Brad DeLong's Semi-Daily Journal (2004)

That 20 m bucks is in effect all wages -- retraining, redesigning work process, rearranging relations with vendors... all this is wage... and so is software; software is almost pure value added... and, so is hardware, in fact; what is a computer physically, after all? it is a bit of petroleum, a bit of sand, a bit of metal (gold but very very small amount)...

and something else is happening too: with every new version of software, all that organizational investment comes to you as imbedded in the software, thus software replacing labor a bit more with every new version, "freeing" a few more ERP employees...

I keep saying every body needs to be given a broad based college education like Harvard arts and sciences and nobody believes me... why do I think so? so that they can learn new things quickly, that's why..

Bulent: "so that they can learn new things quickly, that's why.."

And I thought for some time that your desire was to produce well-rounded and culturally rich people ... :-)

Ah. "complementarity" Sounds so wonderful! ( Let's name a cologne after it.)
But does it work?
Is it true that in those spots ( not Zambezi) where computers and interconnectivity have been incubating for years, establshing this complementarity, that we find astonishingly high productivity numbers?
Oh wait... there's that CBS ( complementarity buzz saw) now: making furniture out of that tree right before my very eyes!
Or is it just damn Marketing?

>>What is striking, however, is how large the investments in organization capital - "computer-enabled assets" to Mr. Brynjolfsson - are compared with technology investments.>>

As much of the organization capital is bound into workers own skills, and these skills are built through working, learning by doing, much of the mechanisms regarding capital formation has to be rethought.

Consider that a the classical 'old economy' link between (real)interest rates and expected growth - even though its theoretically firm - has shown little empirical strength.

http://bonoboathome.blogspot.com/2003_12_28_bonoboathome_archive.html#107282098276017616

http://blogofpandora.blogspot.com/2003_12_01_blogofpandora_archive.html#107031621675437382

CM: Well rounded and culturally rich people do learn quickly. In fact, well rounded and culturally rich people are the only people who learn quickly without being smart alecs and without going for short cuts at expense of others. With well rounded and culturally people, you get work force that is competent, honest, and well mannered -- well mannered includes aesthetics as well.

So let's go for higher education for all -- a la Harvard FAS.
----

Calmo:

Where you simply "computerize" the existing process, you don't get helluva distance in productivity.

Where you re-design re-engineer your process treating hardware software just like another new machine tool option, and do all that properly, you get good productivity results.

--------

Mats:

"...much of the mechanisms regarding capital formation has to be rethought...."

Bravo!

What is "capital"?

The science of economics does need to be re-invented.

The cost of implementing an ERP system can certainly be impressive. I watched such a thing at some remove, and the initial $4M investment slouched towards $10M over a couple of years. By the time I dropped out, this monstrosity, with its many mincomputers, was still short of the functionality of the micro-based client-server MRP system we had been using, of which the software portion of the cost was more like $30-100K.

The general point - that the entire company benefits from learning the relevant disciplines - is right on.

Other details, like what to centralize and what not, may deserve quibbling. Office workers take to email like ducks to water. Likewise to Travelocity: "Cool! I can pick my own seat!" A quickly developing Web can easily beat a sluggish corporate staff.

Has Mr. Brynjolfsson ever had a real job in his life? Or has he just been a professor talking about what real people do in real jobs while never coming near any of it himself?

I think the latter. Therefore, his opinion must be discounted, about 85% or so.

Chuck Noolan --Brynjolfsson is probably the best productivity analyst around. By the way, as a private consultent he probably makes several times what you do on top of his MIT salary.
He works very closely with many firms in designing their personell - training, knowledge, systems.

I use to work for an insurance company, and we had a saying that every night at 5:00 our
capital walked out the door. Technology is that way,the physical capital is relatively, insignificant. The real "capital" of most firms now is their people and their inbedded knowledge.
That is one reason I have been so concerned about the sever downsizings of the past few years.
When firms lay off so many people the "institutional knowledge" they represent significantly cuts the ability of the firms to actually continue to do the job they use to do.

