Much of my research relies on datasets I construct from public, administrative, and regulatory sources. One is publicly released; the others support ongoing projects and are not currently distributed.
Much of my research relies on datasets I construct from public, administrative, and regulatory sources. One is publicly released; the others support ongoing projects and are not currently distributed.
Natural Disasters and Local Labor Markets Panel (Publicly Released) - County-level labor market outcomes linked to disaster exposure and federal assistance.
Kumar, I. (2026). Natural Disasters and Local Labor Market Dynamics: County-Quarter QWI–NOAA–FEMA Research Dataset. Texas State University Dataverse. https://doi.org/10.18738/T8/OCZ1JK
Overview
A county-by-quarter panel covering 2005–2025, about 157,000 observations, linking local labor-market outcomes to disaster exposure and federal assistance.
Data Construction
Public-use Quarterly Workforce Indicators (employment, hires, separations, earnings) merged with NOAA Storm Events records and damage measures, FEMA claims, and FEMA Public Assistance obligations at the county-quarter level.
What This Enables
Study how natural disasters affect hiring, separations, and earnings
Compare responses by damage severity and human harm
Examine differences across regions
Link to other county-level data on local economic and financial conditions
Related Research
Natural Disasters and Labor Market Dynamics
Do Financial Advisers Respond to Local Disasters? Evidence on Demand and Supply
Financial Advisory & Misconduct Dataset (In Use) - This dataset forms the backbone of multiple research projects on financial advisory markets and misconduct
Overview
Comprehensive dataset combining FINRA disclosures at the individual advisor level with firm-level characteristics to study financial advisory markets, including misconduct, labor dynamics, and firm behavior.
Data Construction
Constructed from FINRA BrokerCheck disclosures, IAPD records, and SEC ADV filings to link individual advisors with firm characteristics and regulatory information. The dataset is structured to allow analysis at both the advisor and firm levels. The dataset is further enriched with FINRA arbitration awards, enabling analysis of dispute resolution and case outcomes.
What This Enables
Study how market structure shapes financial advisory firms
Analyze dispute resolution outcomes in misconduct cases
Examine the role of regulation in shaping advisor behavior
Investigate labor market dynamics within financial advisory firms
Related Research
Automation and High-Skill Labor Markets: Evidence from Robo-Advisors
Where Gender Gaps Arise: Firm Complaint Handling, Arbitration, and Customer Recovery
Do Financial Advisers Respond to Local Disasters? Evidence on Demand and Supply
Labor Market Data Infrastructure (Under Development) - Higher-frequency measures of labor demand to study how local labor markets respond to shocks.
Overview
Extends the released panel with richer and more frequent measures of hiring and labor demand.
Data Construction
Integrates LEHD public data products, including Job-to-Job Flows, with Lightcast job-postings microdata acquired through my Translational Health Research Center Faculty Fellowship. Job postings let labor demand be measured weekly rather than quarterly. IRS migration data will be integrated, to study population mobility.
What This Enables
Measure labor demand weekly after disasters and other shocks
Track which occupations, skills, and wages employers advertise during recovery
Link changes in labor demand to firm characteristics and local conditions
Related Research
Natural Disasters and Labor Market Dynamics
Workforce Recovery after the 2025 Central Texas Floods
Risk Disclosure and Firm Labor Demand
13F Holdings Dataset (Under Development) - Detailed holdings-based dataset to study portfolio behavior and institutional investment decisions.
Overview
Dataset of institutional equity holdings based on SEC Form 13F filings, designed to analyze portfolio allocation, risk-taking, and investment behavior at the manager level.
Data Construction
Constructed from SEC Form 13F filings, standardizing holdings across institutional managers for consistent portfolio-level analysis. Coverage runs from 2004 to the present, combining the structured XML filings available from 2013Q2 onward with earlier filings from the pre-XML text format. Planned integration of 13F securities listings will improve security identification across managers and time.
What This Enables
Study portfolio risk-taking and asset allocation decisions of institutional investors
Analyze how financial intermediaries adjust investment behavior in response to technological and market changes
Examine heterogeneity across managers, strategies, and time
Link portfolio behavior to firm-level and market-level outcomes
Related Research
Robo-Advisors and Investor Welfare
COVID-19 Policy, Mobility, and Health Outcomes Dataset (In Use) - Integrating policy, mobility, and health data to study how pandemic shocks affected economic activity and behavior.
Overview
Integrated dataset combining multiple sources to capture the evolution of the COVID-19 pandemic across policy responses, population mobility, health outcomes, and economic activity at the county and state levels.
Data Construction
Constructed by integrating Google Mobility data, New York Times COVID-19 deaths, OxCGRT policy measures, HHS hospital capacity data, and CDC vaccination data. The dataset is organized at both the county and state levels, with monthly aggregation to enable consistent analysis over time.
What This Enables
Study how policy interventions affected mobility and economic activity
Analyze the relationship between health shocks and labor market outcomes
Examine heterogeneity across regions in pandemic response and recovery
Investigate interactions between public health measures, behavior, and economic outcomes
Related Research
Workplace Visibility and a Hidden Gendered Time Burden: Evidence from Personal Time Use
Election and Political Uncertainty Dataset (In Use) - To study how electoral dynamics and political uncertainty shape firm behavior and economic outcomes.
Overview
Dataset capturing electoral outcomes and political uncertainty measures across U.S. states, designed to study how elections and political transitions influence firm behavior, misconduct, and economic activity.
Data Construction
Constructed using data from MIT Election Lab and Harvard Cage combined with state-level political and economic information. The dataset includes election outcomes across presidential, gubernatorial, and senate races, and is organized at the state-time level to enable panel analysis. Includes measures of electoral competitiveness based on vote margins and indicators for political turnover across offices.
What This Enables
Study how political uncertainty affects corporate behavior, including misconduct
Analyze the impact of close elections and political turnover on economic outcomes
Examine how firms respond to changes in political leadership and policy environments
Investigate heterogeneity across states and election types
Related Research
Political Turnover and the Regulatory Enforcement Pipeline: Evidence from Corporate Misconduct
Local Financial Institutions Dataset (Under Development) - Understanding how local financial institutions shape credit supply and economic resilience.
Overview
Institution-level panel of credit unions linked to the places they operate, built to study local credit supply and how small lenders respond to economic shocks.
Data Construction
NCUA 5300 Call Reports combined with branch locations, so credit unions can be assigned to the counties they serve rather than to headquarters alone.
What This Enables
Study how disasters change credit union funding and lending
Separate credit supply from local loan demand using branch footprints
Examine how regulatory limits on capital and borrowing shape lending
Related Research
Credit Unions and Local Lending after Natural Disasters