Brad DeLong's Semi-Daily Journal (2004)

OK, I read it and I have a hard time understanding it. I always have a hard time with these types of commentaries. I don't know whether it is my lack of knowledge of macro finance, or that financial people think about things differently than I do.

It seems to me that he is talking about the effects of an increase in the long run trend in productivity growth. But when I look for an explanation of where that change in the long run trend come from, I get a cyclical explanation:

"So if faulty data don’t explain [higher productitivy] , what does? ... companies purged the hiring excesses of the 1990s over 2002-3, cyclically boosting productivity. And the high ‘fixed’ costs of healthcare and pension contributions also stymied hiring ... But as companies have now gone a little overboard in the effort to purge excess employment, I think pent-up demand for hiring is beginning to build."

Am I ignorant, or dumb. Or both. Or what. Would one of those macro/finance mavens help me here?

I am with you jml, so don't construe this as help. I like the remark that despite the extra security costs across the board due to 9/11, the productivity numbers shone even brighter. That despite higher health care costs for those employees not fired, these numbers still shone.
Was there that much labor slop in the economy in q3-4 of 03? This "cyclical theory" seems plausible for the period pre-9/11, but not for this more recent period.
I think the (long run) productivity growth trend as charted is contaminated with financial instruments rather than better mousetraps ( especially of the 'complementarity' flavor). If it were real physical innovations, we would not see the volatility in these quarterly measurements. Not that better mousetraps do not appear but they do not account for the lion's share of these (unbelievably) large numbers.
There is a point where the numbers confront rather than confirm, reality, no?

I find interesting the expectation that increase corporate governance regulations would lead to reduced economic efficiency. IMHO, it's clear that upper management has massive power of possession. If you're a stockholder, you can always sell, but most stocks you could move to should be similar. With lots of lack of information to make it worse.

shorter version.

world does not have ability to use fiscal or monetary policy to stimulate if weakness emerges.

so risk reward balance heavily skewed to negative.

Is productivty rebound real.
Yes, but much of it is cyclical-- normal pattern is for productivity to be highly cyclical. Very strong in recovery but slowing sharply in expansion -- where we are now
and evidence building we are now seeing sharp slowing of productivity growth.

Question -- does problem of poor data on outsourcing mean growth & productivity overstated. Yes, but it is not significant.

Rest of world highly dependent on exports to the US to stimulate growth. But if weak dollar
works it means that source of growth for rest of world goes away.

Bottom line, environment very risky because of inability of govts to stimulate & if US imports do not growth sharply will dampen rest of world growth.

If inflation emerges & fed tightens could have very negative consequences much sooner than people realize.

That is my version of what article said.

thanks for your summary spencer. I am still waiting for some one to tell me where this new long run productivity increase comes from. I will kind of agree with calmo (even though he admits he didn't *mean* to help me out) until I see a straightforward explanation. I don't see the article being helpful on that point.

It can't be as simple as 'Technological innovation has this economic characteristic: productivity growth is positive. In the long run, we are able to produce more with less hours worked, thanks to inquisitive and original minds.'
So much for the general "long run productivity increase" and it's origins.
But jml is after the "new" long run productivity increase. There are some ( bakho et al) who believe it's robotics and generally high tech.
So much of our "Product" is in the housing industry that one would expect to see this "new" productivity there. Maybe a version of Henry Ford production lines dropping yet another house on the lot or...? I don't see it and I work there.
I don't work in Finance but I can see that not only does the US have the lion's share of that industry but it also has the highest productivity numbers. So what counts as a unit of product in Finance?
My feeling is that this is as hard to measure as the latest derivative species. Issuing a $500M bond registers as twice as much product as a $250M bond? Someone out there knows.

calmo: If you follow this line of thought through, you will arrive at a point where productivity can only be compared for occupations that produce countable items of some standardized quality. The universal (and unfortunately most reliable) countable unit of economic output is the dollar (or other currency unit) of revenue.

So it basically comes down to summing the revenues minus input and operating costs in the respective industries and dividing by hours worked. This is combined with price indexing (in an attempt) to remove the portion of revenue growth/decline that comes from price hikes/drops. This is what leads to the GDP number and the GDP deflator, respectively.

And there is of course a lot of leeway for systematic error: for example, how do you measure changes in product quality properly -- maybe a manufacturer cuts support cost by having the 800 number put to a phone menu that takes customers through 15 mins of waiting time so that 90% give up before getting to the operator, or replaces a part of the product by something cheaper and more flimsy, so that the product does not last 5 years but 2, etc. How do you put a number on that?

There are also other controversial ways of boosting productivity -- feeding drugs and hormones to farm animals, using abbreviated (bio)chemical processes in food processing leading to hopefully safe but inferior products, and so on.

And of course, "hours worked" are quite flexible as well as we know (_where_ are they worked, is overtime properly reported, are there uncounted externalities, etc.).

All of this makes it very hard to compare productivity numbers even within the same economy.

