Interactive Measurement Guide for Tax Research

Rebecca Lester (Stanford GSB) & Marcel Olbert (University of Mannheim)

Journal of Accounting and Economics (2025)

Welcome to the Interactive Measurement Tables

An interactive resource from "Firms' Real and Reporting Responses to Taxation: A Review"

Rebecca Lester (Stanford GSB) and Marcel Olbert (University of Mannheim)

Journal of Accounting and Economics 80(2-3), 2025

Access Full Paper (Open Access)

Purpose of This Resource

The database comprises measures used in academic research that are discussed or cited in Lester and Olbert (2025). It is not intended to be a comprehensive collection of all measures used in tax research, but rather a curated resource reflecting the scope of the review.

How to Use This Resource

📊 Explore by Outcome

Browse measures organized by research outcome (investment, employment, innovation, etc.)

🗂️ Browse Data Sources

Search our comprehensive database of 49 datasets used in tax research

🔍 Search Measures

Find specific measures by keyword, formula, or citation

ℹ️ About

Information on how to cite this resource and the original paper

Citation

Please cite the original paper as:

Lester, Rebecca and Marcel Olbert. 2025. "Firms' Real and Reporting Responses to Taxation: A Review." Journal of Accounting and Economics 80(2-3): 101837. https://doi.org/10.1016/j.jacceco.2025.101837

BibTeX:

@article{lester_olbert_2025,
  author = {Lester, Rebecca and Olbert, Marcel},
  title = {Firms' Real and Reporting Responses to Taxation: A Review},
  journal = {Journal of Accounting and Economics},
  year = {2025},
  volume = {80},
  number = {2-3},
  pages = {101837},
  doi = {10.1016/j.jacceco.2025.101837}
}

Explore by Outcome

Browse measures organized by the type of outcome you wish to study. Select a category below to view detailed measurement tables.

📈 Investment Measures

Capital Expenditures: Annual Changes and Investment Levels (15 measures)

🧪 R&D and Innovation

R&D Expenditures, Patents, and Citations (14 measures)

🤝 M&A Measures

Deal Measures: Probabilities, Counts, Values, and Deal Profitability (14 measures)

👥 Employment & Compensation

Employees, Wages, and Satisfaction (27 measures)

🌍 Geographic Allocation

Foreign Subsidiaries, Investment Flows, and Lockout Effects (13 measures)

Investment Measures

Capital Expenditures and Fixed Assets

Legend

[*] BEA/IRS/CENSUS: Data source requires application/review; outputs are disclosure-reviewed; BEA and IRS access limited to US citizens.

Paper References: [a] Lester (2019); [b] Albertus et al. (2022); [c] Curtis et al. (2022); [d] Amberger et al. (2021), Bethmann et al. (2018), Chen (2023), De Simone & Olbert (2024), Jacob et al. (2019); [e] Samuel (2022); [f] Liu (2020); [g] Arena & Kutner (2015); [h] Green & Kerr (2022); [i] Guenther et al. (2020); [j] Zwick & Mahon (2017); [k] Dobridge (2021); [l] De Simone et al. (2022); [m] Fox, Krull & Rane (2021); [n] Yagan (2015); [o] Coles et al. (2022).

Panel A: Annual Change in Fixed Assets (10 Measures)

# Measure Pro(s) Con(s) Data Source(s)
1 CapExt / PP&Et-1 Adjusts for size; improves cross-unit comparability; simple and directly interpretable magnitude Sensitive to small/noisy denominators COMP[a]
2 CapExtDom/For / PP&Et-1 See #1 See #1 BEA-MICRO[*],[b],[a]
3 CapExi,t / PP&Ei,t-1 See #1 See #1 CM, ASM, LBD, ACES[*],[c]
4 Δ Net Fixed Assetst / Assetst-1 Captures period-to-period investment changes; adjusts for size; improves cross-unit comparability See #1 ORBIS[d]
5 (Δ NFAt + Deprt) / Assetst-1 See #1; see #4; approximates gross investment via depreciation add-backs See #1 ORBIS[e]
6 (Δ NFCAt + D&At) / Assetst-1 See #5 See #5; includes intangibles, which are a separate category of investment. Intangibles on the balance sheet represent only a subset of the firm's intangibles ORBIS[f]
7 CapExDom/For / Total Assets See #1 See #1 WSP[g]
8 (CapEx - PP&E Sale Proceeds) / Assetst-1 See #1; see #4; removes asset sale effects See #1 WSP[h], COMP[i]
9 Investmentt / Capital Stockt-1 See #1 See #1 IRS-SOI-4562[*],[j]
10 Log(Investment) Reduces skew and stabilizes relative variation; enables multiplicative interpretation Undefined for non-positive values; zeros need adjustment CM[*],[c], QWI[c]

Panel B: Level of Investment (5 Measures)

# Measure Pro(s) Con(s) Data Source(s)
11 log(Dom CapEx / For CapEx) See #10; see #1 Undefined for non-positive values; zeros need adjustment; sensitive to small/noisy denominators COMP[k]
12 Log(Tangible fixed assets) See #10 See #10 ORBIS[l]
13 CapEx (level) Simple and directly interpretable magnitude Lacks size adjustment COMP[m]
14 Annual cash outlays for depreciable PP&E See #13 See #13 IRS-SOI-4562[*],[n]
15 Equity issuances minus dividends See #13 See #13 IRS-SOI-4562[*],[o]

R&D and Innovation Measures

R&D Expenditures, Patents, and Citations

Panel A: R&D Expenditures (3 Measures)

# Measure Pro(s) Con(s) Data Source(s)
1 R&D Expend / Assetst0 Adjusts for size; improves cross-unit comparability; simple and directly interpretable magnitude Sensitive to small/noisy denominators COMP[a]
2 Log(R&D expenditure) Reduces skew and stabilizes relative variation; enables multiplicative interpretation Undefined for non-positive values; zeros need adjustment COMP[b],[c]
3 R&D expenditures Simple and directly interpretable magnitude Lacks size adjustment CRA[d], HMRC[e]

