Matt Yglesias’s China Syndrome

I think Yglesias’ point is (fairly) sound. In theory, under the classical economic paradigm, if there is a wedge between worker productivity in the agricultural and the industrial sectors, you expect workers to pour out of one sector into the other until their productivity aligns. In practice, this doesn’t happen. For example, in China labor mobility is extremely restricted (mostly in order to prevent urban areas from becoming centers of misery and political unrest).

My how things change. Just a few short decades ago the American left was profusely praising the PRC under Mao and its labor policies. Now American leftists are calling it a “repressive state” and other insults. If China had not broken with the USSR in the 1960s and had not introduced capitalist reforms in the 1980s I am guessing that no American leftists would be criticizing it today. But, in reality it is hard to see how Chinese workers are any worse off today than they were in the 1960s and 1970s when Maoism was really popular among the New Left in the US and Europe.

“My how things change. Just a few short decades ago the American left was profusely praising the PRC under Mao and its labor policies. ”

Wow. That is a very trolly first sentence my friend.

“So either there is large-scale mobility, and agricultural workers actually are competing for factory jobs, or it’s something else that’s keeping factory wages down. ”

Given that Chinese manufacturing wages have gone up by a factor of four since the year 2000, quite possibly the largest rise in recorded history anywhere, I’m not sure why anyone thinks wages have been kept down.

There are endless stories of the first areas to industrialise facing labour shortages. That’s why factories are moving west, Foxconn is installing robots and so on.

Since the post mentions Krugman, his essay “Ricardo’s Diffcult Idea” explains much of it. Or Marx of course. It’s the reserve army of the unemployed that keep wages down. Once that army is being employed then wages rise. As is happening in China.

Worstall: imagine what it would be like if they didn’t torch CCP offices.

Corey, I often mix them up, but I don’t think I’m doing so this time.

(Note, for the purposes of this analysis I’m assuming that “productivity growth” is a (mean) average. It appears to be that way in the linked article which is further linked to this paper Productivity growth is defined as “Labour productivity is calculated by dividing real expenditure-based GDP, obtained from BEA NIPA tables, by the number of
hours worked for all workers, including the self-employed, from unpublished BLS data.”)

Take 10 workers in 1970 in the XSA. All of them produce 10 units of production. That makes for a model GDP of 100 and a mean productivity (the ‘average’ productivity of workers in the XSA) of 10. Since they also have all the same productivity, the median productivity (the productivity of the average ‘most typical’ worker) is also 10.

Now lets fast forward to 2010 in the XSA. 8 workers still produce 10 units of production. 1 worker has been trained on a really cool machine. He, with the help of the machine produces 50 units of production. 1 worker has come up with a really cool method of providing important services over the internet to lots of people. He also produces 50 units of production. The GDP is now 180. Average (mean) per worker labor productivity is up 80%. Hooray. However MEDIAN worker productivity is up 0%, because the most typical worker is still only making 10 units of production.

Now please, for the love of God, understand that I’m not arguing that is actually what has happened in the economy. I’m just showing how the mean/median thing can be very important in these statistics. The quoted paper is super careful about the mean/median distinction for wages and compensation, but breezes right past it on the productivity question. I suspect that isn’t anything nefarious on their part, but rather the fact that such data would be really tough to get. But it is nevertheless important because it isn’t at all inconceivable that technology/machines might end up skewing things heavily (and which might bear strongly on the questions in nearby threads about predistribution vs redistribution).

Similarly on the lawyer question, I suspect there would be fewer law school applicants if they realized that the average lawyer (median) is barely going to make a white collar salary even though the average law salary (mean) looks perfectly respectable.

An important point about the US is that, relative to other developed countries both the pre-1970 and the post-1970 periods are anomalous.

US Workers did exceptionally well in the postwar settlement, to the point where unskilled factory workers earned middle-class wages. (Arguably, this predistribution muted support for redistribution).

Then since 1970 workers in the US have done much worse than anywhere else – in most places the labor share has declined only modestly if at all.

US Workers did exceptionally well in the postwar settlement, to the point where unskilled factory workers earned middle-class wages.

This is the only known successful recipe for creating the strong and vibrant middle class politicians are always talking about. Ever. In the history of any nation. The fact that politicians that talk about promoting the middle class rarely make a serious effort to recreate the actual golden age of the middle class seems pretty telling, to me.

Of course, any given politician is constrained by all the others he/she has to work with, but still, it seems like they could at least start by acknowledging that Reagan crushed the US middle class on purpose and no real progress can be made without first undoing his agenda. Or would that itself constitute jumping out the Overton window?

Related to what John Q said @18, the choice of 1973 as your starting point is very much a cherrypick. Wages were unsustainably and temporarily ABOVE measured productivity in 1973. If you don’t pick a cyclical high as your start the basic point about increasing returns to capital and high-wage labour remains but it is far less dramatic.

We get annoyed when sundry right-wing pundits do this sort of thing – we shouldn’t do it either.

derrida derider @22:

1973 is actually not “very much a cherrypick”–1973 was the first of the series of oil shocks in the 1970s. There’s a huge academic literature that argues that in 1973 the Fordist era of production changed into the modular post-Fordism that characterizes the industrial and labor worlds today. (The usual phrase is “on or about 1973.”) The choice here is pretty clearly reflective of those debates.

Point being, you can argue that 1973 isn’t the right choice for Corey’s point, but it’s not out of nowhere.

I really hate the whole thought paradigm, of a $1/hour agricultural worker dragging down the $2/hour manufacturing wage, as well as the whole vision of people pouring out of one sector and into another, driven by differences in pay. It is really not a very good analysis, and presumes a degree of substitution of labor for capital, which seldom obtains.

