Research Policy

halo_rd and halo_educ are Stata files containing the data used for the regression tables; figrddata, fig3data, and fig4data were used to generate the figures in R/ggplot2. (rrrrcode.txt contains the replication code for tables and figures alike.)

The database is a novel and unbalanced panel of 657 startups in the U.S. cleantech sector between 2008 and 2012. The panel includes 2,015 alliances (2008 to 2012) of all U.S. startups interacting with other firms, government organizations, research institutes, universities, and other not-for-profit organizations from the U.S. and globally. The startups operate in one or more of the following 17 sub-sectors: solar, wind, biomass, geothermal, hydro and marine power, nuclear, energy storage, smart grid, energy efficiency, advanced materials, transportation, biofuels and biochemicals, conventional fuels, fuel cells and hydrogen, air, recycling and waste, or water and wastewater. We considered a firm to be a startup if it is younger than 5 years in year t, with t0 being 2008. In addition to the network data, the dataset includes information on the patenting activity and financial investments (number of financing deals) of the startups.

see detailed in data file

This is the main dataset that we used to test our hypotheses as stated in this paper.