At a lunch a while back a friend said he is now the firms "old fart". He says his job is to attend meeting and doze-off and not do much. But ever once in a while he is suppose to wake up and say, Wait a minute, we tried this back in 1987 and this is what happened.

The findings that IT investments alone being little if any productivity benefits, while IT investments accompanied by appropriate organizational changes can explain much of the 1990s productivity surge, is far from new: The Economics and Statistics Administration of the Department of Comemrce reported it in "The Digital Economy" reports for 1999 and 2000. Those reports and the two preceding them represented the first rigorous attempts to measure the impact of IT on growth and the economy.

Very interesting article. Given the fuss over transfer pricing and shifting U.S. taxable income to tax havens, consider the following. U.S. parent spends resources on generating organizational capital, which benefits both domestic and international markets. It then sets up a Cost Sharing arrangement with a tax haven subsidiary that will reap all of the profits from international markets. Should it not pay for the value of this organizational capital? Has it when it establishes the new arrangement?

All interesting comments. I had a conversation with an exec at Northrup about 25 years ago and he pointed out a problem with the double humped age distribution of the employees that Northrup had: The old guys did not talk or socialize with the young guys so there was an information/skills gap developing that worried him. Wonder how they worked around this problem. The current "downsizing" may affect the only top end of the age distribution thus the same problem may not develop, however, there will be a price paid.

I am going to use this to ask a question that has been on my mind the last year.

I wonder how much of the business spending on computers over the last year or two is driven by the fact that many firms now lease rather than buy computers. When a lease expires a firm has to
make a decision and it usually ends up being to take out a new lease on new computers. If the computer had been purchased when the two or three year life of the computer expires the firm can make a decision just to keep on using the old computer and put off buying a new one. I know that in the auto sector leasing vs buying has created a mechanism that leads to more new auto "sales" because individuals and firms can not just continue to use the old auto after it is paid for. Does anyone have any experience to suggest that this has played a significant role in business computer demand over the past year or two.

Another factor along this line. In the 1990s the pace of computer and software change was such that a two or three year old system was really obsolete. But now a three or four year old computer and software system is just about as good as an old system because the rate of change in technology --especially software--has slowed sharply.

Has anyone looked at how these two trends are impacting demand?

Spencer, my comment still stands. Unless he's been responsible for the day-to-day development and implementation of systems such as he describes (and such as I do in my highly-paid consulting work), he's just theorizing.

In case any one wants to know, I am one who is just theorizing -- don't know about Brynjolfsson.

"Very interesting article. Given the fuss over transfer pricing and shifting U.S. taxable income to tax havens, consider the following. U.S. parent spends resources on generating organizational capital, which benefits both domestic and international markets. It then sets up a Cost Sharing arrangement with a tax haven subsidiary that will reap all of the profits from international markets. Should it not pay for the value of this organizational capital? Has it when it establishes the new arrangement?"

Posted by: Harold McClure on February 2, 2004 07:03 AM

I've heard of this being done domestically, let alone for international markets. The examples I heard of were for copyrights, patents and other intellectual property. A corporation is set up in a tax haven; the US corporation then sells its trademarks/patents/etc. to the foreign corporation, for a nominal fee. The foreign corporation then leases the use of them back to the US corporation, at a rate which makes sure that the US corporation never officially makes a profit. The profit is, in effect, transferred to the foreign corporation.

Barry:

For international transactions, tax law under section 367(d) and section 1.482 argues that the (nominal) fee be arm's length or fair market value. But some of the "studies" that purport to have established that the value should be low would not even be publishable at National Review Online. Sen. Grassley is rightfully angry over this - but so far the reaction has been to hamstring even further the government's enforcement efforts in lieu of giving them adequate resources.

Goes over much the same area as Brynjolfsson and Hitt, Journal of Economic Perspectives 14:4 (Autumn 2000).

Is the question: whether ALL increases in productivity should be connectable to INVESTMENT spending, for purposes of policy?