I'm curious about the source of the postulated permanent increase in productivity growth because the last review article with actual numbers and statistics and evidence and other good stuff like that couldn't decide between the following hypotheses:
1. it is some kind of national accounts statistical illusion that we don't understand yet,
2. it is from efficiencies in making more computers, but which won't have a long term effects, because the productivity increase is only showing up in making computers and nothing else.
3. it is from efficiencies in making more computers, these will have a long term effect, but we can't measure it well because it is working through various kinds of financial and distribution services that cannot be measured very well,
4. we don't know.

It seems to me that this is a very important issue with important short and long run implications for economic welfare, ss and medicare viability, etc. The idea of its existence seems to already have affected policy. Didn't Greenspan basically say he believed in a permament change in economic conditions, in some kind of mix of reduction in systematic risk and increase in productivity growth when he was in his "we took a little corrective action, and it turns out to not be a big bubble anyway" stage of thought about a year ago?

So I am always on the prowl for new info on this. I think the article was very informative. But still, it seems like he discusses some of the important implications of the long term change, but when he turns to the source, I don't get any long run explanation at all. So that made the article a little odd for me.

jml, everybody: Does my contention that productivity is revenue minus cost base (i.e. "value added") divided by hours worked not explain it nicely?

There may be an illusion, but there is not necessarily malevolent tampering with the numbers, at least by public agencies. (What numbers companies report I would trust less; the have incentives to please shareholders and others.)

For example, if you outsource labor or other parts of your cost base (like component prices if you are an integrator) to cheap locales and manage not to pass on all the savings to your customers, your productivity should increase. You basically are able to produce the same number of revenue dollars or even product units with less labor (as part of your expensive labor has been converted to cheap imported _components_, and foreign worked hours are not reported). Whether that counts as a productivity increase by your definition I have to leave to you. I have some doubts.

Similarly, if you outsource service-type work, it is likely that the ensuing "import" of labor is not reported as an import. So if you are let's say a software company with 1 million annual labor hours (2000h/yr * 500) and you outsource 10% of your labor to Asia, I would suspect you will report 900000 labor hours afterwards. When assuming 1 domestic hour costs you $70 (don't forget benefits and payroll taxes) and the foreign replacement costs you $20, you save $5 million in labor cost. Assuming your revenue is reduced somewhat due to harsh business climate, you may still see more productivity due to 10% less domestic labor.

And companies with a high percentage of exempt (from the FLSA) employees will report the contractual hours of 40h/week, regardless of how much exempt (and thus untracked) overtime is actually being worked.

cm:

Free trade under the doctrine of comparative advantage is supposed to raise productivity rates, regardless of whether it is due to outsourcing of labor inputs or to the direct trade in goods. Also, the contract cost of outsourced services would be counted as an imported input.

One source of economic growth through productivity increases in real output that most standard economic theory does not consider is the diversity in an economy of products, processes and sectors. Products and processes tend to come in matched sets of substitutes or complements. The greater their diversity, the higher the probability that innovations in products or processes can form links so that they can find market niches from which to grow and that widely distributed or dispersed, but small innovations can contribute in aggregate to overall growth. Also the higher the probability that cascades of innovations resulting in extended chains of such links can occur. It would seem that micro-chip technology, in particular, would have an indeterminately large potential for applications and adaptions across a wide variety of processes, products and sectors. One of the stars of the recent productivity figures has be the growth in service sector productivity, "the Walmart effect". Previously, productivity growth was thought to be constrained by the fact that technical improvements in manufacturing led to a shrinking of the manufacturing sector in favor of the growth in low productivity services, shrinking and diluting the effects of such technically induced productivity gains.

I don't have much quarrel with the benefits of technical productivity growth per se. My questions would be about how, by what specific pathways or mechanisms, such gains are distributed in the medium term. In particular, could the benefits of such productivity gains be increasingly unequally distributed and concentrated, leading to damaging, countervailing socio-economic effects? Could the concentration of economic control and the shortfall in aggregate demand resulting from such an increase in inequality lead to such productivity gains colliding with a lowered growth path?

In the long-run, my questions to economists would concern the possibility of managing and steering technically induced economic growth in the direction of technological "build-down", that is, the development of technical improvements in the direction of an increasingly efficient use of resources, in the place of relying on the maximalization of resource exploitation, in the light of ecological, natural resource, and human population constraints on the production and distribution of social wealth. No doubt such a prospect would require a considerable degree of coordination and planning, which would imply some modes of government action and intervention, rather than reliance on the vagaries of free markets alone. But perhaps such wisdom is so improbable as to be impossible to expect, since wisdom is always in short supply, low demand and incapable of being technically produced.