Panel B: Patents & Innovation Signals (2 Measures)

# Measure Pro(s) Con(s) Data Source(s)
4 log(1 + Patents post-IPO) See #2 See #2; ignores patent quality; adding 1 is a statistical convenience, not theory-driven, and can bias results if zeros are frequent UVA-DARDEN[m]
5 Country Inventorsi,t / Total Inventorsi,t See #1 See #1 USPTO[n]

Panel C: R&D Intensity (1 Measure)

# Measure Pro(s) Con(s) Data Source(s)
6 R&D Expend / Total Sales See #1 See #1 COMP[f], CHINA-TAX[g], IRS-SOI-4562[h]

Panel D: Number of Patents (5 Measures)

# Measure Pro(s) Con(s) Data Source(s)
7 Citation-weighted patent counts See #3; quality-weighted proxy See #3 PATSTAT, ORBIS[i]
8 Log(Patent grants) See #2 See #4 PATSTAT[j]
9 Log(Patent applications) See #2 See #4 PATSTAT[k]
10 Number of patents See #3 See #3 PATSTAT[k]
11 Log(Future patent applications) See #2 See #4 ORBIS[l]

Panel E: Patent Citations (3 Measures)

# Measure Pro(s) Con(s) Data Source(s)
12 Log(Forward citations) See #2 See #2 G-PAT, USPTO[m]
13 Log(Non-self-citations (3-year ahead)) See #2 See #2 G-PAT, USPTO[n]
14 Citations per patent See #1; see #7 See #1 NBER-PAT[o]

M&A Measures

Deal Probability, Count, Value, and Performance

Panel A: Probability of M&A (1 Measure)

# Measure Pro(s) Con(s) Data Source(s)
1 Dummy for acquisition Simple and directly interpretable magnitude; bounded support Loses information on intensity, size, and magnitude ZEPHYR[d],[e], SDC[d],[e],[c]

Panel B: Number of Deals (3 Measures)

# Measure Pro(s) Con(s) Data Source(s)
2 Number of Deals Simple and directly interpretable magnitude Lacks size adjustment ZEPHYR[i],[h], SDC[i],[h],[c]
3 Log(M&A deals) Reduces skew and stabilizes relative variation; enables multiplicative interpretation Undefined for non-positive values; zeros need adjustment ORBIS, ZEPHYR[e]
4 Log(M&A bids) See #3 Adding 1 is a statistical convenience, not theory-driven, and can bias results if zeros are frequent SDC[d]

Panel C: Value of Deals (2 Measures)

# Measure Pro(s) Con(s) Data Source(s)
5 Log(Domestic acquisition value) See #4 See #4 SDC[d]
6 Log(Total deal value) See #4 See #4 SDC[d],[h]

Panel D: Location of M&A (1 Measure)

# Measure Pro(s) Con(s) Data Source(s)
7 Acquirer origin country indicator See #1 See #1 COMP[h], ZEPHYR[a],[j]

Panel E: Profitability (6 Measures)

# Measure Pro(s) Con(s) Data Source(s)
8 Acquisition premium over target price Adjusts for size; improves cross-unit comparability Proceeds/book mismatches may overstate negatives; sensitive to small/noisy denominators SDC[f],[d]
9 Short-window stock price reaction See #2 See #2 SDC[f],[d]
10 3-day CAR for acquiring firm See #2 See #2 SDC[f],[d]
11 Change in industry-adjusted ROA Captures period-to-period investment changes See #2 SDC[f],[d]
12 Post-acquisition 3-year BHAR See #2 See #2 SDC[f],[d]
13 Forward 24-month BHAR See #2 See #2 SDC[f],[d]

Panel F: M&A Financing (1 Measure)

# Measure Pro(s) Con(s) Data Source(s)
14 Tax-free stock-for-stock indicator See #1 See #1 ZEPHYR[g], NYSE-AMEX-DEL[b]

Employment & Compensation Measures

Employees, Wages, and Satisfaction

Panel A: Number of Employees (7 Measures)

# Measure Pro(s) Con(s) Data Source(s)
1 Number of employees Simple and directly interpretable magnitude Lacks size adjustment QCEW[a], NETS, CRSP[a], LBD[b], IAB-LIAB[d]
2 Number of private-sector employees See #1 See #1 BEA-REA, QCEW[e]
3 Log(Domestic employees) Reduces skew and stabilizes relative variation; enables multiplicative interpretation Undefined for non-positive values; zeros need adjustment IRS-1120[*],[f]
4 Log(Domestic employees) See #3 See #3 IRS-1120+1040[*],[g], NETS[h],[i], LBD[b], ORBIS[j],[k], CBP[l], BEA-MICRO, COMP[k]
5 Residual of expected foreign employees Adjusts for scale and fundamentals Not easily interpretable 10-K, COMP[n]
6 Number of offshored jobs See #1 See #1 DOL-TAA, COMP[o]
7 Growth rate of employment Captures period-to-period changes See #1 NETS[p], ORBIS, SWE-ADMIN[s]

Panel B: Wage (18 Measures)