Might be difficult. Productivity appears to be the “poor relation” among macro variables, since it is spawned by so many things. Sources of productivity-growth range from mechanical-tool Physics to worker-motivation Psychology. Maybe computer-enabled Organization is somewhere in the middle of the range?

Not really much different from apes inventing head-&-haft ax to chop logs better, causing some rearrangement of preparations for the big pig roast. Which may take a few roasts ‘til they get it right.

Indeed one thing the NYT article and Brynjolfsson & Hitt do NOT consider is a tertiary stage, after (1) computers are introduced and (2) the workers are reorganized, which is (3) everybody gets thoroughly comfortable with the whole system and they start having their own creative ideas and ballsy attitudes, on how to make it connect and work even better and more comfortably. This might not start until a year or two after. Then, I suggest, the fur REALLY starts flying.

At least for a while. We take things in stride quickly, and we are communicating faster. Unlike others, I wouldn’t be surprised if the computer productivity revolution tapers as quickly as it began. Until the next big jump, when we’re jacked-in as cyborgs.

Since reorganization is the rearrangement of PATTERN, it seems highly unlikely that any QUANTIFICATION will ever be predictive of it (Gregory Bateson, “Mind and Nature”, chapter 2, number 10). But of course we will continue to try.

Policywise, we could be in micro territory. There’s a formidable list of barely-priceable adjuncts to basic transactions: asymmetric information, transactions costs, externalities, public goods. Institutional economics might be revamped to consider how reorganization handles some, and creates others anew.

This idea of “organizational capital” seems highly analogous to the old notion, much disputed, of “social capital”, and ought to enable a policy reconsideration of same.

One more nail in the coffin of market fundamentalism?

For example, productivity via worker motivation might as well be improved by government spending and regulation (workplace environment makes workers feel safer; job security gives them a stake, and pride in the proceedings) as well as consumer spending (automobiles and refrigerators enabling the postwar happiness boom).

All of the above is no doubt, relatively, true.

Still, there are huge improvements in productivity to be found in North America just from people getting stuff in out of the rain and snow. (I mean that quite literally: I see construction jobs all the time where the managers' profit is less than the value of the waste in rusted materials that aren't available to become profit on the next job.)

One is reminded of Solzhenitsyn's "1914," where the Russian peasant soldiers wonder at the neatness and wealth of Prussia's farms.

Colour me Prussian, looking at North America's peasant manufacturing operations.

The biggest problem with the implementation of these ERP systems is that companies routinely force their business model into the limitation of the software/hardware that they purchase. Or as someone else mentioned they just instantiate current practices into a computer centric system. These are both the biggest drawbacks to new ERP systems.

Poor business practices will not be aided by an new ERP system. Similarly an ERP system will not make up for a poorly motivated and unknowledgeable workforce.

AS for a Harvard educated workforce, who would do all the work? Just my dig after 15 years in Boston I can tell you that no institution is guaranteed to provide a superior worker, no matter how much they paid for their education. A broad knowledge is of little use unless you can apply it in more abstract ways.

The relationship between software and the organization using it is not sufficiently understood. Managers sometimes try to change the way things are done by specifying software that is only compatible with the new way of doing things. I've seen this fail as resistance to the new way surfaces as complaints about the software.

If you have to build or buy all new software, or a big software item, to change the way things are done, then you can't improve your organization. Improvability requires flexibility in the software.

Too much flexibility means things are not done the right way, or are done inconsistently, so that measurements become meaningless and policies are not followed. If managers can only implement decisions by having the software changed, changes can only happen too slowly. Over time, management will continue to have a need to continually change the software. Continually changed software is not reliable and is not well tested.

There is an internal contradiction between the use of software to implement uniformity, correctness and efficiency, on the one hand, and the need of managers to make decisions daily and implement them in a timely manner, on the other.

So far as I know, the technology to change software behavior in only correct and trustworthy ways does not yet exist.

bulent
"and something else is happening too: with every new version of software, all that organizational investment comes to you as imbedded in the software, thus software replacing labor a bit more with every new version, "freeing" a few more ERP employees..."

