The determinants of innovation: New evidence from nineteenth-century world fairs

Innovation is a key determinant of economic growth. But what determines innovation? This question has proven difficult to answer because innovation is difficult to measure, especially across countries. This dissertation introduces a new internationally comparable dataset on innovation. I have constructed such data from the catalogues of two nineteenth-century world fairs: the Crystal Palace Exhibition in London in 1851 and the American Centennial Exhibition in Philadelphia in 1876. I use these data to address three questions: What do inventors patent? How do patent laws influence innovation? And what determines the location of innovation?; Patent data have been criticized as a measure of innovation because only a fraction of innovations are patented. Patents measure innovation accurately, however, if innovations are patented at a constant rate. I use the exhibition data to compare patenting rates across industries and geographic locations. My analysis shows that patenting rates are almost identical across rural and urban areas, but that they vary significantly across industries. Patenting rates are especially low in industries where innovations are difficult to reverse-engineer and can be easily kept secret. Innovations in these industries appear to be less dependent on patent protection.; Patent laws are designed to create incentives for innovation. Yet we know little about how exactly this works. Using exhibition data, I find that strong patent laws may fail to raise the number of innovations, but they do influence the distribution of innovative activity across industries. The absence of patent laws guides innovative activity towards industries where mechanisms other than patent laws protect intellectual property. Inventors in countries without patent laws concentrate on industries where secrecy is effective. In the nineteenth century, secrecy was especially effective in scientific instruments. Every fourth exhibit from countries without patent laws comes from that industry while no more than one seventh of other countries' innovations are scientific instruments.; Exhibition data also allow me to study the geographic location and the quality of innovation. Recent economic research emphasizes the influence of localized knowledge spillovers on innovation. If knowledge spillovers exist, innovators should concentrate locally and the quality of innovation should be highest within concentrated locations. Analyses of exhibition data confirm that innovation clusters locally and that the quality of innovation is highest within clusters of innovative activity. Concentration is most pronounced in industries where innovations are complex and can be protected through secrecy rather than patents.