# Measure Pro(s) Con(s) Data Source(s)
8 Log(Median full-time wage) See #3 See #3 IAB-LIAB[q]
9 Compensation scaled by assets Adjusts for size; improves cross-unit comparability Sensitive to small/noisy denominators ORBIS[k]
10 Growth of total wages (non-owners) See #7 See #7 SWE-ADMIN[r]
11 Log(Staff costs) See #3 See #3 ORBIS[s]
12 US compensation / Assetst-1 See #9 See #9 BEA-MICRO[*],[t]
13 Actual - Expected Pension Contribution See #1 See #1; proceeds/book mismatches may overstate negatives COMP[u]
14 County employees * Avg Annual Wage See #1 See #1 QCEW, NETS, CRSP[a]
15 Avg annual wage per employee (county) See #9 See #9 QCEW, NETS, CRSP[a]
16 Total payroll (county) See #1 See #1 LBD, QCEW[v]
17 Total Payroll / Total Employment See #9 See #9 LBD, QCEW[v]
18 Avg annual real wage per job See #9 See #9 BEA-REA, QCEW[e]
19 Log(Domestic compensation per employee) See #3; see #9 See #3; see #9 IRS-1120[*],[w],[f]
20 Log(Real wage per headcount) See #3; see #9 See #3; see #9 AMADEUS[y]
21 Avg log daily wage (full-time) See #3; see #9 See #3; see #9 IAB-LIAB[d]
22 Log(Total payroll) See #3; see #9 See #3; see #9 IRS-1120+1040[*],[g]
23 Log(Payroll per headcount) See #3; see #9 See #3; see #9 IRS-1120+1040[*],[g]
24 Log(Total wage expense) See #1; see #9 See #3; see #9 ORBIS, QCEW[z]
25 Average wages See #1 See #1; see #9 IRS-W2-1099[*],[g],[d]

Panel C: Reputation & Employee Satisfaction (2 Measures)

# Measure Pro(s) Con(s) Data Source(s)
26 Median employee ratings See #1 Selection bias GLASSDOOR, LEXIS[aa]
27 Annual avg of monthly satisfaction See #1 See #1; see #9 CULTUREX, COMP[ab]

Geographic Allocation Measures

Subsidiaries, Investment Flows, and Lockout Effects

Panel A: Foreign Subsidiary Holdings (7 Measures)

# Measure Pro(s) Con(s) Data Source(s)
1 Foreign equity holding company dummy Simple and directly interpretable magnitude; bounded support Loses information on intensity, size, and magnitude ORBIS, COMP[a]
2 Count of tax haven subsidiaries Simple and directly interpretable magnitude Lacks size adjustment COMP, ANN-REP[b]
3 % of subsidiaries in tax havens Adjusts for size; improves cross-unit comparability Sensitive to small/noisy denominators COMP, ANN-REP[b]
4 Δ(Haven Subs / Total Subs) Captures period-to-period investment changes; adjusts for size; improves cross-unit comparability See #3 COMP, ANN-REP[b]
5 Indicator for "dot" haven usage See #1 See #1 COMP, ANN-REP[b]
6 "Double Irish" structure indicator See #1 See #1 SEC-EDGAR-21[c]
7 Number of EP patents transferred See #2 See #2 EPO, MPI-PTD, PATSTAT[d]

Panel B: Capital Investment Flows and Lockout Effects (6 Measures)

# Measure Pro(s) Con(s) Data Source(s)
8 Foreign-target cash acquisition indicator See #1 See #1 SDC, SEC-8K, COMP, CRSP[e]
9 Foreign CapEx / Assetst-1 See #3 See #3 COMP-GEO[f], WSP[g], COMP, BEA-DIA, 10-K[h]
10 Foreign Cash / Assets ("trapped cash") See #3 See #3 COMP, BEA-DIA[i]
11 Excess Foreign Cash / Assets See #3 See #3 COMP[j]
12 Repatriation Dividends / Assets See #3; direct measure of capital flows to parent See #3 BEA-DIA[k]
13 Survey-based lockout measures Provides qualitative evidence directly from tax executives Survey self-reporting biases; limited external validity Proprietary survey[m]

Browse Data Sources

Comprehensive information about 49 unique data sources used in tax research. Select a panel below to view details.

Type Definitions:

  • Admin Restricted administrative microdata (often requires US citizenship or special clearance)
  • Public Freely available official source (open access)
  • Private Proprietary/commercial license required (subscription-based)

Panel A: Firm Financial Data

7 sources including Compustat, Orbis, Worldscope, Amadeus

Panel B: M&A and Stock Market Data

4 sources including CRSP, SDC Platinum, ZEPHYR

Panel C: Employment and Establishment Data

8 sources including CBP, QCEW, LBD, ACES, Census of Manufactures

Panel D: Government Administrative Data

12 sources including IRS (1120, W-2, SOI), BEA, HMRC, CRA, SEC

Panel E: Other Establishment and Labor Data

3 sources: IAB-LIAB, NETS, DOL-TAA

Panel F: Patent and Innovation Data

7 sources including USPTO, PATSTAT, NBER, EPO, UVA-Darden

Panel G: International Investment Data

5 sources including BEA-DIA, CPIS, IBFD, Treasury TIC

Panel H: Employee Sentiment and News Data

3 sources: Glassdoor, Culture X, LexisNexis

Panel A: Firm Financial Data

Data Source Abbrev. Type Coverage Pros Cons
Amadeus AMADEUS Private European Public & Private Firms Harmonized European financial data using standardized formats; covers both listed and unlisted European companies Expensive; data quality varies significantly across countries; sparse coverage in some regions
Annual Report Subsidiary Lists ANN-REP Public Various Companies Detailed subsidiary ownership structures and geographic presence; official company disclosures Manual collection and parsing required; inconsistent reporting standards across firms; potential for strategic omissions
Compustat COMP Private North America (US & Canada) Public Firms Standardized accounting variables enable cross-firm comparisons; long historical coverage back decades; excellent coverage of large US public firms Expensive; accounting standard changes (GAAP/IFRS) create time series breaks; limited to public companies only
Compustat Geographic Segment Files COMP-GEO Private North America Public Firms (Geographic Segments) Provides geographic breakdown of firm operations including foreign and domestic CapEx; enables study of multinational firm allocation decisions Part of broader Compustat subscription; geographic reporting requirements may create incomplete coverage; firms may aggregate segments strategically
Orbis ORBIS Private Global Public & Private Firms Global coverage of both public and private companies; includes ownership trees showing parent-subsidiary relationships; valuable for multinational firm studies Expensive; data quality varies significantly across countries; sparse coverage in some regions
SEC 10-K Annual Reports 10-K Public US Public Firms Comprehensive annual financial statements and business descriptions; free public access via EDGAR; standardized format Manual extraction required for non-standardized data; text-heavy documents require parsing; filing delays
Worldscope WSP Private Global Public Firms Global accounting fundamentals for international companies; good for cross-country comparative studies Expensive; limited replicability; uneven country coverage; variable data quality in emerging markets; inconsistent accounting standards across firms

