Is the Information Technology Revolution Over?

By David M. Byrne, Stephen D. Oliner, and Daniel E. Sichel

2013 ASSA MeetingsDiscussant: Chad Syverson

Overview

Paper’s analysis has three main parts:

1. Decomposition of LP growth into IT and non-IT components and comparison between 1995-2004 and 2004-2012

2. Prediction of steady state LP growth going forward

3. Re-assessment of potential IT-driven growth in light of evidence from semiconductor manufacturing

1. LP Decomposition: Is a Productivity (Re-) Slowdown Historically Unprecedented?

The prior GPT diffusion event was the electrification (& plumbing) era

Medium-run LP trends were not altogether different at that time

1. LP Decomposition: Is a Productivity (Re-) Slowdown Historically Unprecedented? No.

1890 1895 1900 1905 1910 1915 1920 1925 1930 1935 1940

1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

40

60

80

100

120

140

160

180

Labor Productivity, Electrification and IT Eras(1925 = 100 and 1995 = 100)

ElectrificationIT

1. LP Decomposition: Is a Productivity (Re-) Slowdown Historically Unprecedented? No.

1890 1895 1900 1905 1910 1915 1920 1925 1930 1935 1940

1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

40

60

80

100

120

140

160

180

Labor Productivity, Electrification and IT Eras(1925 = 100 and 1995 = 100)

ElectrificationElectrification (cont.)IT

1. LP Decomposition: Is a Productivity (Re-) Slowdown Historically Unprecedented? No.

Period Annual Avg. LP Growth (%)

Period Annual Avg. LP Growth (%)

1896-1915 1.0 1976-1995 1.4

1915-1924 3.3 1995-2004 3.1

1924-1932 1.0 2004-2012 1.6

1932-1940 2.7 2012- ?

2. Predicted Steady-State Future LP Growth

Forecasting is hard

Average future growth rates are hugely important

But recent patterns have convinced me median growth matters a lot too

Forget normative issues—even positive implications will be inescapable if pattern of past 3 decades continues

2. Predicted Steady-State Future LP Growth

1950 1960 1970 1980 1990 2000 20100

50

100

150

200

250

300

350

400

Real Labor Productivity Growth vs. Real Median HH Income (1950 = 100)

Labor ProductivityMedian HH Inc

3. Intel and Semiconductor Innovation: A More Pessimistic View

Authors: semiconductor mfg. technology cycle hasn’t slowed •Lithography process generation has held at ~2 years since 1993•Intel’s CPU cycle has held up similarly

However, Pillai (forthcoming): while transistors per unit area have continued to increase, chip designers have not been able to fully harness these improvements

NYT, 9/1/09: “The computer industry has a secret. Yes, the number of transistors on modern microprocessors continues to multiply geometrically, but no one really knows how to get the most out of all these new transistors.”

3. Intel and Semiconductor Innovation: A More Pessimistic View

Wrap-Up

Thought-provoking paper: critical issues at hand