Innovation and Productivity

Firms make large investments in intangible capital, which drive innovative activities and market values. These investments impact productivity but are difficult to measure and thus mislead researchers who restrict attention to only those assets in the BEA fixed asset tables. For example, in previous work with Ed Prescott ("Labor Productivity Puzzle"), we showed that measured productivity will be overstated in recessionary periods when there are large declines in intangible investments because measured GDP does not fall by as much as total output. This is important because the rise in productivity during 2008--2009 has led many economists to the natural conclusion that financial disruptions---not changes in firm productivity---are the primary cause of the large fluctuations in real activity. While our previous work has shown that there is no logical inconsistency with aggregate observations on GDP, hours, and labor productivity, my current work considers both the macro and micro data, extending theory to include both intangible investments (as in McGrattan and Prescott) and sectoral input-output linkages (as in earlier work by Long and Plosser, 1983). There is microevidence showing that intangible investments are large, especially for the high-technology industries that have important input-output linkages with other industries. They are also large for U.S. multinationals, implying both domestic and foreign linkages. The primary goal is to assess the theoretical predictions for post-war business cycles and ask, how much of the variation in GDP, investment, and hours can plausibly be attributed to changes in firm-level and industry-wide TFPs? This work will also be useful in a second project that extends the analysis of Holmes, McGrattan, and Prescott ("Quid Pro Quo: Technology capital transfer for market access in China," forthcoming ReStud) who study the impact of quid pro quo policies governing foreign direct investment (FDI) into China in a multicountry model with technology capital (e.g., nonrival intangible capital). The model has an aggregate technology with a single consumption good and thus no scope for "wool for wine" trade. As a conceptual matter, it is straightforward to generalize the model to include multiple industries, generating trade in consumption goods based on comparative advantage. Such an extension would allow for industry-specific intangible capital and for technology and policy parameters to vary across industries. The goal here is to study regulation of FDI---which varies by industry---and its impact on global innovation and capital flows.