Semi-Daily Journal

Correct that typo:

>> Gross Domestic Putchases

Although Putchases is kind of a cool word.

how is service outsourcing counted in the GDP? For instance if a company moves the call center for their billing operations overseas it is still treated as an expense, but how do the economists figure out where the expense took place (since there are no import/ export charges save the phone bill)?

Brad DeLong

This website is quite wonderful, but have you noticed how dominated it is by male posts and references and commentators? Of course, you have the odd post by the typically rightest typically blond Virginia P. But, a Mureen Dowd or Molly Ivins are merely "women." The male commentators on the site typically respond to male commentators. No matter, women probably are not ready for economics, politics, even owls.

Mimi

The whole thing has me wondering: Even if the Bush tax cuts had been *correct* economically, i.e. had been targeted at the poor and middle class, who would have spent the money, generating demand, and in theory, production, it might not have stimulated job growth in the US, since more and more consumer goods are manufactured abroad.

is a putchase a putsch via a purchase of insiders? just kidding

Could wage rigidities lead to the the gap between GDP and employment growth? seems possible without thinking through a formal model

I think much of the ironic combination of job loss and high productivity derives from working the same people more hours instead of hiring more people. This is basically due to two things:

1. The companies just want to get more work out of each person at the same cost - not good for you, if you are the worker. They are facilitated in this by the ill-advised prevalence of "exempt" (from 1.5 OT) status. I know so many people who obviously aren't supervisors, etc., being paid a salary, and working 50-70 hours a week. This is a scandalous white-collar sweatshop environment. The law should be ammended to eliminate almost all exempt categories, and to put the rest on at least straight time over time.

2. The second reason is not the company's fault/greed, but derives from how insurance etc. costs work. The company has to pay out per head, giving it an incentive to hire fewer workers and overwork the rest.

Replace that system with a health care plan, and possibly for other costs, that allows (requires?) the company to pay out a straight percentage of employee compensation to insure all its workers. (Putting aside for now the question of how they charge, "incidence," the cost to the employees.) That would have the two benefits of encouraging employment, and making it easier for lower paid workers (and lower-paying and earning) companies to give insurance to workers.

Neil

I think much of the ironic combination of job loss and high productivity derives from working the same people more hours instead of hiring more people. This is basically due to two things:

1. The companies just want to get more work out of each person at the same cost - not good for you, if you are the worker. They are facilitated in this by the ill-advised prevalence of "exempt" (from 1.5 OT) status. I know so many people who obviously aren't supervisors, etc., being paid a salary, and working 50-70 hours a week. This is a scandalous white-collar sweatshop environment. The law should be amended to eliminate almost all exempt categories, and to put the rest on at least straight time over time.

2. The second reason is not the company's fault/greed, but derives from how insurance etc. costs work. The company has to pay out per head, giving it an incentive to hire fewer workers and overwork the rest.

Replace that system with a health care plan, and possibly for other costs, that allows (requires?) the company to pay out a straight percentage of employee compensation to insure all its workers. (Putting aside for now the question of how they charge, "incidence," the cost to the employees.) That would have the two benefits of encouraging employment, and making it easier for lower paid workers (and lower-paying and earning) companies to give insurance to workers.

Neil

Yeah, the solution to Brad's question would seem to be that the internet serves to locate low-value-added inputs overseas, rather than the high-value-added inputs Roach claims. That would allow for a much bigger wedge between output and employment. Productivity in the US would also go up, just as a result of the math. If the average output per hour is $2 worth of goods and services (don't even start, I just made the number up for the example) and a job that produces only $1 worth of stuff per hour is exported, productivity goes up without producing a single extra thing. If we also add a $3/hour in output job, then GDP goes up and so does productivity. The data seem to argue that Roach is looking at the exception, not the rule, for exporting jobs.

Welcome, Mimi,

Meet Anne and Lise. They are smart and fun and seem to be women. Otherwise, how can you tell outside those case where full names are given (sometimes not even then) the sex of posters at Brad's place?