Panel B: M&A and Stock Market Data

Data Source Abbrev. Type Coverage Pros Cons
CRSP CRSP Private US Stock Market (NYSE, AMEX, NASDAQ) Comprehensive US stock returns since 1926 with clean corporate action adjustments; widely accepted academic standard Returns are upward biased in any month where there are multiple distributions on the same day; monthly returns table often does not contain monthly returns for several months before the merger
NYSE/AMEX Deletions NYSE-AMEX-DEL Private (constructed) US Delistings/Deletions Clean event study samples by identifying specific exit reasons; reduces survivorship bias in analyses Requires significant data construction work; classification choices affect results
Orbis M&A (ZEPHYR) ZEPHYR Private Global M&A Transactions Detailed M&A deal information including values, dates, and company characteristics; good global coverage of major transactions Issues related to the continuous overwriting of firm-level data and inaccuracies in announcement dates
SDC Platinum SDC Private Global M&A Transactions Long historical coverage back to 1970s; considered gold standard for M&A deal databases; detailed participant information Expensive; limited replicability for academic research; data coverage can be inconsistent across countries and time; event classifications sometimes ambiguous or error-prone

Panel C: Employment and Establishment Data

Data Source Abbrev. Type Coverage Pros Cons
Annual Capital Expenditures Survey ACES Admin US Capital Investment by Establishment Detailed capital investment by establishment and asset type; separates structures from equipment Residency requirement; highly restricted access; strict disclosure rules; limited sample sizes for detailed breakdowns
Annual Survey of Manufactures ASM Admin US Manufacturing Plants (Annual) Annual manufacturing plant data between economic censuses; consistent definitions with Census of Manufactures Residency requirement; rotating sample coverage, not universe; restricted access; disclosure review required
Census of Manufactures CM Admin US Manufacturing Plants Detailed plant-level production, materials, and energy data for manufacturing; comprehensive coverage every 5 years Residency requirement; conducted only periodically (5-year intervals); restricted RDC access; long publication lags
County Business Patterns CBP Public US Establishments & Employment Annual subnational economic data by industry at county level; includes employment and payroll by establishment size; free public access Suppressed for confidentiality in small geographies; excludes some industries; limited firm-level detail
Longitudinal Business Database LBD Admin US Business Dynamics (Census) Universe of US establishments with longitudinal links; comprehensive employment, payroll, and industry data Residency requirement; access restricted to Census Research Data Centers (RDCs); strict disclosure rules limit publishable results
National Establishment Time-Series NETS Private US Establishment Dynamics Long-run establishment panel tracking moves, closures, and ownership changes; covers small businesses missed elsewhere Expensive; vendor-curated with proprietary methods; inconsistent data quality
Quarterly Census of Employment & Wages QCEW Public US Employment & Wages Near-census coverage of US jobs with quarterly employment and wage data; free public access; consistent methodology over time Geographic and industry detail suppressed for confidentiality; industry reclassifications complicate long-term analysis; no firm-level or worker-level information
Quarterly Workforce Indicators QWI Public US Job Flows & Worker Demographics Job flows and worker demographics (age, sex, race) with turnover rates; publicly available without restrictions Data "noise infused" for confidentiality, reducing precision; incomplete state coverage; complex methodology

Panel D: Government Administrative Data

Data Source Abbrev. Type Coverage Pros Cons
BEA Firm-level Microdata BEA-MICRO Admin US Multinational Activities Detailed multinational enterprise data including foreign affiliate operations; official government source with comprehensive coverage US citizens only; requires security clearance; extremely strict publication restrictions; respondents subject to reporting thresholds that vary over time
BEA Regional Economic Accounts BEA-REA Public US Regional GDP, Employment, Income Comprehensive regional GDP, employment, and income data; free public access with long time series Aggregated data only (no firm detail); frequent revisions change historical trends; complex methodology; respondents subject to reporting thresholds that vary over time
Canada Revenue Agency CRA Admin Canadian Tax Records Canadian corporate tax data with detailed financial information; comprehensive coverage of Canadian corporations Restricted access; requires research approval; Canada-only coverage; tax vs. book accounting differences
Chinese State Administration of Tax CHINA-TAX Admin Chinese Corporate Tax Records Comprehensive Chinese corporate tax records; large sample of Chinese firms Restricted access; requires government partnerships; China-only coverage; language barriers
Her Majesty's Revenue and Customs HMRC Admin UK Tax Records UK corporate tax and financial data; official government source for British firms Restricted access; requires approval process; UK-only coverage; complex application procedures
IRS Corporate Tax Returns IRS-1120 Admin US Corporate Tax Filings Complete universe of corporate tax returns; includes all corporations regardless of size or public status Highly restricted access requiring IRS approval; tax vs. book accounting differences; strict confidentiality rules; requires US citizenship
IRS Corporate+Individual Returns IRS-1120+1040 Admin US Corporate & Individual Tax Filings Links corporate returns to individual owners; enables study of pass-through entities and owner characteristics Highly restricted access requiring IRS approval; tax vs. book accounting differences; strict confidentiality rules; requires US citizenship
IRS SOI (Form 4562) IRS-SOI-4562 Admin US Tax-based Capital Expenditures Universe coverage of tax-based capital expenditures from depreciation schedules; precise investment timing Highly restricted access requiring IRS approval; tax vs. book accounting differences; strict confidentiality rules; requires US citizenship
IRS W-2 & 1099 IRS-W2-1099 Admin US Worker-level Earnings Worker-level earnings from tax records; near-universal US employment coverage; job-to-job transitions Highly restricted access requiring IRS approval; tax vs. book accounting differences; strict confidentiality rules; requires US citizenship
SEC 8-K Current Reports SEC-8K Public US Material Corporate Events Timely disclosure of material corporate events including M&A announcements; free public access Event-driven filings may be incomplete; requires text parsing; variable disclosure quality
SEC EDGAR Exhibit 21 SEC-EDGAR-21 Public US Subsidiary Ownership Structures Detailed subsidiary ownership structures from 10-K filings; free public access via EDGAR Manual extraction required; firms may strategically omit subsidiaries; annual frequency only
Swedish Administrative Tax Data SWE-ADMIN Admin Swedish Tax Records Rich owner-firm linkage data from tax records; high-quality Nordic administrative data Restricted to approved researchers; Sweden-only coverage; requires understanding of Swedish institutions

