http://www.federalreserve.gov/BoardDocs/Speeches/2000/20000306.htm

''from the above speech ...''

In the last few years it has become increasingly clear that this business cycle differs in a very profound way from the many other cycles that have characterized post-World War II America. Not only has the expansion achieved record length, but it has done so with economic growth far stronger than expected. Most remarkably, inflation has remained largely subdued in the face of labor markets tighter than any we have experienced in a generation. 

A key factor behind this extremely favorable performance has been the resurgence in productivity growth. Since 1995, output per hour in the non-financial corporate sector has increased at an average annual rate of 3-1/2 percent, nearly double the average pace over the preceding quarter-century. Indeed, the rate of growth appears to have been rising throughout the period. 

My remarks today will focus both on what is evidently the source of this spectacular performance -- the revolution in information technology -- and on its implications for key government policies.

----

''What amazes me - despite everyone typically has jobs solving problems caused only by computers these days - is that per capita productivity is actually up, after a long stagnant period. Someone's doing something right. -- PhilipCraigPlumlee''

It amazes me rather more how long it's taken for IT to show through in this way. MooresLaw has been doing its extraordinary stuff for how many decades now? The wonder is that the benefit has filtered through so sporadically. Is it the rise of really broad de facto standards - Wintel and TCP/IP/Web especially - that has begun to burn the fat off the older "technology providers" or "innovation inhibitors", including in telecoms, allowing some real benefit to be seen by the "masses"? And made just a few companies with proprietary technology right in the middle of the current interlocking de facto standards much, much fatter ... -- RichardDrake

''You remember the idiotic experiments The Telecom (AT&T) did in the 1970s to invent a "Video Phone"? The ones that never went anywhere? We have them now. And the only way we got them was the most round-about path imaginable. We invented one simple minor system at a time, and deployed it until it achieved critical mass (and a slight productivity boost). First modems then Web pages then streaming media then Web Cams. -- PhilipCraigPlumlee''

Hmm ... are we saying IETF style ''working standards'' + ''one step at a time'' = real productivity? Thanks for this point.
----
Productivity as measured by standard economics is mostly B.S. anyway. Two reasons:

	1 It measures productivity as GNP / full-time jobs, not GNP / man-hours worked. Which means if companies start forcing their employees to work uncompensated overtime, their productivity will magically appear to rise, when all that happened was that they squeezed more hours out of their labor force.
	1 Productivity uses a thing called "hedonic pricing" to try to account for quality of goods produced. So, if this year the car you make is 5% more reliable than the car you made last year, and you didn't spend any more hours doing it, your productivity has increased by 5%. Sensible in theory, bloody hard in practice. It's worth noting that most of the productivity gains in the U.S. economy in the last economy came out of the tech sector, because of hedonic pricing being applied to things like telecom switches and microchips. If your CPU were ten times as fast, would it be worth paying ten times as much? Does value scale with certain metrics that easily? Or are economists just taking the best numbers they can get and covering up the uncertainty of their results?
----
CategoryEconomics