K

I think much of the ironic combination of job loss and high productivity derives from working the same people more hours instead of hiring more people. This is basically due to two things:

1. The companies just want to get more work out of each person at the same cost - not good for you, if you are the worker. They are facilitated in this by the ill-advised prevalence of "exempt" (from 1.5 OT) status. I know so many people who obviously aren't supervisors, etc., being paid a salary, and working 50-70 hours a week. This is a scandalous white-collar sweatshop environment. The law should be amended to eliminate almost all exempt categories, and to put the rest on at least straight time over time.

2. The second reason is not the company's fault/greed, but derives from how insurance etc. costs work. The company has to pay out per head, giving it an incentive to hire fewer workers and overwork the rest.

Replace that system with a health care plan, and possibly for other costs, that allows (requires?) the company to pay out a straight percentage of employee compensation to insure all its workers. (Putting aside for now the question of how they charge, "incidence," the cost to the employees.) That would have the two benefits of encouraging employment, and making it easier for lower paid workers (and lower-paying and earning) companies to give insurance to workers.

Neil

K Harris

I think I was just in the mood for a morning moomph. This really is a wonderful website and I will later be reading about owls. Your comments have been interesting as the others.

Mimi Better Spirited

Since I have opened and shut this window a couple of times, I'm going to assume my earlier attempt at a comment evaporated. Excuse any double post.

So, as I was saying...data seem to suggest that Roach has identified the exception, rather than the rule, in internet-based overseas outsourcing of jobs. For all that sending software engineering and accounting and factory jobs offshore is getting all the headlines, the aggregate GDP, employment and productivity data indicate that it is still low-wage, low-output jobs that are headed overseas. A decline in aggregate hours and a rise in GDP is easiest to arrange if you ship lots of low-output jobs overseas while adding a few high-output jobs. Sure, there are plenty of other sources of productivity gain to make this not absolutely essential to make sense of the data, so Roach could be right about the overall picture in outsourcing. Still, the easiest fit with headline GDP, hours and productivity data is shedding lots of low-output jobs and adding a few high-output jobs.

OOOOOPs. Sorry. Prior version way the heck up there.

Wow - this gives me a great idea for those masters of data manipulation at the White House. They should change GDP reporting in the following way. Have GDP defined as only domestic demand and just exclude net exports altogether. My math says there would be an immediate jump in demand equal to 5%. Now I know this is dishonest, but remember when NRO wanted to exclude government spending - or was that government purchases? I bet the Bush White House could do this and none of the readers of NRO would even notice.

So, when can we bring back Smoot-Hawley?

http://www.epinet.org/content.cfm/webfeatures_econindicators_jobspict

While manufacturing employment once again contracted sharply in August (-44,000), job losses were spread broadly across industries, with service-producing industries shedding 67,000 jobs.

Within services, losses occurred in most subsectors, excluding health care, one area that continues to post strong job growth. The information industry also continues to contract, losing 16,000 jobs in August. Since the recession began in March 2001, IT employment has fallen in all but one month. The IT sector has shed 12% of its jobs since then, comparable to the 14% decline in factory jobs.

Whether productivity growth is quite as high as recorded this last year is less important than noticing the lack of sufficient global demand to spur job creation in America. Private debt limits American demand. Demand from European and Asia cosumers for American products is too limited. Fiscal policy has had a muted stimulus effect and there is little more the Fed can do to spur demand.

We are growing too slowly for significant job creation and likely will continue to grow too slowly for at least the rest of the year.

I agree with jjj's comments implying a divergence between GDP and Gross Domestic Purchases depends on accounting of costs.

Where are expenses and profits booked for goods and services produced in China? Hong Kong with its corporate tax rate of 16% seems a much better place to park profits than the U.S. (35%?) How "profitable" are China operations on paper?

If expenses are inflated there will be no divergence.

If job losses were indeed spread across sectors, that would be pretty interesting and an argument against the restructuring hypothesis. We'll have to see how the data is revised in the months ahead.

Also, I'm wondering what percentage of total mfg employment 44,000 is, and what fraction of total service employment 67,000 is.

This is more like it! Where *do* those darn numbers come from? Thanks, K Harris et al.

Just as an aside, it would not surprise me if foreign demand for US goods was low not just because of low growth levels in those economies, but also for political reasons. Not only do many governments have reason to want to make us sweat, but polls show individuals in those countries having a very strong animus against the US. That can easily translate into refusal to buy American-produced even if they do start growing.