Panel E: Other Establishment and Labor Data

Data Source Abbrev. Type Coverage Pros Cons
German Linked Employer–Employee Data IAB-LIAB Admin German Linked Employer-Employee Data High-quality linked employer-employee data covering most German workforce; includes wages and firm characteristics Restricted researcher access; Germany-only coverage; requires German language skills
National Establishment Time-Series NETS Private US Establishment Dynamics Long-run establishment panel tracking moves, closures, and ownership changes; covers small businesses missed elsewhere Expensive; vendor-curated with proprietary methods; inconsistent data quality
US Department of Labor Trade Adjustment Assistance Program DOL-TAA Public US Layoffs due to Trade/Outsourcing Identifies trade-related mass layoffs with firm and worker information; free public access Selection bias from application process; may not reflect true trade impact; limited historical coverage

Panel F: Patent and Innovation Data

Data Source Abbrev. Type Coverage Pros Cons
European Patent Office Patent Register EPO Public European Patents Official European patent data with comprehensive coverage of EP patents; includes patent transfer records Limited to European patents; requires technical expertise to navigate; patent family relationships can be complex
Google Patents G-PAT Public Global Patents Easy-to-use search interface with basic analytics; free access with no registration Not official database; Google Patents ingests records from multiple patent offices, but the completeness, update frequency, and metadata standardization can vary. Some filings, especially older or international ones, may be missing or incomplete
Max Planck Institute Patent Transfer Database MPI-PTD Public Patent Transfers Specialized database tracking patent transfers and ownership changes; enables study of patent mobility Limited coverage period; complex methodology for identifying transfers; requires understanding of patent law
NBER Patent Data Project NBER-PAT Public US Patents (Historical) Cleaned USPTO data with citation links and disambiguated names; widely used by researchers; free access Updates lag USPTO by several years; coverage ends around 2006; methodological choices may not suit all uses
PATSTAT PATSTAT Private Global Patents (EPO) Global patent coverage from multiple offices worldwide; includes patent families and international citations Expensive; complex database requiring technical expertise; computationally intensive processing
US Patent & Trademark Office USPTO Public US Patents Official source of all US patents and applications; includes full text and inventor information; free public access Extensive name disambiguation required; complex data structure; US patents only
UVA Darden Patent Dataset UVA-DARDEN Public US Patents Matched to Public Firms Curated patent portfolios matched to public firms; ready-to-use for corporate finance research Patents awarded only by the US Patent and Trademark Office (USPTO), between 1980 and 2017 to publicly listed firms internationally

Panel G: International Investment Data

Data Source Abbrev. Type Coverage Pros Cons
Bureau of Economic Analysis Direct Investment Abroad BEA-DIA Admin US Outbound Investment Official US data on multinational enterprise operations abroad; includes detailed foreign affiliate financial data US citizens only; requires security clearance; strict publication restrictions; reporting thresholds vary over time
Coordinated Portfolio Investment Survey CPIS Public Global Portfolio Investment Standardized cross-country portfolio investment data; enables international comparisons; free access Countries use different methodologies despite common framework; data quality varies by reporter
Deloitte Tax Treaties Database DELOITTE-TT Private Global Tax Treaties Practitioner-oriented treaty summaries; includes implementation guidance; regular updates Coverage can lag behind official treaty updates, and details are often presented as practitioner summaries rather than verbatim legal texts
International Bureau of Fiscal Documentation IBFD Private Global Tax Information Authoritative tax treaty texts and international tax law; comprehensive global coverage; expert analysis Expensive; subscription-locked, and not easily replicable for other researchers. Its content is curated summaries of tax law rather than always the raw primary sources, which can introduce interpretation bias. The database is also geared toward legal practitioners, so it's not machine-ready for large-scale quantitative analysis
Treasury International Capital System TIC Public US Cross-border Financial Flows Official cross-border financial flow data involving US entities; long time series; free public access Highly aggregated with no firm-level detail; frequent historical revisions; complex reporting may affect quality

Panel H: Employee Sentiment and News Data

Data Source Abbrev. Type Coverage Pros Cons
Glassdoor by Culture X CULTUREX Private Employee Satisfaction Indices Curated employee satisfaction indices using advanced analytics; claims to reduce noise from raw reviews Expensive; proprietary black-box methodology; difficult to understand index drivers; limited research applications
Glassdoor Ratings GLASSDOOR Private Employee Reviews & Salaries Large-scale employee reviews and company ratings; includes salary data and interview experiences Expensive; strong selection bias toward disgruntled employees; potential for company manipulation
LexisNexis News LEXIS Private Global News & Legal Comprehensive news and business information from thousands of global sources; searchable full text archives Expensive; requires extensive deduplication work; search algorithms may miss relevant content

Glossary of Terms

Definitions of key financial and economic terms used in this resource

Acquisition Premium

Definition: The difference between the price paid to acquire a target company and its pre-announcement market value. It represents the additional amount an acquirer is willing to pay above the target's standalone value.

