PUB - Publikationen an der Universität Bielefeld

This data publication contains all data and R-code used for the paper: #### Hötte, Kerstin (2023): "Demand-pull, technology-push, and the direction of technological change".
https://doi.org/10.1016/j.respol.2023.104740 ### Abstract: This paper studies the impact of Demand-pull (DP) and Technology-push (TP) on growth, innovation, and the factor bias of technological change in a two-layer network of input-output (market) and patent citation (innovation) links among 307 6-digit US manufacturing industries in 1977-2012. Two types of TP and DP are distinguished: (1) DP and TP are between-layer spillovers when market demand shocks pull innovation and innovation pushes market growth. (2) Within-layer DP arises if downstream users trigger upstream innovation and growth, while TP effects spill over from up- to downstream industries. The results support between- and within-layer TP: Innovation spillovers from upstream industries drive market growth and innovation. Within the market, upstream supply shocks stimulate growth, but this effect differs across industries. DP is not supported but shows a factor bias favoring labor, while TP comes with a shift towards non-production work. The results are strongest after the 2000s and shed light on the drivers of recent technological change and its factor bias. -------------------------------------------------------------------------------- This data publication offers all material to reproduce the data and the results. Please do also consult the paper and the online supplementary material for a documentation of the data.
The core data in this publication is empirical data on two coupled network layers inferred from cross-industrial citation links (patent citation network) and input-output flows among industries (input-output (IO) network). These data is available in quinquennial time steps for the years 1977, 1982, 1987, 1992, 1997, 2002, 2006, 2012.
The core data is available at the 6-digit level, but the code coming with this data publication provides tools for aggregation and all scripts used for the statistical analyses can be run on more aggregate data (no warranty).
The results presented in the paper rely on 6-digit level data and a balanced panel of manufacturing industries. Principally, all data is also available for non-manufacturing, but the coverage of non-manufacturing sectors in the patent data is poor, which is due to the concordance table by Goldschlag et al. 2020 (doi: 10.1080/10438599.2019.1648014) that almost exclusively maps to manufacturing.
The 6-digit level data on manufacturing industries is supplemented by data from the NBER-CES productivity database: https://www.nber.org/research/data/nber-ces-manufacturing-industry-database [accessed 13/01/2023]. This data publication does also contain the raw data which allows the full reproduction of all steps of data processing and analysis. -------------------------------------------------------------------------------- ### Steps to reproduce the results in brief ### To reproduce the results and the data from the raw data, you must run the code provided in the following order: 1. CREATING THE DATA:
(a) All required code and data to produce the patent and input-output network are available in the subfolders "patent_data_R_files", and "io_data_R_files", respectively.
(b) The folder "R_scripts_both/create_merged_panel" provides all code needed to create the merged panel data and the network measures (centrality, spillovers) used in the analysis. 2. REPRODUCING THE ANALYSES:
The folders "R_script_both/descriptives" and "R_script_both/regressions" provide all code needed to reproduce the figures, tables, statistics and regression analyses, and provides all material to conduct robustness tests using other model and data specifications.
Further detail is provided below and in the "read.me" files in the respective folders.
This data publication also provides additional results on the analyses at different levels of data aggregation. Some of these results are provided in the folder "statistical_output" but you can also produce additional results running the code provided. ---------------------------------------------------------------------------------------------- ### This data publication consists of 5 folders: ### (1) patent_data_R_files
(2) io_data_R_files
(3) R_scripts_both
(4) data_combined
(5) statistical_output
---------------------------------------------------------------------------------------------- ### Details: ### ---------------------------------------------------------------------------------------------- #### (1) "patent_data_R_files" This folder contains 2 subfolders: "code", "data" - "code": This subfolder contains the R-scripts of all single steps executed to process the patent raw data. These steps are explained in detail in the Supplementary Material of the paper Hötte, K (2023): "Demand-pull, technology-push, and the direction of technological change [forthcoming]". - "data": This subfolder contains the processed data at different levels of aggregation and a folder with the raw data. Please do check the "patents_read.me" for more information and licenses/references. ---------------------------------------------------------------------------------------------- #### (2) "io_data_R_files" This folder contains 3 subfolders: "code", "data" - "code": This subfolder contains the R-scripts used to compile the IO data. The single steps are explained in detail in the Supplementary Material of the paper Hötte, K (2023): "Demand-pull, technology-push and the direction of technological change [forthcoming]". - "data": This subfolder contains the processed for different aggregation levels and the raw data downloaded from the Bureau of Economic Analysis (BEA) websites. Please do check the "input-output_read.me" for more information and licenses/references. ---------------------------------------------------------------------------------------------- #### (3) R_scripts_both This folder contains 1 file ("settings_data.r") and 3 subfolders ("create_merged_panel", "descriptives", "regressions"): - "settings_data.r": Here you can specify the data settings for the analyses. In the provided version, the default specification such that it can reproduce the results presented in the paper (6-digit, manufacturing, non-normalized, etc). This script also sets the output directory and prefixes to tag the output files. Please check the script itself for further detail and explanations. - "create_merged_panel": This subfolder contains all scripts to compile the merged panel data from the data compiled before (using the code in "patent_data_R_files", "io_data_R_files", and having pre-processed the raw data from other sources (see "data_combined/other_data")). To create the industry level panel data, you have to run the "main_create_data.r" script. It sources functions and subscripts from the "auxx" folder. - "descriptives": This subfolder contains all scripts needed to reproduce the descriptive statistics and exports them to latex tables. It also creates all figures. To produces these statistics, you need to run the "main_descriptives.r" script, which sources the scripts from the "src" and in the "auxx" folder. - "regressions": This folder contains all scripts needed to reproduce the regression analyses. It additionally contains scripts that were used for purposes of data exploration and model selection (e.g. different model specifications, sets of explanatory variables, specification of spillovers, measurement of network centrality). To reproduce the results, you have to run the "main_regression.r" script. Additional information about the different possibilities to run the code are provided in the script. The script generates RData objects with the results which are saved in the folder "statistical_output". The outputs are objects with the results of multiple model specifications. To create the latex tables with the results provided in the paper, run the code in the folder "process_output". This subfolder contains additional scripts that may be used to process and explore the results (including alternative specifications for robustness checks). ---------------------------------------------------------------------------------------------- #### (4) "data_combined" This folder contains the merged data at different aggregation levels. The data can be created from the processed data by running the code in "create_merged_panel" explained above. The data includes: industry panel data, technological similarity networks, data on link formation processes, data on spillovers and a sample of network metrics. Some of these data are merged in the regression analyses scripts.
Please consult the "data-combined_read.me" file for additional information on the data and the subfolders containing data from external sources and the references to be cited. ---------------------------------------------------------------------------------------------- #### (5) statistical_output This folder contains the results presented in the paper and additional results for other aggregation levels and data subsets. These results are based on the original data (see disclaimer above). These data in this folder will be overwritten when you run the R-scripts for the regressions and descriptive statistics. ----------------------------------------------------------------------------------------------- ### LICENSE: ### This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.
To give appropriate credit, please cite the most recent version of the paper:
Hötte, Kerstin (2023): "Demand-pull, technology-push, and the direction of technological change". https://doi.org/10.1016/j.respol.2023.104740