If you're claiming that, as your ERP system becomes more imbedded in the organization, you can make do with fewer 'ERP staff' you're mistaken.

You don't need the crew that installed the systems, and redesigned the process. You'll need a small army of folks to maintain your investment. And don't forget there is always an upgrade lurking next quarter or next year.

spencer
"I wonder how much of the business spending on computers over the last year or two is driven by the fact that many firms now lease rather than buy computers."

It depends. A few years ago a 'pc' desktop was worthless if it was more than three years old. You can 'get by' with older desktops as long as you don't need to run new softare or a new operating system. If you do, then you need new equipment.

Servers and data center equipment is much the same; the older gear still works, but the demands almost require new systems every 18 months.

RC
"Poor business practices will not be aided by an new ERP system. Similarly an ERP system will not make up for a poorly motivated and unknowledgeable workforce."

Garbage in, garbage out. The more things change ...

Productivity and efficiency are different things.

Productivity increase does require investment, and I mean capital investment.

Management and labor performing better under a given capital setting is efficieny. That too requires investment, but that's in a different sense, that's investment in human resources.

Productivity increase is not possible without investment in fixed capital because productivity increase involves two simultaneous processes: 1- Replacement of labor by capital. 2- Incorporation of knowledge into production process. (I got that from a document of the French Senate, by the way, and I was surprised to find that the French Senate had such a clear notion of productivity while the French in general appeared to consider productivity a dirty word invented by capitalists to cheat the working class out of surplus value.)

Prussian farms looking neater than Russian farms with comparable technology around 1920s to 1940s is explained by efficieny differential.

The fact that three percent of US labor force produces more foodstuff that the nation needs while 35-40 (or more?) percent of workforce in Turkey is still employed in agriculture is explained by productivity.

Brian:

"If you're claiming that, as your ERP system becomes more imbedded in the organization, you can make do with fewer 'ERP staff' you're mistaken."

Good point. I should have rather said something like "... "freeing" a few thousands of manhours here and there in the enterprise, or bringing about more output and/or shorter lead times with the same work force..."

bulent:

I wonder whether you have read Joseph Schumpeter's "Capitalism, Socialism, and Democracy" (1942). After four chapters of the most precise demolition of Karl Marx ever, he goes on to explain why HE thinks capitalism will end, out of its own SUCCESS. It is similar to the line of thought you pursue, and has lots and lots of other great stuff, too.

U-uh, haven't. Heard of Schumpeter, of course, even read a bio of him, I believe, and perhaps read a few paragraphs he wrote, but not the book you refer to. I'll go after it.

Capitalism ending out of its own success does make sense, obviously. And it probably makes sense to Neocons as well -- that's why they are trying to make capitalism less successful, for a while at least, and they hope to figure out, in the mean time, a way to keep their privileges in what follows capitalism.

What outfit was it that Putin working for before glasnost and perestroika?

I follows lazy tactics again! Instead of obtaining and reading the book, I reads comments on it. I found this "CSandD":

http://www.phil-books.com/Capitalism_Socialism_and_Democracy_0061330086.html

And, ho!Look what it reads:

"...One of Schumpeters main points for the end of capitalism is the decline of need for the entrepreneur...."

That stands 180 degrees contrary to what I envisage: Building upon market economy and free enterprise system...

But probably, very probably, in fact, necessarily, a different breed of entrepreneurs...

That's customer review titled "Dated and fatally flawed, but sweeping", via link in my previous post above.

Ho! Somebody even gave a name to it: the "Schumpeterian tension" -- the tension between the Schumpeter who comes to praise entrepreneurship and the Schumpeter who comes to bury it ... that somebody is a Richard. N. Langlois (of University of Connecticut in 1987)

Link:

http://www.ucc.uconn.edu/~LANGLOIS/SCHUMPET.HTML

Imitation is the sincerest form of television.

Even a philosopher gets upset with a toothache.

Just because there's a pattern doesn't mean there's a purpose.

In his errors a man is true to type. Observe the errors and you will know the man.

Keep the good work.