Innovation

This analysis relates standardized firm-year MQ to patent counts, citations, and real patent value.

flowchart LR
    panel[(Public-firm analysis dataset)]
    prepare[[Filter, transform outcomes,<br/>winsorize, and standardize MQ]]
    estimate[[Estimate OLS and Poisson grids]]
    csv[(table9_innovation.csv)]
    patents[/table9_patent_count.tex/]
    cites[/table9_cites_weighted.tex/]
    value[/table9_value_weighted.tex/]

    panel --> prepare --> estimate --> csv
    csv --> patents
    csv --> cites
    csv --> value

    classDef input fill:#eee5f5,stroke:#76528c,stroke-width:2px
    classDef code fill:#f3f3f3,stroke:#666,stroke-width:2px
    classDef support fill:#fff0cc,stroke:#a66b00,stroke-width:2px
    classDef tableOutput fill:#e2f2e7,stroke:#3b7a4b,stroke-width:2px
    class panel input
    class prepare,estimate code
    class csv support
    class patents,cites,value tableOutput

The current implementation uses full firm-wide MQ, not the disconnected R&D-worker MQ panel. It estimates OLS models for log-one-plus patent counts, citations, and patent value, plus Poisson models for counts and citations. The industry, fixed-effect, clustering, and control grids match the firm-performance analysis.

Input: Public-firm analysis dataset Module: analyses/modules/table9_innovation.R

Rscript pipelines/run_table9_innovation.R \
  --config config/config_one_percent.yaml
Rscript pipelines/run_table9_innovation.R \
  --config config/config_tables89_full.yaml

/artifacts/<coverage>/analyses/table9_innovation/
  table9_innovation.csv
  table9_patent_count.tex
  table9_cites_weighted.tex
  table9_value_weighted.tex

Preliminary manuscript formatting

The module writes simple outcome-specific fragments and does not save fitted models or metadata. The output filenames retain legacy table numbering.