Does the Productivity Paradox really exist?
On Sunday, July 12, 1987, in the New York Times’ Book Review, Robert
Solow wrote a review of a book called “Manufacturing Matters” by
Stephen Cohen and John Zysman. Back then, in 1987, the world looked much like
it looks today. Many manufacturing jobs were moving overseas as companies sought
less expensive hourly labor, and a popular topic of debate was whether the “service
economy” could grow to provide jobs for recently displaced American factory
workers.
Mr. Solow panned the book, and its authors’ premise that computers would
replace manufacturing as the powerhouse of American productivity. In summing
up his assessment of the book, Mr. Solow writes a sentence that has been very
frequently quoted: “You can see the computer age everywhere but in the
productivity statistics.”
Mr. Solow’s book review, and in particular his “productivity paradox”
has become part of our economic lexicon. An internet search using the Google
search engine on the phrase “productivity paradox” returns not fewer
than 10,000 references to this term. Many of these references are scholarly
works.
It’s hard to imagine that there’s any debate at all about whether
or not computers aid productivity. In this year of 2003 we are surrounded by
very productive computers and computer systems.
Computers power the sophisticated load-balancing and tracking systems that
make mobile telephones possible. Microwave ovens with embedded computer chips
make it possible to speed ever faster through the drive-through at the local
fast-food burger joint. Painting and welding robots powered by artificial intelligence
software enable automotive plants to build cars faster with fewer workers, and
sophisticated just-in-time delivery systems have enabled giant retailers such
as Walmart and Best Buy to sell merchandise across the country at only the tiniest
of profit margins.
When we look more closely, we see that there is a cost. To create and maintain
all this technology requires labor. Stories of massive system failures abound.
Computer viruses bring down entire companies for days at a time. Attempts to
implement complex software systems are abandoned, at a cost often in excess
of tens of millions of dollars. Is it possible that the costs of computer technology
cancel out its gains?
To answer this question, it is useful to compare computer/information technology
against other technologies that have been revolutionary in the past. Nicholas
Crafts, of the London School of Economics, does just this. In his paper, “The
Solow Productivity Paradox in Historical Perspective”, Mr. Crafts compares
improvements in productivity brought about by computer/information technology
against those brought about by steam and electricity, respectively.
Mr. Crafts finds, contrary to Mr. Solow’s assertion, that computer/information
technology is indeed increasing our national productivity. It is, in fact doing
so at a comparatively rapid pace, and that its effects would be much more apparent
were not the world economy so much larger now than it was 100 or 200 years ago.
Mr. Crafts’ conclusion is that “the Solow productivity paradox stems
largely from unrealistic expectations.”
Another possibility is that computers have in fact greatly increased productivity,
but that the statistics used to measure productivity are not able to accurately
account for the increase. Eric Brynjolfsson of the MIT Sloan School of Management
and Lorin Hitt of the Wharton School of Business look at this possibility in
their paper, “Beyond the Productivity Paradox”.
Mr. Brynjolfsson and Ms. Hitt note that computers do not simply make existing
tasks quicker and easier to perform; they make it possible to perform new types
of tasks, and many of those new types of tasks involve increases in convenience
that are not measured by conventional productivity statistics.
The conclusion arrived at by Mr. Brynjolfsson and Ms. Hitt is similar to that
of Mr. Crafts. There is a caveat, they note. Computer/information technology
is not a panacea. If the implementation of information technology is not also
combined with a realignment of workflow in an organization, then information
technology can actually result in reduced productivity. In the end, they too
find that computer/information technology is increasing our national productivity.
I’d like to raise a third, related possibility. In a talk given before
the American Economic Association on January 3, 2003, Stanley Fischer, President
of Citigroup International presented a paper titled “Globalization and
Its Challenges”. In this paper, Mr. Fischer covers many interesting topics
related to globalization. One in particular is of particular relevance to measurement
of the increase in productivity brought about by information technology.
Mr. Fischer notes that since 1987 there has been a significant decrease in
the global poverty rate as measured by the World Bank. In China and India, which
together represent a large percentage of the world population, per-capita income
has been rising at a rate almost double that of the US or Britain.
Is it possible, I wonder, that wide-scale implementation of information technology
has in fact greatly increased productivity—but the productivity not of
the US, but of the vast, underdeveloped Asian region. Information technology
is making it possible to move jobs from the US to China and India; to where
those jobs are beyond the reach of the productivity statistics.
References
Solow, R.M (1987), “We’d Better Watch Out,”
New York Times Book Review, July 12, 1987.
Crafts, Nicholas (2001). The Solow Productivity Paradox in Historical
Perspective. Centre for Economic Policy Research, Research Papers, November
2001.
Brynjolfsson, Eric and Hitt, Lorin M. (1998). Beyond the Productivity
Paradox: Computers are the Catalyst for Bigger Changes. Communications of the
ACM. August 1998
Fischer, Stanley (2003). Globalization and Its Challenges. Ely
Lecture, American Economic Association, January 2003.