Context in Research: Used to assess whether tax considerations affect the pricing of M&A transactions.

Administrative Data / Microdata

Definition: Data collected by government agencies for administrative purposes (tax collection, program administration, regulation) rather than for statistical research. Microdata refers to individual-level or firm-level records, as opposed to aggregated statistics. Sources marked with [*] in this resource require special access.

Context in Research: Provides near-universe coverage and precise measurement but typically requires citizenship requirements, security clearances, and disclosure review before publication.

Buy-and-Hold Abnormal Return (BHAR)

Definition: The difference between the return from holding a stock over a specified period and the return from a benchmark (such as a market index or matched firm). Unlike CAR, BHAR compounds returns over time, making it suitable for measuring long-term performance.

Context in Research: Used to evaluate post-acquisition performance over extended periods (e.g., 24 months or 3 years).

Capital Expenditures (CapEx)

Definition: Funds used by a company to acquire, upgrade, and maintain physical assets such as property, plants, buildings, technology, or equipment. CapEx is often used to undertake new projects or investments by a company.

Context in Research: Often used as a measure of real investment response to tax incentives.

Capital Stock

Definition: The total value of a firm's or economy's accumulated physical capital—buildings, machinery, equipment, and other durable goods used in production. Capital stock is typically measured at replacement cost or book value and evolves as new investment adds to the stock while depreciation subtracts from it.

Context in Research: Used as a denominator to scale investment flows, creating investment rates comparable across firms or industries of different sizes.

Citation-weighted Patent Counts

Definition: A measure of innovation output that weights each patent by the number of citations it receives from subsequent patents. This approach gives greater weight to patents that are more influential or technologically important.

Context in Research: Used as a quality-adjusted proxy for innovation output, addressing the limitation that raw patent counts treat all patents as equally valuable.

Cumulative Abnormal Return (CAR)

Definition: The sum of all abnormal returns over a specified time period, where abnormal return is the difference between actual stock return and expected return. It measures the total impact of an event on a stock's price.

Context in Research: Used to measure short-window market reactions to M&A announcements (e.g., 3-day CAR).

Depreciation / D&A

Definition: Depreciation is the systematic allocation of the cost of a tangible asset over its useful life. D&A (Depreciation and Amortization) also includes amortization, which applies the same concept to intangible assets.

Context in Research: Added back to changes in net assets to approximate gross investment, since depreciation reduces book value but does not represent a cash outflow.

Disclosure Review

Definition: A process required by statistical agencies (e.g., Census Bureau, IRS) before research outputs can be released publicly. Reviewers check that results do not reveal information about individual respondents, applying rules on minimum cell sizes, dominance, and complementary disclosure.

Context in Research: Limits what researchers can publish from administrative data; may require aggregation or suppression of detailed results.

Double Irish Structure

Definition: A tax planning arrangement that exploited differences between Irish and U.S. tax rules by using two Irish subsidiaries—one tax resident in Ireland and one tax resident in a low-tax jurisdiction—to shift profits and minimize global tax liability. Ireland closed this structure to new entrants in 2015.

Context in Research: Used as an indicator variable to identify firms employing aggressive international tax planning strategies.

Employment

Definition: In tax research, employment refers to the level, growth, and composition of a firm's workforce, as well as worker compensation. Employment outcomes are studied as "real" effects of tax policy because they represent changes in firms' allocation of resources to labor inputs.

Context in Research: Tax policy affects employment indirectly through its impact on investment (via labor-capital complementarity or substitutability) and through tax incidence effects on wages. Common measures include headcount, payroll, wage levels, and employment growth rates.

Excess Foreign Cash

Definition: Foreign cash holdings that exceed the amount needed for normal overseas operations. It represents accumulated foreign earnings retained abroad beyond operational requirements, often due to tax considerations related to repatriation.

Context in Research: Measured as foreign cash scaled by assets, used to study the extent to which repatriation taxes create distortions in capital allocation beyond the "trapped cash" concept.

Fixed Assets / Net Fixed Assets (NFA)

Definition: Fixed assets are long-term tangible assets used in operations, such as land, buildings, and equipment. Net Fixed Assets (NFA) equals gross fixed assets minus accumulated depreciation, representing the remaining book value of these assets.

Context in Research: Changes in NFA, combined with depreciation add-backs, are used to construct investment measures.

Forward Citations

Definition: The number of times a patent is cited by subsequent patents. Forward citations serve as a quality-weighted measure of a patent's technological importance and influence.

Context in Research: Used to measure innovation quality, with non-self-citations (citations from other firms) often preferred to avoid bias from self-referencing.

GAAP / IFRS

Definition: Generally Accepted Accounting Principles (GAAP) are U.S. accounting standards set by FASB. International Financial Reporting Standards (IFRS) are global standards set by the IASB, used in over 140 countries. The two frameworks differ in areas like revenue recognition, inventory accounting, and lease treatment.

Context in Research: Differences between GAAP and IFRS can create measurement challenges when comparing firms across countries or when accounting standards change over time.

Headcount

Definition: The total number of employees at a firm or establishment, typically measured as full-time equivalents (FTEs) or total persons employed. Headcount provides a straightforward measure of labor input that is comparable across firms and over time.

Context in Research: Used as a denominator to calculate per-employee metrics (e.g., payroll per headcount, compensation per employee) and as a direct measure of employment levels in studies of tax incidence on labor.

