UC Berkeley

This dissertation improves our understanding of how the actions of the Federal Reserve and other central banks affect real activity in the economy. Chapters one and two show how secular change in the composition of investment spending in the United States has weakened the ability of conventional monetary policy tools (which alter the federal funds rate) to affect labor income and consumption. Specifically, Chapter one uses data on software developers collected from GitHub to show that firms are slow to adjust R&D activity and other so-called ''intangible investment'' in response to changes in interest rates because of congestion in onboarding the workers who produce it. Chapter two takes this result seriously as the explanation for the low observed responsiveness in the cross section for intangible investment as compared to tangible investment. Given the shift towards more and more intangible investment in the U.S. economy, this implies that investment spending overall is becoming less sensitive to changes in interest rates. Combined with two other secular changes documented in Chapter two---a rising import share in investment spending and a decline in the labor share of domestically produced investment---this implies that consumption and labor incomes for hand-to-mouth agents (and thus consumption and labor incomes overall) respond less to monetary policy shocks in general equilibrium. Finally, Chapter three uses novel historical data collected from the Bank of England to show that sterilized foreign exchange intervention (interventions in currency markets that hold policy rates fixed) can affect exchange rates.