China overtakes US as world's biggest CO2 emitter

[The Magnificent Goldberg]

It's very encouraging if you are saying that countries whose economies are service-dominated can do well. It would imply that America has a future. Right now, here in the Midwest, all we see is economic decline owing to our industrial decline.

Yes, but you have to watch WHERE in the country the different economic growth occurs. I'm not at all familiar with the regional scene in the US. For illustrative purposes, here's the regional GVA figures for Britain.

http://www.statistics.gov.uk/pdfdir/gva1206.pdf

About half way through the document, there's a table setting out the changes in the period 1995-2005, when Britain generall was doing reasonably well. But you'll see that most of Britain was in decline. London, the South East, The East and the South West were increasing their share of GVA at the expense of the former manufacturing areas.

MG

[Guy Berger]

1) Since agriculture and services can be important sectors of economy, "industrial strength" and "economic strength" are not the same thing.

While you are correct to a degree, I don't think it's as big of a degree as you think. The British thought "services" could replace manufacturing, as did other major economic powers of the past. When they reached that stage, they were already fading. As are we.

To dredge up this topic, there was an interesting article in the WSJ yesterday. (Quoted below.)

I'm not an expert on productivity but IMHO the idea that in the United States services have low productivity growth and manufacturing have high productivity growth is either BS, or irrelevant.

If there really is a big gap in productivity growth between the two sectors, why aren't resources flowing from services to manufacturing?

Is Baumol's disease back?

Named for economist William Baumol, the theory argues that the labor-intensive nature of some services acts as a constraint on productivity growth in an economy that increasingly produces services. It's relevant again years after some economists pronounced it "cured."

Between the mid-1970s and mid-1990s, annual productivity growth in the U.S. nonfarm business sector averaged about 1.5%. In the past decade, it has averaged about 2.6%. For a couple of years in the early 2000s it was near 4%, and Mr. Baumol's idea looked destined to join others in economics that looked good in theory but wrong in practice.

Yet last week's downward revisions of U.S. productivity growth for the past three years suggest that the trend is closer to 2%, and shows that productivity growth has slowed for four straight years.

At the same time, employment in traditionally less-productive sectors such as health care and leisure is growing rapidly, while employment in higher-productivity business services is growing more slowly; in manufacturing and retail trade, it is flat or shrinking. Therein may lie clues into whether the recent dip in productivity growth is a major turn or a temporary lull.

That's where Baumol's disease comes in.

In the 1960s, Mr. Baumol, now at New York University, and William G. Bowen, an economist who later became president of Princeton University, argued that because productivity growth in labor-intensive service industries lags behind that in manufacturing, productivity growth in service-oriented economies tends to sag.

Their famous example was a classical string quartet -- there are always four players in a quartet and it always takes about the same amount of time to perform a set piece of music. You can't get any more music out of the same number of musicians over that same period of time. Broadening that to other types of services, the implication is that rich countries such as the U.S. that tend to veer toward services would face higher prices as wages and costs rise.

But something happened in the last decade.

The information-technology boom led to rapid efficiency gains not just in the production of high-tech equipment, as expected, but also in services such as retailing -- which had long been assumed to have little prospect for much improvement. The quartet can now be heard on iPods and even cellphones, meaning that even if musicians themselves aren't more productive, the methods of distribution are.

In fact, it is in services -- particularly in retail and wholesale trade, in something called the "Wal-Mart effect" -- where economists credit a good part of productivity gains in the late 1990s and early 2000s. That led Brookings Institution economists Barry Bosworth and Jack Triplett to conclude in an influential 2003 paper that Baumol's disease "has been cured."

Along similar lines, a pair of economists from the Federal Reserve -- Carol Corrado and Paul Lengermann -- argued in a recent paper that much of the growth in the U.S. economy since 2000 can be accounted for by strong multifactor productivity growth -- which includes labor as well as inputs such as capital and materials -- in industry, a "remarkable turnaround" in finance and business services, and an "end to the drops" in productivity in personal and cultural services.

Still, the Fed economists and their co-authors estimated that from 2000 to 2004, multifactor productivity growth averaged just 0.2% a year in personal and cultural services (which include health care, social assistance, recreation and food services, among others), compared with almost 2% for finance and business, 2.6% for distribution and 5.4% for high tech.

Sectors where productivity is high and average labor cost low "are those things that can be automated and mass-produced," Mr. Baumol, now in his mid-80s and still teaching, said in an interview. "And things where labor-saving is below average are things that need personal care -- these are health care, education, police protection, live stage performance... and restaurants."

Uh-oh.

U.S. job growth has been concentrated in those latter sectors. More than half of the 1.6 million jobs added in the private sector in the past year have been in food services, health care and social services. Food services alone account for more than 20% of all new jobs this year, including government.

Going back to Baumol's disease, it still takes a bartender the same two minutes it always has to make a gin-and-tonic. And an "end to the drops" in productivity referred to in the Fed paper may not be enough to sustain living standards over generations.

[Productivity]

While there's surely a cyclical component to recent job gains and losses -- areas such as health care and personal services tend to lag the business cycle -- there are longer-term forces at work.

