation by investing in education. It entirely misses the fact that population pressure can impoverish countries, and at any rate birth control and family planning are necessary to move women from the production of babies to the production of new wealth. It dwells on manufacturing but says nothing about service economies. It is overall tailored to the 1980s, when the shock of American deindustrialization was at its peak, Germany and Japan were forward-looking innovators, China had barely recovered from Maoism, and Ronald Reagan was busy hiking military spending and running unsustainable deficits.
Two points of fuzziness stand out. The first is that Jacobs leaves cities undefined. It is implied that they all work like London or New York, that is have a core surrounded by rings of decreasingly urbanized areas. But that is not the only way for a city to arise. The biggest urban area in Germany is not Berlin, but the Ruhr, an agglomeration of many relatively small industrial cities, none of which dominates the region. On a larger scale, Jacobs leaves out megalopolises, which severely complicate her proposed scheme wherein each city region mints and prints its own currency. How can New York and Philadelphia have separate currencies when their metropolitan areas overlap?
The second fuzzy point is the definition of the third world. The third world is not defined by economic passivity, but by various social problems centered around poverty. Delhi, Kolkata, Mumbai, Beijing, and Bangkok are perfectly dynamic, and fit perfectly into the third world. Moscow has gotten far more economically active since the fall of the Soviet Union, but its upsurge in poverty and breakdown in public health triggered a painful process of third-worldening spreading in Russia.
To some extent, it is hard to fault Jacobs for consistently preferring anecdotes to data. When the economic mainstream focuses on nations, it is very hard to find accurate data about the economies of cities. Jacobs is by and large forced to talk about import replacement in the almost magical terms she uses, invoking it whenever there is no other explanation for economic growth that fits her theory. And the historical overviews that stay away from handwaved import replacement are strong.
But the solution to problems with data is to look for empirical clues, such as the number of bicycles or cars or computers a developing country imports every year. Import replacement will occur whenever we see a decrease in the imports of a lower-level good without a corresponding decrease in its consumption. Supply-oriented growth will occur whenever increased exports of primary goods account for big enough a fraction of economic growth.
Jacobs' policy suggestions span the entire gamut from politically insane to extremely cogent. It is not especially hard to divert subsidies to areas where they increase productivity: education, worker retraining, public transportation instead of roads and cars, direct scientific research rather than military research, minimum income as opposed to a mishmash of welfare programs that cost too much and reduce poverty too little. There, Jacobs is completely right. Her suggestion that budding empires not squelch city development in colonies the way Britain did in Ireland and tried to in the US could work, but is politically difficult to implement. In contrast, small currency regions will never work, and neither will ending trade with backward areas.
Make no mistake about it: Jacobs understands macroeconomics. Her theory has a fairly sound core, even if it requires tweaking to account for changes brought in the previous decade and in this one. The problems only start when she heaps onto the theory sundry sub-issues that only detract from it.
Posted by Alon Levy at 06:20 AM | Permalink