Dana Gardner, formerly with the Yankee Group and the Aberdeen Group, recently launched Interarbor Solutions, a consultancy focusing on enterprise applications, software infrastructure, RSS and other topics. Dana has agreed to post some of his insights on Between the Lines. Today, he takes issue with economists who forecast that IT can't possibly maintain significant productivity gains.
Economists are missing the trees due to the forest when it comes to IT productivity. I draw this conclusion from a New York Times story (registration required) this week.
The consensus in question, which apparently is helping to guide the U.S. Federal Reserve in its deliberations over short-term interest rate adjustments, is that IT is not going to achieve the productivity boost it did during the 1990s. The economists don't seem to think that the wonderful tango of rapidly lowering IT costs and rapidly increasing functionality and efficiency reach has any quick-step left for the next five years. IT, yawn ... will just sit this one out.
They are probably seriously wrong, which is really good news. First, the reasoning for why IT spending shot up over the past 10 years is flawed. The story says, "...many analysts say they believe that plunging technology prices were responsible -- along with the dot-com stock market bubble -- for a big part of the huge investment in [IT] equipment during the 1990s."
Nope, the relationship is that IT standardization around such key technologies as TCP/IP, Java, SQL, Ethernet, UNIX, SMTP, LDAP, Windows, and HTML spurred volume purchases, i.e. investment, of servers, PCs, routers, LAN/WANs, and applications. It was the volume and fierce competition within standardized (rather than closed) products that drove prices down, and then led to the speculative bubble (as if standards-oriented sellers could reap profits like monopolies). The engine of growth was fueled first by innovation and standardization, then volume, then lower prices, then more adoption, and more lower prices, which all leads to generally higher productivity. Let's get the order straight, and recognize that the process takes years (even at Internet time).
Now, the future is not so bright, we're told. "The rate of technical change, which is the most difficult thing to measure, seems to be slowing from the unprecedented pace of a few years ago," said Mark Zandi, chief economist at Economy.com, in the Times story.
Technical change is hard to measure if you only look at IT sales.
Allow me to point out a few gargantuan trees in this forest: open source tools and development frameworks; standardization around XML; open source middleware stacks, databases, and business applications; low-cost multi-core 64-bit x86 processors; hardware and software virtualization; even more dirt-cheap and super-fast storage; RSS; application lifecycle management methods and uniform modeling and testing; 64-bit file formats; grid and utility computing efficiencies, Java virtual machines running on silicon; advanced search and indexing; Web services standards; IPTV, cheap and ubiquitous land-line and mobile broadband; services-oriented architectures; composite applications; G3 handsets; software as a service; the Foveon camera; embedded devise development standards and embedded Linux; subscription-based purchasing models for IT products and services; and let's not forget outsourcing, off-shoring, and mixed-source purchasing.
Now most of these trends have yet to kick in sufficiently to significantly impact the IT investment and productivity cycle. But is there much doubt that they will? If we measure only the revenues of the IT vendors, we see mid-single-digit growth, and we may see some leveling of the price decreases for what are or are becoming commodities, which by definition resist price weakness below their established floors.
But when we examine the buy side, the enterprise, SMBs, SOHOs, and carrier-host organizations, we can expect a much different impact over the next five to 10 years. I'm no economist, but these IT trends and business model shift impacts will in no way slow down productivity growth. I believe they will accelerate it beyond what the 1990s offered, and without the waste of the stock market bubble and Y2K drag.
For the enterprises and business buyers, the total cost of computing operations and IT-driven business process innovation is going down in real terms, based on total productivity. This total cost-benefit relationship will improve a lot more for a long time. IT buyers are now getting a lot more bang for the buck, and a rapid return on investment. IT innovation is rampant, global, and easier to absorb than ever given all the qualified labor in locations far and wide.
Enterprises have a lot to bite off to enjoy these benefits, granted, but they have large coffers to begin the process. My research shows them eager.
"Over the past decade, the U.S. economy has benefited from a remarkable acceleration of productivity," Mr. Greenspan told the Senate Banking Committee last month. "But experience suggests that such rapid advances are unlikely to be maintained in an economy that has reached the cutting edge of technology."
We may be standing on the shoulders of giants, but we have not reached the cutting edge of technology. The IT vendors may have reached some turning point in terms of pricing and lock-in, but for those buying and using IT smartly -- let the good times roll, and roll, and roll, and roll.
So maybe, just maybe, the IT productivity effect on GDP higher-growth-at-low-inflation trend isn't over after all.