HOUSE_OVERSIGHT_014548.jpg

It follows that if real GDP is understated, then what appears to be a slow recovery is not as slow as reported and what appears to be a period of low productivity growth is not as low as reported. We believe that the evidence favors the mismeasurement argument. At a September 2016 Brookings Institution conference on productivity, Martin Feldstein, Harvard professor and president emeritus of the National Bureau of Economic Research, also concluded that “the official statistics substantially underestimate the real growth of output” after studying the methods used to measure price indices.°° We point to three examples to illustrate the mismeasurement argument. First, our colleagues in Goldman Sachs’ Global Investment Research (GIR) have pointed out that the official price indices for information and communication technology show an implausible gap between the price deflation in computers and that in communications equipment, software and other IT equipment (see Exhibit 14). They question how “a given dollar outlay now buys about 10 times as much computer in real terms as 20 years ago, but it only buys about 10% more software.” Our colleagues’ conclusion that the official price indices for the information and communication technology sector are overstated matches that of a 2015 study of microprocessor pricing by David Byrne of the Federal Reserve Board, Professor Stephen Oliner of UCLA, and Sichel.** The trio created an index showing that prices for microprocessor units used in desktop personal computers declined by an average annual rate of 43% between 2008 and 2013, while the official Producer Price Index (PPI) for these units declined by an average annual rate of 8%—substantially mismeasuring the real value created by this sector of information technology equipment. They point out that because microprocessor units represent about half of US shipments of semiconductors, the rate of innovation in this sector is inevitably mismeasured. A second example of mismeasurement that we can all readily appreciate involves the quality and product improvements in smartphones. Hal Varian, chief economist at Google and emeritus professor at the University of California at Berkeley, has estimated that globally, people took over 1.6 trillion photos in 2015 using their smartphones, compared with 80 billion in 2000 using cameras and film. The price of each photo taken has gone Exhibit 14: US Technology Price Indices The implausible gap with hardware suggests IT and communication price indices are likely overstated. Q1 1995 = 100 140 Software “----- Hardware (Computers and Peripherals) 120 Communications Equipment Other IT Equipment 114 100 += PR \ 90 so * 60 ‘ 40 N 20 0 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Data through 2015. Note: Other IT Equipment represents medical and non-medical equipment and instruments. Source: Investment Strategy Group, Goldman Sachs Global Investment Research, Bureau of Economic Analysis. from 50 cents to zero for smartphone users; 1.6 trillion photos that would have contributed $800 billion to GDP have no impact on GDP in the current framework. GDP has declined since camera and film sales have fallen without a commensurate quality adjustment for smartphones. Of course, fewer photos would have been taken had the smartphone not been developed, but the point still stands. Similarly, Varian shows that with the onset of the commercial application of GPS technology, productivity growth in trucking was twice the aggregate US productivity growth, yet when GPS functionality was added to smartphones basically at no additional charge, GDP declined because sales of stand-alone GPS systems fell.*4 Finally, a third example, also provided by Varian, shows that, because GDP does not fully count the export of intangibles such as software and design, GDP is understated. He shows how an iPhone manufactured by Foxconn in China using parts from 28 countries and exported to France has no direct impact on US GDP. Varian concludes that in a global supply chain, US design and software that is replicated outside the US through offshore manufacturing and exported to a third country never impacts US GDP measures directly, particularly if the profits are not repatriated and redeployed in the US.* Our colleagues in GIR continue to estimate that such mismeasurements lower reported annual real Outlook | Investment Strategy Group 1S HOUSE_OVERSIGHT_014548