Initial Public Offering (IPO)

Definition: The first sale of stock by a private company to the public. After an IPO, the company's shares trade on a public stock exchange, subjecting the firm to disclosure requirements and providing access to public capital markets.

Context in Research: IPO timing can affect innovation measurement; post-IPO patent counts are used to study how going public affects innovation incentives.

Intangible Assets

Definition: Non-physical assets such as patents, trademarks, copyrights, and goodwill. They are critical for innovation but often harder to value than physical assets. Only a subset of intangibles appear on the balance sheet (typically acquired intangibles), while internally developed intangibles are usually expensed.

Context in Research: Relevant when distinguishing tangible investment from total investment; IP transfers to low-tax jurisdictions are a key focus in profit-shifting research.

Investment

Definition: In the context of tax research on real effects, investment refers to firms' expenditures on productive capital assets, including tangible assets (property, plant, and equipment) and intangible assets (R&D, patents). The neoclassical theory of investment predicts that corporate taxes affect investment through changes in the after-tax cost of capital.

Context in Research: Investment is the primary "real" outcome studied in tax research. Common measures include capital expenditures (CapEx), changes in fixed assets, and investment ratios scaled by assets or prior-period capital stock. Tax depreciation rules and corporate income tax rates directly affect investment incentives.

Linked Employer-Employee Data

Definition: Administrative datasets that match information about individual workers (wages, demographics, job history) to information about their employers (industry, size, financials). Examples include Germany's IAB-LIAB and matched IRS W-2 data in the U.S.

Context in Research: Enables studying how corporate tax changes affect worker outcomes like wages and employment, while controlling for both worker and firm characteristics.

Lockout Effect / Trapped Cash

Definition: Under a worldwide tax system (pre-2018 U.S. law), the phenomenon where U.S. multinationals held foreign earnings abroad to defer U.S. tax on repatriation. "Trapped cash" refers to the foreign cash holdings that firms were reluctant to bring home due to the associated tax cost.

Context in Research: Measured as foreign cash scaled by total assets; used to study how repatriation taxes affect capital allocation and investment location decisions.

Log Transformation

Definition: A mathematical transformation that replaces a variable with its natural logarithm. Log transformations reduce right-skewness in distributions (common for firm size, investment, and patent counts), stabilize variance, and allow regression coefficients to be interpreted as percentage changes.

Context in Research: Commonly applied to investment, employment, and R&D measures. A limitation is that log is undefined for zero or negative values, requiring adjustments like log(1+x) which can introduce bias.

Mergers & Acquisitions (M&A)

Definition: Transactions in which the ownership of companies or their operating units are transferred or consolidated. A merger combines two firms into one, while an acquisition involves one company purchasing another.

Context in Research: Tax considerations can influence M&A activity through deal structure (taxable vs. tax-free), target selection, and financing choices.

Multinational Enterprise (MNE)

Definition: A corporation that owns or controls production of goods or services in at least one country other than its home country. MNEs operate through foreign subsidiaries, branches, or affiliates and face tax obligations in multiple jurisdictions.

Context in Research: Central to studying profit shifting, transfer pricing, and the real effects of international tax rules on investment location decisions.

Non-self-citations

Definition: Citations to a patent from patents held by different firms or inventors. Non-self-citations exclude citations from the same patent holder's other patents, focusing on external recognition of an invention's influence.

Context in Research: Used as a quality measure for innovation output that avoids the bias introduced when firms cite their own prior work. Often measured over a fixed forward window (e.g., 3 years ahead).

Patent Applications vs. Patent Grants

Definition: Patent applications are filings submitted to a patent office requesting protection for an invention. Patent grants are applications that have been approved after examination. Applications reflect innovation inputs closer to when R&D occurs, while grants confirm the invention met patentability standards.

Context in Research: Applications capture innovation timing more precisely; grants provide a quality screen but introduce lag.

Patent Families

Definition: A set of patents filed in multiple countries that protect the same invention. Patent families arise when inventors seek protection in multiple jurisdictions, with each national filing considered part of the same "family" linked by priority claims.

Context in Research: Using patent families rather than individual patents avoids double-counting the same invention filed in multiple countries.

Patent Transfers

Definition: The reassignment of patent ownership from one entity to another, either within a corporate group (e.g., from parent to subsidiary) or between unrelated parties. Patent transfers can occur through sale, licensing arrangements, or corporate restructuring.

Context in Research: Used to study how firms relocate intellectual property across jurisdictions, often to low-tax locations. The number of EP (European) patents transferred across countries is a key measure in profit-shifting research.

Payroll / Total Payroll

Definition: The aggregate wages and salaries paid to all employees of a firm or establishment over a given period. Total payroll represents the firm's labor cost and can be measured at the firm, establishment, or geographic (e.g., county) level.

Context in Research: Used as a measure of labor compensation in studies of tax incidence on workers. When scaled by headcount, payroll provides average wage measures comparable across firms.

PP&E (Property, Plant & Equipment)

Definition: Long-term tangible assets used in the production of goods and services, including land, buildings, machinery, and equipment. PP&E appears on the balance sheet at historical cost less accumulated depreciation.

Context in Research: Commonly used as a denominator to scale capital expenditures, creating an investment rate that is comparable across firms of different sizes.

Repatriation

Definition: The process of returning foreign-earned profits or financial assets back to the company's home country, typically through dividends from foreign subsidiaries to the parent company.

Context in Research: Repatriation dividends scaled by assets measure capital flows back to the parent; central to studying lockout effects and the 2017 Tax Cuts and Jobs Act transition tax.

Research and Development (R&D)

Definition: Activities that companies undertake to innovate and introduce new products and services. R&D expenditures include wages for researchers, materials, and overhead costs directly related to innovation efforts.

Context in Research: A key measure of innovation investment; R&D intensity (R&D/Sales) and R&D scaled by assets are common outcome variables in studies of R&D tax credits.

