Symbolic Institutional Traps: Language Regimes, Legal Legacy, and Organizational Constraint in Postcolonial Economies

README: Symbolic Institutional Traps and the Liability of Foreignness Scott M. Brown (University of Puerto Rico)Email: [e-mail] DOI: 10.5281/zenodo.15050209 Overview This project empirically tests how language regimes embedded in legal and administrative systems create institutional traps that constrain multinational enterprise (MNE) operations and economic integration.The study combines national and subnational data across four key datasets to measure how symbolic misalignment (such as monolingualism in non-commercial languages) affects regulatory quality, business formation, and workforce access. 📂 Datasets You must upload the following four files into your Google Colab session before running the code: Uploaded File Description /content/2020_Rankings.xlsx World Bank Ease of Doing Business (EODB) — Global regulatory efficiency indicators (2020 Edition) /content/DBNA 2022 Rank and Scores.xlsx Doing Business North America (DBNA 2022) — City-level institutional performance across 83 U.S. cities /content/Spanish_Speakers_All_States.xlsx U.S. Census American Community Survey (ACS) — State-level Spanish-speaking and English proficiency data /content/wgidataset.xlsx World Governance Indicators (WGI) — Governance quality measures (Regulatory Quality, Government Effectiveness, etc.) 📋 How to Run the Study Open Google Colab. Upload the four Excel files listed above. Copy and paste the Python code provided below into a Colab notebook cell. Run the code to automatically load the datasets, clean the data, and estimate key regression models. 🚀 Required Python Code python # --- 0. Imports --- import pandas as pd import statsmodels.api as sm import statsmodels.formula.api as smf # --- 1. Load Clean Datasets --- dbna = pd.read_excel('/content/DBNA 2022 Rank and Scores.xlsx') acs = pd.read_excel('/content/Spanish_Speakers_All_States.xlsx') wgi = pd.read_excel('/content/wgidataset.xlsx') # Optional: Governance analysis # --- 2. Standardize Column Names --- dbna.columns = dbna.columns.str.strip().str.replace(' ', '_') acs.columns = acs.columns.str.strip().str.replace(' ', '_') wgi.columns = wgi.columns.str.strip().str.replace(' ', '_') # --- 3. Merge Datasets --- # Merge DBNA and ACS on 'State' merged_dbna = dbna.merge(acs, on='State', how='left') # --- 4. Regressions: Language vs Institutional Outcomes --- # H1: Language (% Spanish) and Starting a Business Score model1 = smf.ols('Starting_a_Business_Score ~ Percent_Spanish_Speakers', data=merged_dbna).fit() print("\nRegression: Starting a Business Score ~ Percent Spanish Speakers") print(model1.summary()) # H3: Language (% Spanish) and Land and Space Use Score model2 = smf.ols('Land_and_Space_Use_Score ~ Percent_Spanish_Speakers', data=merged_dbna).fit() print("\nRegression: Land and Space Use Score ~ Percent Spanish Speakers") print(model2.summary()) # H3: Language (% Spanish) and Getting Electricity Score model3 = smf.ols('Getting_Electricity_Score ~ Percent_Spanish_Speakers', data=merged_dbna).fit() print("\nRegression: Getting Electricity Score ~ Percent Spanish Speakers") print(model3.summary()) # H4: Language (% Spanish) and Employing Workers Score model4 = smf.ols('Employing_Workers_Score ~ Percent_Spanish_Speakers', data=merged_dbna).fit() print("\nRegression: Employing Workers Score ~ Percent Spanish Speakers") print(model4.summary()) # --- 5. (Optional) Governance Analysis: Percent Spanish vs. WGI Regulatory Quality --- # If WGI includes 'State' or 'Country' to merge, otherwise skip # Example assuming WGI has 'Country' to match 'State' #wgi_merged = wgi.merge(acs, left_on='Country', right_on='State', how='left') #model5 = smf.ols('Regulatory_Quality ~ Percent_Spanish_Speakers', data=wgi_merged).fit() #print("\nRegression: Regulatory Quality ~ Percent Spanish Speakers") #print(model5.summary()) # --- 6. End --- print("\nAll regressions completed.") 🧠 Key Concepts Symbolic Institutional Traps: Language regimes act as hidden barriers, complicating regulatory navigation and labor market integration. Symbolic Misalignment: Misfit between administrative languages and global commercial norms raises onboarding costs for MNEs. Institutional Friction: Language encapsulation isolates economies and reduces foreign direct investment (FDI) attractiveness. 📜 Data Documentation Each dataset has been: Cleaned for consistent formatting. Harmonized for cross-dataset integration. Standardized to facilitate reproducible econometric analysis. Full codebooks and metadata are available in the appendix of the research paper. ⚡ Notes The EF EPI (English Proficiency) dataset was not uploaded here. If available, further regressions on symbolic distance can be run. If any columns do not match exactly (e.g., different spellings), modify the variable names slightly based on print(dbna.columns). 📈 Planned Outputs The code generates: Regression outputs on how Spanish-speaking prevalence correlates with: Starting a business Ease of Doing Business Regulatory quality Subnational institutional performance differences (Puerto Rico vs. U.S. states). 🌍 License and Reuse Open Data: CC BY 4.0 License Citation Requested:Brown, S.M. (2025). Symbolic Institutional Traps and the Liability of Foreignness: Language Regimes as Hidden Barriers to Multinational Entry. University of Puerto Rico. DOI: 10.5281/zenodo.15050209