/Measuring-Technological-Innovation-Over-the-Long-Run-Replication-Kit

This repository provides the replication code for the analysis results in Kelly, B., Papanikolaou, D., Seru, A. and Taddy, M., 2021. American Economic Review: Insights.

Primary LanguageMATLAB

Replication Code and Data for Kelly, B., Papanikolaou, D., Seru, A. and Taddy, M., 2021 AERI

This package provides the replication code for the main results in Kelly, B., Papanikolaou, D., Seru, A. and Taddy, M., 2021. Measuring Technological Innovation Over the Long Run. American Economic Review: Insights. The paper is available at https://www.aeaweb.org/articles?id=10.1257/aeri.20190499.

The folder ./code contains all programs, while the folder ./input_data includes all needed input data files.

Code

The folder ./code includes code files of two types of programs: Stata and Matlab.

Preliminaries:

Please complete the following steps in this order:

  1. The file winsorizeJ.ado is program which allows for winsorizing the data by groups (here, year). It should be copied to your personal ado directory.

  2. The file CreateMergedData.do creates a merged dataset that is needed to run the patent-level regressions.

  3. The file IdentifyBreakthroughPatents.do identifies a set of breakthrough patents and outputs the data to an intermediate file named "Breakthrough_Patents.dta".

  4. The files Create_Innovation_Indices_NAICSx.do take the input above and create the time-series indices at the Naics x-digit level. Note: the 5-digit and 6-digit files use the 1997 NAICS definitions, to better match to Kendrick's definition of industries.

  5. The file Create_Innovation_Indices_Agg.do creates aggregate innovation indices.

Paper Figures and Tables:

  1. To create Figure 1, run the matlab file CreatePlotsSimilarityCites.m

  2. To create Figure 3A and 3B and Tables A2 and A3:

    a. Run the files patent_level_figs_CITES.do and patent_level_figs_KPSS.do for the figures

    b. Run the files patent_level_regressions_CITES.do and patent_level_regressions_KPSS.do for the regressions

  3. To create Figure 4, run the matlab file TSaggplot.m

  4. To create Figure A4 and A5, do the following in sequence:

    a. Run the file Innovation_Productivity_NAICS4.do for the industry productivity regressions

    b. Run the matlab file TimeSeriesRegressions_Industry_MatlabFigs.m to generate tikz plots for Figures A.4 and A.5 in the Appendix

  5. To create Figure 5, do the following in sequence:

    a. Run the file CreateIndustryPlots.do to aggregate the industries at a relatively coarse level; it generates as outputs "Fig_IndInnovationLR.csv"

    b. Run the matlab file TSindustryplot.m to generate Figure 5 in the paper

  6. To create Table A.1 and Figure A.1, follow these steps in sequence:

    a. Run the file HistoricallyImportantPatents_List.do to generate an intermediate output file named "important_patents_list.csv" which contains the list of patents in Table A.1

    b. Run the matlab file Hist_Important_Patents_Fig.m to generate Figure A.1 in the Appendix

  7. To generate Figure A.2, run the file BreakthroughInnovation_by_TechClass.do

  8. To generate Figure A.3, running the file Aggregate_productivity_regs.do first and then run the matlab file TimeSeriesRegressions_MatlabFigs.m

Contact

Please contact Dimitris Papanikolaou (d-papanikolaou@kellogg.northwestern.edu) or Amit Seru (aseru@stanford.edu) for any questions regarding the codes or data.

Please see the paper for more information on the codes and data. If you use these codes files or data, please CITE this paper as the source.