Asymmetrical Information

From the desk of Mindles H. Dreck:

Productivity and Technology

As the recession that started last year unfolds, mostly predictable patterns have emerged in offical economic data. Jobless claims are up and the inflation, with the exception of medical costs and some tobacco and energy-related anomalies, is under control.

Productivity, defined as "output per hours worked", usually decreases dramatically during recession quarters. Two quarters of decreased output is a recession, so the numerator is going down. Usually, job or hours cuts lag and measured productivity dips before recovering with output (GDP). But the third quarter productivity gorwth statistics released in December were unexpectedly strong at 1.5% in the non-farm business sector overall, and 2.5% in manufacturing.

Productivity growth is critical. Any increase in productivity translates one-for-one into the inflation-adjusted ("real") growth potential of the economy. Every dollar of increased output that does not require a dollar of labor inputs is a "pressure-less" increase in growth. As I and many others have said, the solution to our fiscal and social security issues relies entirely on the highest possible rate of sustained real economic growth (not lockboxes or budget austerity in recession,Daschlenomics or backward-looking tax rebates). Furthermore, productivity growth also adds to our competitiveness with other economies, so it can have a positive impact on trade imbalances and job opportunities in the U.S. For those reasons, we watch productivity carefully, and debate it often.

From 1995-2000 the United States recorded productivity gains of 2.5% per annum, much higher than the prior long-term (1972-1995)average of 1.4%, and dramtically higher than the 1987-1995 average of 0.9%. Economists have sharpened their slide rules and entered a pitched battle to decide whether this was a one time phenomenon or a secular change in our economy. As you can imagine, the release of counter-cyclical figures on December 6 has all their tongues wagging.

"New Economy" types like to pin productivity growth on the application of information technology. Their critics say it was entirely a by-product of a stock-market boom, spiking demand for PCs due to the internet valuation bubble, or a problem measuring hours worked in the growing service sector. Nonetheless, the productivity numbers have been sticky even as the market is down 30% and internet executives can only raise money by stripping naked.

Enter the McKinsey Global Institute, known for funding exhaustive studies on these topics. McKinsey's response to the New Economy debate is "a little of both." But what's really interesting is their research defining the sector attribution of productivity growth - which parts of the economy were responsible for this economic gift we call "productivity growth"?

McKinsey's work suggests that only six sectors of the economy, Wholesale, Retail, Securities/commodities brokers, Electronic and electric equipment, Industrial Machinery and Equipment, and Telecommunications, accounted for all of the incremental productivity growth increase in the 1995-2000 period over the 1987-1995 rate. The other 53 sectors of the economy added nothing cumulatively and very little individually.

McKinsey examines the relationship between IT spending and productivity growth across all sectors and finds...none. Even in the magic six sectors they find an insignificant correlation. In the securities and finance businesses they correctly identify market valuation as accounting for half of the productivity growth (our revenues, in almost all businesses, grow directly with market valuation). Furthermore, the sectors that had the highest contribution are the largest employers (retail and restaurants), so any incremental productivity gain in these industries will have a large effect on national productivity.

The technologies McKinsey indentifies as having a major impact on are essentially 1980s technologies - warehouse distribution, electronic data interchange, and the like. The actual hardware for these techniques is not sophisticated and has become very cheap. Implementing these technologies requires not only buying them, but changing your business processes around them to reap the benefits. McKinsey estimates these sectors have further to go, as they estimate only 25% penetration of modern inventory and information management technologies in wholesale and retail.

The other major factor they identify is competition as a catalyst. Each of these sectors had major new, fast-growing entrants with radically different business models. Examples include Schwab and E-Trade, Dell, and AMD. The ultimate example of combining technology, process change and competitive stimulus is Wal-Mart, of course.

Which brings me to my point (at last, you say). I am much more impressed with the internet as a technology now than I was in 1999 and 2000, the years of peak investment. In my opinion, my industry is just beginning to reap the benefits of connectivity. Our productivity requires the rapid movement, filtering and analysis of information. While we have had this kind of information flow from most of the markets for some time, we are barely scratching the surface in customer contact. In my own business, it has only recently become economically worthwhile to use web-based technologies to deliver highly customized interactive information. A simple example -we use SAS technologies to slice up insurance industry data and deliver it in interactive graphical format to clients and prospects. This process used to take weeks and was delivered in a printed publication. Now it takes hours, and the user can easily filter the data that interests them.

In general, despite the "dot.bomb" critics, my sense is that many internet and other IT innovations are only beginning to deliver the goods. In 1990, one could have looked at all the innovations in warehouse technology (bar coding, etc. - remember Symbol Technologies?) and claimed it was all hype. These technologies only made their contribution starting in 1995.

There is a benefit curve in technology adoption. It may have been overstimated in 2000 internet valuations, and the current content on the internet may still be, like this site, a labor of love. But I still think we are just beginning to ride the curve.

Posted by Mindles H. Dreck at January 11, 2002 09:02 AM | Technorati inbound links