The actual capital that gets invested in applying technological possibilities and making high productivity possible is remarkably lumpy. There’s seldom an alternative high-labor-intensity method of making anything; the usual pattern, is complete domination by dedicated capital embedding an advanced technology. People are mostly employed in miscellaneous control functions, and not for an homogenous flow of labor services, which can vary neatly at the margin. There’s only one seat on a farm tractor, as Pareto observed long ago. There’s no farm tractor design, which is a labor intensive alternative.

A key foundation for the U.S. middle class of the 1950s and 1960s was one of the most successful industrial policies ever adopted — the second version of the agricultural adjustment act, put in place in 1935, which stabilized agricultural finance, in ways that enabled rapid investment of capital, setting off steady improvements in productivity in agriculture, coincident with the migration of labor off the farm. Economists, by and large, think AAA was a giant failure, when it was, in fact, a resounding, astonishing success. Classical analysis wants sectors to either attract resources or shed resources, but, in real life, the best result may require that a sector both attract resources and shed resources simultaneously, as U.S. agriculture did, from 1935 on, creating a foundation for rising middle class incomes, out of the widespread agricultural poverty of the 1920s, which had been a significant contributer to the Great Depression.

China, as is well-known, deliberately manipulated its currency and internal price structures, to make investment in high-tech manufacturing, low-risk and profitable. I don’t say that to criticize what they have done as somehow “unfair” — they wanted to build up a manufacturing sector, as a means to drive up production and incomes, and this is what they had to do. Economists often seem to think that such policy manipulation should be unnecessary, but such economists are idiots. The economic production processes that dominate international trade and deliver high incomes are precisely the ones with increasing returns, network externalities and so on. Without policy intervention, the “free” market, left to its own libertarian calculation, would leave China forever undeveloped and on the short-end of highly disadvantageous terms of trade, vis a vis the already developed countries. Every developing country, after the English pioneer, has had to engage in protectionist or industrial policies, to succeed.

Part of China’s policy scheme for manipulating the price matrix in ways that made rapid investment in manufacturing was using the domestic agricultural, service and state-owned sectors to make it possible for manufacturing workers to earn a pretty good living, on a wage, which was very low in international currency exchange terms, in part by keeping that domestic price structure low, and, also, by encouraging the Chinese to save fantastic shares of their incomes. Chinese consumer spending is way less than half of GDP even now, which is unsustainable. Years of sterilizing massive currency exchange rate interventions has exacerbated the institutional shortcomings of their banking system, in a country which is lousy with money savings.

Chinese agriculture was made to retain a large farm population. Agricultural productivity can be depressed by congestion effects — too many hands actually depress the productivity of a fixed amount of land and other agricultural resources. It’s a problem.

But, the bigger problem is that the whole price matrix will have to rise and Chinese consumers will have to spend some serious money, absorbing the country’s output. The factory worker will have to be paid more, and, also, his barber and his lawyer. It is only a matter of time, before the Chinese figure out that they don’t need to be subsidizing Wal-Mart, or Sony or Philips or Apple, and the margins they once conceded to make manufacturing investment a no-brainer, can be reversed in a shift in the terms of trade in their favor.

The shifting of gears, though, is likely to be political and economic chaos for several years. I would be very surprised, if there isn’t considerable political upheaval, if only because China desperately needs to institute a social welfare state and a more modern consumer banking system, to enable people to stop saving such massive amounts of money (and cushion the blow, as much of that accumulated savings evaporates in adjustments of the price matrix and reform of banking).

is there any reason to think that M Yglesisas actually knows anything about china, other then what he reads in the paper ?
I think most of us have college degrees, and can imagine a senior english paper – write a knowledable essay about something you know nothing about

James Fallows, OTOH, has lived and taught in the PRC for quite a while (various periods over the last 10 years), often in areas outside the direct all-seeing eye of Bejing. His reports are always prefaced, couched with exceptions and cautions, careful to distinguish between what is reported to the international press and international economics statistics agencies vs what appears to happen in actual daily life, and always end by saying “but China is a lot bigger, and things there are a lot more varied and complex than is possible to report – for good and bad”.

Cranky

Glen Tomkins: “Yglesias was raised in a society in which unions in particular, and demand economy considerations in general, had become truisms. Nixon himself, at the Moscow trade show in 1959, was famously the champion of demand and consumption against the Soviet Union’s monomaniacal fixation on production. Even the champions of the party that served the capitalists most assiduously used to be doctrinaire demand-siders.”

Actually, Matt graduated from Harvard several years ago; he’s probably just under 30.
That would mean that he was born in the early 80’s, and his awareness of things like this would have started in the 90’s, well after the crushing of most private sector unions, and the establishment of the political theme that we are a post-union country.

@9, @13, @14 — if they are intended to explain U.S. wage trends, these comments about median vs. mean all seem to be rooted in a mistake. It is a basic tenet of labor economics that unless there are barriers to labor mobility, wages follow the MEAN productivity of all workers in the labor pool and not the productivity of any particular worker in question. So it doesn’t matter if only one out of ten workers (the one who works on the cool machine) is more productive, the wage for everyone should still be pulled up. This is true unless there are barriers to competition or unless the ‘more productive’ worker truly has a unique skill that is impossible to imitate. (Not true if his producivity can be ascribed to the machine he works on).

The classic example of this is barbers. A barber/hairstylist is probably no more productive than he was in the middle ages. (The electric shear adds some productivity but scissors date back thousands of years). But he makes more money since he competes in a labor pool where he could choose to be an auto worker, computer worker, what have you.

I will add to the chorus saying 26 is a terrific comment.

The factors Bruce is pointing to about the lumpiness of capital complicates the straight neoclassical point I was getting at in 34 but does not completely eliminate it, you still have to think about how job market search works.