"Yeah, the solution to Brad's question would seem to be that the internet serves to locate low-value-added inputs overseas, rather than the high-value-added inputs Roach claims. That would allow for a much bigger wedge between output and employment. Productivity in the US would also go up, just as a result of the math."

Posted by K Harris at September 8, 2003 10:31 AM

K, I am admittedly not the brightest bulb in the pack, but I cannot see why that isn't the point of the math if we are measuring U.S. productivity growth.

As to the overall question of whether this type of productivity growth is sustainable (i.e., will re-employment of those displaced cause an equal decline productivity), I personally can't think of any reason to assume it is unsustainable. Then again I'm not the brightest...

"I'm wondering what percentage of total mfg employment 44,000 is, and what fraction of total service employment 67,000 is."

At this stage K I think it doesn't matter. The interesting thing is its starting to happen.

"the internet serves to locate low-value-added inputs overseas, rather than the high-value-added inputs Roach claims".

I think we have to wait and see K. I don't see what is the real defence against doing a lot of US high-value input as relatively low-value input elsewhere.

Isn't the answer to Brad's puzzle to be found in the trade deficit. Isn't this the point where the system has sprung a 'leak'.

"what fraction of total service employment 67,000 is".

Well here's the It data from the BLS report:

Employment in the information sector fell by 16,000 over the month. Since its recent peak in March 2001, the number of jobs in this sector has declined by 459,000, or about 12 percent.

"the internet serves to locate low-value-added inputs overseas, rather than the high-value-added inputs Roach claims".

"A decline in aggregate hours and a rise in GDP is easiest to arrange if you ship lots of low-output jobs overseas while adding a few high-output jobs...........so Roach could be right about the overall picture in outsourcing."

Look, I think the point is not so much where we are now, as where we might be going. How do you protect high-value jobs in the US from becoming cheaper high-value jobs outside. Saying no to globalisation. That sort of about-turn on the part of the US would look even more silly than the current Iraq one. I think to understand Roach's point you need to think about global imbalances.

The solution to Brad's connundrum. Perhaps the trade deficit might provide the key. Here is where the leak is. It doesn't really matter whether the import is petrol or pre-prepared software, or genetic code. It still has to be paid for. In GDP terms - assuming we're talking intermediate products here - it's an input that doesn't have a labour component.

BTW I was trying to square the circle of job loss numbers:

More losses in computer design:

"Computer systems design lost 8,000 workers over the month. Since peaking in March 2001, employment in this industry has declined by 232,000."

Whilst two areas gained:

"A gain of 25,000 jobs in health care and social assistance in August was about in line with its average monthly employment increase over the
prior 12 months."

"Construction employment edged up over the month. Since February, the industry has added an average of 20,000 jobs per month."

Anyone notice anything interesting about these numbers?

OOps, I thought the first post was lost.

Sorry.

While I'm back: look at services in the trade balance (source: BEA stats):

"The surplus on services decreased to $14.4 billion in the first quarter from $16.1 billion in the fourth.

Services receipts decreased to $74.6 billion from $75.3 billion. Large declines in travel and passenger fares, reflecting concerns about the war in Iraq and the SARS virus, were partly offset by increases in "other" private services (such as business, professional, and technical services, insurance services, and financial services) and in royalties and license fees.

Services payments increased to $60.2 billion from $59.2 billion. Declines in travel and passenger fares, largely reflecting concerns about the war in Iraq and the SARS virus, were more than offset by increases in all other services categories combined."

Bottom line. you've sprung a leak and it needs fixing.

OOps, I thought the first post was lost.

Sorry.

While I'm back: look at services in the trade balance (source: BEA stats):

"The surplus on services decreased to $14.4 billion in the first quarter from $16.1 billion in the fourth.

Services receipts decreased to $74.6 billion from $75.3 billion. Large declines in travel and passenger fares, reflecting concerns about the war in Iraq and the SARS virus, were partly offset by increases in "other" private services (such as business, professional, and technical services, insurance services, and financial services) and in royalties and license fees.

Services payments increased to $60.2 billion from $59.2 billion. Declines in travel and passenger fares, largely reflecting concerns about the war in Iraq and the SARS virus, were more than offset by increases in all other services categories combined."

Bottom line. you've sprung a leak and it needs fixing.