I admire your pragmatic/managerial approach john, but I am stuck on 'what is this thing called productivity growth (love)?'. Berner even writes about 'productivity acceleration' and who knows, maybe there's 'productivity jerk'. What I do know is that if we can't get a grip on 'productivity', we should not be using it to forcast GDP. ( I pass on the overwhelming temptation to make a jerk joke.)
Looking at other countries' productivity numbers, I found that the studies were mostly 5-7 yrs in duration; that there were similar changes across those countries studied ( Japan, US, UK, most of Europe); that the US had ( usually) higher numbers. Not exactly what I am looking for but still reading.
cm -your (re)definition of productivity (revenue less cost base, divided by hours worked) is? original and has the promise of being measurable, but how is this different from profitablity?
I want to argue that when Lehman Bros issue a bond for $500M (for basically, a percent fee) their productivity on this transaction is identical to that for a $250M issue. Your view would have it that the first deal was twice as productive ( assuming the same hours worked).
I am looking at the banks only because they seem to be doing as well as these productivity numbers.

john, calmo: My point was that Walmart's productivity is probably not measured in how many tons or units of product they move through the sales channels, but how much money they make on it, scaled by price indexing. (But John, prove me wrong and say that the government is counting how many cups of jello Walmart is selling each quarter.) The productivity of a cleaning firm is probably not measured by how many square feet of floor they are sweeping, how many calls to remove messes they process, etc.

I work in a software company that has growing offices in India and China, and where my business unit is producing a range of products with a 3-digit million revenue. All US engineers are exempt, and are probably reported as working 40 hours weekly. John, are you going to say that my company has to report their foreign labor hours? Of course they have to report how much money they are transferring in foreign paychecks, which will make its way into import figures. But regardless, all overtime of exempt employees will go unreported. If foreign labor hours are not reported (and I don't see how they reasonably could), this will distort domestic productivity numbers.

Any ideas? I plan to read up on the topic as time allows, but for the time being my working hypotheses is that money is being measured, fully in line of the view of many people that apparently production of and bragging about increasing financial numbers is the primary purpose of our economies.

There is always a gap between the "real" economy and the nominal economy measured in monetary prices. This is one of the core objections against the neo-classical turn to marginal analysis. But "real" wealth is the increase in the size of the distributable surplus product, that is, the total product minus that portion of the total product that is used up in the process of production. And clearly technical improvements in production processes are the primary source of per capita increases in the surplus product in an economy. What is actually exchanged in markets amounts to producibilities and their ratios, which is what is behind the supposed law of supply and demand. In this sense, a lower long-run price for something amounts to a measure of increased productivity, (though there is inflation, which could very roughly be taken as a measure of unmet demand and as a lure to further investment in the productive possesses that reduce the costs of production in meeting an ever-shifting demand.) To be sure, such an economic view abstracts from how final consumption products are situated in social organization, good old-fashion use-value, and any final consumption product involves the several combination of different labors that can not be separated out into distinct values.But averaged out, technical and organizational innovations do lead to a secular trend increasing the real surplus product distributable between capital and labor. Markets do both stimulate and promulgate technical innovations, but, unto themselves, they do not guarantee that they will occur.

I am not so much interested in the precise measurement of productivity gains, whether they amount to 4.8% or 3.5% in the last two years and such-like. I would leave that to the earnest efforts of the statisticians. (And let's not forget the ticklish issue of hedonic pricing, when making international comparisons.) What should be noted is the difficulty of actually achieving productivity gains. (Consider a learning curve that reduces costs of production by 10% per annum; after 7 years, gains have been reduced by half and it's downhill from there.) But if there are consistent reports of an increase in the secular trend to productivity gains in advanced economies, then the best guess would be that the long anticipated effects of IT are being distributed throughout such economies. My concerns would not be about the benefits of such innovations, but with how they are distributed, with the cultural and social shocks they entail, and with how they might lead to an increased concentration of economic power.

As for the issue of outsourcing, I think it is obvious that such practices do not lead to any actual productivity gain. To the contrary, they lead to lower levels of (labor) productivity at decreased levels of capital intensity. Call it the cultural contradiction of capitalism. The counter-argument would be that they free up resources for alternative uses. But how such resources would be used, if effectively at all, would depend on the policy environment and the dim discernment of valid ends and highly mediated means.

I can well sympathize with the situation of well qualified engineers who must dance to the dictates of know-nothing MBA's. But since I work at a much lower pay-scale, I can only recommend a robust principle of charity, combined with a richly vindictive fantasy life that might lead to more productive insights.

john c. halasz: Good analysis, and by and large I agree. Your point that the market's outcomes must be viewed relative to policy environment is very valid. The economic policies of the current and past administrations are even in the most favorable view at best misguided, and besides ideological biases, they are the result of questionable measurements. If you are bullshitting yourself about what causes productivity (and "prosperity"), you will try to do more of it, making matters worse.

If you measure the miles that you drive your car by how much gas it burns (because the odometer is broken or missing, or you don't know how to read it), you may think you drive longer distances by looking at your increasing gas bill, where in reality your engine performance goes to hell.

On a different note, what are you doing? Maybe your pay is not really (that much) lower than an engineer's.

Online Casino Directory

nice