On Friday, the very fine economists at the Bureau of Economic Analysis will release their estimate of fourth quarter GDP growth. Current estimates peg the US economy’s growth at roughly 3% for the quarter.
But here’s the rub: The rise of the data-driven economy means government economic statistics may significantly understate US GDP growth and productivity growth. Official numbers are afflicted by huge and growing blind spots that increasingly distort the published figures. To summarize, we are building a mammoth data-driven economy that, perversely, is only partly visible in the economic data.
To give just one simple example: As of January 29, the official statistics report that the real value of of Internet access consumed by households has fallen by 5% over the past year. The official statistics also report that real value of cable and satellite television and radio services has fallen by 2% over the same stretch. And supposedly the real consumer value to households, as measured by the government, of mobile, cable, and internet access together has risen by a measly 0.4% over the past year.
These numbers can’t be right (for more of the theory here, see PPI’s 2012 paper “Beyond Goods and Services:The (Unmeasured) Rise of the Data-Driven Economy” )
Or to give another example, private investment in big data. All sorts of organization are building up huge data stores with long-term value. For example, the shift to electronic health records is predicated on the value of that data for lowering health care costs and improving patient treatments. (see, for example http://www.healthit.gov/providers-professionals/benefits-electronic-health-records-ehrs).
In theory, the investment in big data should be reported as part of GDP. Indeed, the BEA has recently started reporting spending on R&D and “entertainment, literary, and artistic originals” as part of investment spending. And the original researchers on intangible investment did in fact include investment in databases.
However, in practice, the BEA does not include investment in big data in GDP: The tech equipment and the programming, yes, but not the actual labor and costs necessary to collect and clean the data. For example, when a hospital employs medical coders to clean up their electronic patient records, that coder’s salary is recorded as an expense, but not as a contribution to GDP. Similarly, the costs of converting from paper to electronic records is not being counted a part of GDP.
The distortion in the statistics from omitting big data is becoming bigger as big data becomes more important. I cite health care because health care organizations are devoting vast resources to electronic health records, but the same holds for any company collecting big data.
We can list example after example where the data-driven economy is simply missed by the current statistics. An earlier PPI paper, Data, Trade and Growth, showed that the government does a terrible job measuring cross-border data flows, because many of them do not leave a monetary footprint. to the extent that the US holds a commanding position in cross-border data trade, this omission may be important for GDP and productivity growth.
Finally, it’s worth noting that reshoring may be artificially depressing the growth and productivity statistics, just as offshoring artificially inflated growth and productivity gains in the early part of the 2000s (this give me a chance to plug a new conference volume edited by myself and Susan Houseman, entitled “Measuring Globalization: Better Trade Statistics for Better Policy“). I will address this point at length in a future post.