Population aging will shift more of the U.S. economy toward one-on-one services. The Labor Department estimates that between 2004 and 2014, seven of the 10 fastest-growing occupations will be in health care, and health-care employment will double the national average. Employment in leisure and hospitality will also outpace the average, though not by as much.

"If what we're starting to see is an increase in personal services -- nursing homes, care for the elderly -- as they become bigger and bigger, that would create a very important composition story," says Martin Baily, a productivity scholar at the Peterson Institute for International Economics, a Washington think tank.

If employment and sector-specific productivity trends continue, "I would start heading toward 1.5%" annual productivity growth over the long term, says Mr. Bosworth. Still, he doesn't think that will happen; he stands by his notion that Baumol's disease has been cured, in part because medical care -- the ultimate test of Baumol's theory because it's set to account for so much of the economy -- holds promise for "big productivity gains."

There are other reasons for optimism. The U.S. is very competitive globally in high-paying and high-productivity services, so any expansion of trade could positively affect the job mix domestically.

Asked whether Baumol's disease spells doom for productivity, its namesake replies "yes and no."

"It is true that in money terms our productivity will be slowed down by the shift in labor from agriculture, manufacturing and services like telecommunications into services like health care and education," Mr. Baumol says.

"But if you count the number of students who have graduated or the number of people who have been taken care of after a heart malfunction, that is not going down." And the benefits of education and health care, on future output, can be hard to measure.

As for the disease that bears his name, not only does Mr. Baumol say it hasn't been cured, he adds: "I can brag and apologize that we've made the longest-lasting [correct] prediction that's ever been made in economics."

Edited August 14, 2007 by Guy

[The Magnificent Goldberg]

To dredge up this topic, there was an interesting article in the WSJ yesterday. (Quoted below.)

I'm not an expert on productivity but IMHO the idea that in the United States services have low productivity growth and manufacturing have high productivity growth is either BS, or irrelevant.

If there really is a big gap in productivity growth between the two sectors, why aren't resources flowing from services to manufacturing?

Really, what Baumol is saying - and it makes sense to me - is that some sectors of services - those that emphasise or rely on personal performance, such as nurses, lap dancers and prostitutes etc - can't get much more productive per se without sacrificing quality. So if you include figures for those occupations in an average for all services, they drag the averages down.

The question of what makes resources flow or not isn't wholly reliant on productivity changes. What it's actually wholly reliant on is anticipated rate of return - and productivity contributes to that but doesn't absolutely rule it. It seems to me that there are two factors which have a somewhat different impact on the equation.

First, there are market failures. In economics, nothing inevitably happens automatically. A big market failure is lack of knowledge. An investment can only be made where there is knowledge that will enable a return to be made. Some people's imperfect, or limited, knowledge directs them into certain areas of activity (there is a regional effect here, too, it seems to me). Some of those areas may be low productivity services.

The second, and more important, factor is the state of the market; plain ole supply and demand. I think it was you who posted an article here a few days ago which noted the change in price of CDs, going down, and live gigs, going up. It also noted a falling profit margin on CDs. It didn't talk about a rising profit margin for live venues, but I think that was the implication. Profit margins aren't quite the same as rates of return, but they're pretty good indicators, I think.

The WSJ article noted rising demand for some services as baby boomers retire. It didn't note that, because of the generally good condition of the world economy for extended periods since WWII, that generation are uniquely equipped to pay for those care services they are likely to need. But that is almost certainly the case. So there is already more money to be made in this area than was the case a couple of decades ago, and this trend is likely to increase. So resources will flow.

An awful lot of economic theory is based on manufacturing. Economic theory developed in the Industrial Revolution and was very much influenced by what people saw was happening. I'm not saying economists ever ignored services; clearly they didn't. But several good economists I know have told me that, in particular, economics is not very good at cultural stuff. But it is just this area that is a very big part of services growth, now becoming less reliant on manufactured goods as a delivery mechanism.

As ever, we live in intereting times...

MG

[Guy Berger]

To dredge up this topic, there was an interesting article in the WSJ yesterday. (Quoted below.)

I'm not an expert on productivity but IMHO the idea that in the United States services have low productivity growth and manufacturing have high productivity growth is either BS, or irrelevant.

If there really is a big gap in productivity growth between the two sectors, why aren't resources flowing from services to manufacturing?

Really, what Baumol is saying - and it makes sense to me - is that some sectors of services - those that emphasise or rely on personal performance, such as nurses, lap dancers and prostitutes etc - can't get much more productive per se without sacrificing quality. So if you include figures for those occupations in an average for all services, they drag the averages down.

I was somewhat cryptic in my earlier comment, so let me clarify. (And also put the disclaimer that Baumol is a very prestigious economist and I'm just a peon, hence all I write could be complete crap.)

There are two "meanings" of productivity. One is "how many units can you produce an hour." The other is "how much value can you produce in an hour." (There's also a third meaning -- labor productivity. The amount of units or value produced by a unit of worker effort.)

In the 1st sense, it's certainly possible that manufacturing productivity is growing faster than services productivity. But it's not clear to me why this description of productivity is that important for this comparison. Let's say we have a hypothetical economy with two activities -- producing food and producing soap sculptures. People choose to spend 99.9999% of their income on food and 0.0001% on soap sculptures. Let's also say that soap sculpture productivity is growing at 2 times the rate of manufacturing productivity. Does this productivity boom in soap sculptures provide much benefit for the citizens of this economy?

As I said, I'm a complete dilettante on the economics of productivity. But I think it makes more sense when talking about productivity growth, particularly when comparing across sectors or industries, to look at changes in value rather than changes in output.

The question of what makes resources flow or not isn't wholly reliant on productivity changes. What it's actually wholly reliant on is anticipated rate of return - and productivity contributes to that but doesn't absolutely rule it. It seems to me that there are two factors which have a somewhat different impact on the equation.

You're a 100% right, MG.

1) What should actually matter (all things being equal) is (if my economics brain is working properly on a Saturday morning) the level of productivity, not the rate of productivity growth. Of course, given enough time, if activity A is growing faster than B, it will catch up and then leap ahead. But if manufacturing productivity is currently at a lower level than services productivity, then the heavy tilt of resources toward services could be consistent with a higher rate of productivity growth in manufacturing.

2) Marginal, not average, productivity matters. It could be that for existing manufacturing activities, the average level (or growth rate) of productivity is higher than that in existing service activities. This is completely consistent with the marginal productivity of the two sectors being equal.

3) As you say, rate of return on investment is not the same thing as productivity. It could be that the services sector is more monopolistic (for at least some factors of production) than the manufacturing sector. For example, in some industries capital and/or labor may be able to capture productivity gains whereas in others those gains are almost immediately passed onto consumers. However, this wouldn't explain a very long-term deviation, because eventually (in the very long run) only the differential productivity growth rate would matter. And in general, I don't find this story to be very plausible.

4) For whatever reason, manufacturing activity generates positive externalities whereas services activity does not. This would result in a market failure. However, my comments on (3) pretty much apply to this.

5) A market failure due to lack of information. Again, I think my comments on (3) apply to this one. And I find it to be extremely implausible that workers or investors are unaware of large unexploited gaps in returns between sectors of the economy.

First, there are market failures. In economics, nothing inevitably happens automatically. A big market failure is lack of knowledge. An investment can only be made where there is knowledge that will enable a return to be made. Some people's imperfect, or limited, knowledge directs them into certain areas of activity (there is a regional effect here, too, it seems to me). Some of those areas may be low productivity services.

The second, and more important, factor is the state of the market; plain ole supply and demand. I think it was you who posted an article here a few days ago which noted the change in price of CDs, going down, and live gigs, going up. It also noted a falling profit margin on CDs. It didn't talk about a rising profit margin for live venues, but I think that was the implication. Profit margins aren't quite the same as rates of return, but they're pretty good indicators, I think.

The WSJ article noted rising demand for some services as baby boomers retire. It didn't note that, because of the generally good condition of the world economy for extended periods since WWII, that generation are uniquely equipped to pay for those care services they are likely to need. But that is almost certainly the case. So there is already more money to be made in this area than was the case a couple of decades ago, and this trend is likely to increase. So resources will flow.

An awful lot of economic theory is based on manufacturing. Economic theory developed in the Industrial Revolution and was very much influenced by what people saw was happening. I'm not saying economists ever ignored services; clearly they didn't. But several good economists I know have told me that, in particular, economics is not very good at cultural stuff. But it is just this area that is a very big part of services growth, now becoming less reliant on manufactured goods as a delivery mechanism.

As ever, we live in intereting times...

MG

[The Magnificent Goldberg]

There are two "meanings" of productivity. One is "how many units can you produce an hour." The other is "how much value can you produce in an hour." (There's also a third meaning -- labor productivity. The amount of units or value produced by a unit of worker effort.)

In the 1st sense, it's certainly possible that manufacturing productivity is growing faster than services productivity. But it's not clear to me why this description of productivity is that important for this comparison.

As ever, a very interesting response, which I'll think about. For the moment, a question - since I'm not any kind of economist, just an intelligent ex-customer of economists.

If you adjust the second meaning of productivity for inflation, doesn't that produce something that looks very much like the first? (Because a "unit" means something concrete only in relation to a specific operation, otherwise it's an average.)

Oh, and in case other readers of this thread are getting bored with this

MG

[Jazzmoose]

1) Since agriculture and services can be important sectors of economy, "industrial strength" and "economic strength" are not the same thing.

While you are correct to a degree, I don't think it's as big of a degree as you think. The British thought "services" could replace manufacturing, as did other major economic powers of the past. When they reached that stage, they were already fading. As are we.

To dredge up this topic, there was an interesting article in the WSJ yesterday. (Quoted below.)

I'm not an expert on productivity but IMHO the idea that in the United States services have low productivity growth and manufacturing have high productivity growth is either BS, or irrelevant.

If there really is a big gap in productivity growth between the two sectors, why aren't resources flowing from services to manufacturing?

Sorry; I guess you lost me here. I'm trying to figure out what your comment has to do with your quote of my comment, and I can't connect the dots. Help!