R&D Intensity

Definition: A ratio measuring a firm's R&D expenditures relative to its total sales. R&D intensity captures the proportion of revenue devoted to innovation activities, allowing comparisons of innovation focus across firms of different sizes.

Context in Research: A key outcome variable in studies of R&D tax credits and innovation incentives. Higher R&D intensity indicates greater relative commitment to innovation activities.

Research Data Center (RDC)

Definition: Secure facilities operated by statistical agencies (e.g., U.S. Census Bureau) where approved researchers can access confidential microdata. Researchers must pass background checks, complete training, and work in monitored environments where data cannot leave the facility.

Context in Research: Required for accessing Census, IRS, and other administrative microdata; results must pass disclosure review before publication.

Return on Assets (ROA)

Definition: A profitability ratio calculated as net income divided by total assets. It measures how efficiently a company uses its assets to generate earnings.

Context in Research: Industry-adjusted ROA changes are used to evaluate post-acquisition operating performance in M&A studies.

Subsidiary

Definition: A company controlled by another company (the parent), typically through majority ownership of voting stock. Subsidiaries are separate legal entities that file their own tax returns but are consolidated into the parent's financial statements for reporting purposes.

Context in Research: The count and location of subsidiaries (especially in tax havens) is used to measure corporate tax planning intensity and geographic profit allocation.

Staff Costs

Definition: Total personnel-related expenses incurred by a firm, including wages and salaries, employer payroll taxes, pension contributions, and other employee benefits. Staff costs represent the comprehensive cost of labor beyond just direct wage payments.

Context in Research: Used as a measure of labor compensation in studies examining how corporate taxes affect worker outcomes. The log of staff costs is a common dependent variable in tax incidence research.

Tax Haven

Definition: A jurisdiction with very low or no tax rates for foreign investors, often combined with strict secrecy laws. Common examples include Ireland, Luxembourg, Bermuda, the Cayman Islands, and the Netherlands.

Context in Research: The count or percentage of subsidiaries in tax havens is used to measure firms' tax planning intensity and geographic profit shifting.

Tax-Free Reorganization (Stock-for-Stock)

Definition: An M&A transaction structured to qualify for tax-deferred treatment under the Internal Revenue Code, typically involving an exchange of stock rather than cash. Target shareholders defer capital gains tax until they sell the acquirer's stock.

Context in Research: Used as an indicator to study how tax considerations influence M&A deal structure and financing choices.

Total Assets

Definition: The sum of all assets owned by a firm, including current assets (cash, receivables, inventory) and non-current assets (property, plant, equipment, intangibles). Total assets appear on the balance sheet and represent the resources available to generate future economic benefits.

Context in Research: Widely used as a scaling variable to normalize measures across firms of different sizes (e.g., CapEx/Assets, R&D/Assets, Compensation/Assets). Lagged total assets are often preferred to avoid simultaneity concerns.

User Cost of Capital (CoC)

Definition: The minimum rate of return that an investment must yield to cover financing costs and taxes while breaking even. In neoclassical investment theory, the CoC depends on the interest rate, corporate tax rate, and tax depreciation rules. Firms maximize value by investing until the marginal product of capital equals the CoC.

Context in Research: The theoretical foundation for studying real investment responses to taxation. Tax policy affects investment through changes in the CoC: more generous depreciation rules lower the CoC, while higher tax rates (absent full expensing) raise it.

About This Resource

📄 About the Paper

Abstract

This review synthesizes research on firms' real and reporting responses to taxation, providing a comprehensive measurement guide for tax research.

Key Contributions

  • Comprehensive catalog of measurement approaches
  • Detailed data source information with access requirements
  • Pros and cons of different measurement choices
  • Interactive format for easy navigation
👤 About the Authors

Rebecca Lester

Associate Professor of Accounting (with tenure)

Stanford University Graduate School of Business

Rebecca Lester is an Associate Professor of Accounting at the Stanford Graduate School of Business. She is also a Senior Fellow at the Stanford Institute for Economic Policy Research (SIEPR) and a Research Fellow at the Hoover Institution.

Professor Lester's research studies how tax policies affect corporate investment and employment decisions. In particular, she demonstrates the critical role of reporting incentives, disclosure regimes, and information frictions in facilitating or altering the effectiveness of cross-border, federal, and local tax incentives.

Her work has been cited by the U.S. Joint Committee on Taxation and featured in The Wall Street Journal, Fortune, Bloomberg, and The Huffington Post. She was selected as one of Poets and Quants "40 Best Business Professors Under 40" in 2018 and received the Stanford MBA Distinguished Teaching Award in 2021.

Professor Lester received her PhD in Accounting from the MIT Sloan School of Management.

→ Visit Rebecca Lester's Website

Marcel Olbert

Professor of Taxation, Accounting, and Finance

University of Mannheim

Marcel Olbert is a Professor of Taxation, Accounting, and Finance at the University of Mannheim. His research intersects the fields of disclosure regulation, corporate taxation, and corporate finance.

Professor Olbert's work has been published in top-tier journals including The Accounting Review, the Journal of Accounting and Economics, the Journal of Accounting Research, and the Review of Financial Studies.

He currently serves as an Associate Editor at the European Accounting Review and is a Junior Research Associate at ZEW (Leibniz Centre for European Economic Research) and a Research Fellow of TRR 266 Accounting for Transparency. For his work on international taxation, transparency, and multinational firms, he has received research grants from the NBER, the Wheeler Institute, and the ITFP. He was selected as one of Poets and Quants "40-Under-40 MBA Professors" in 2024.

→ Visit Marcel Olbert's Website

🙏 Acknowledgments

Research Assistants: Lauren Andrews, Jensen Ahokovi, Manon Francois, and Qihang Zhang.

Funding Sources: Marcel Olbert acknowledges the financial support of the Research and Materials Development Fund (Olbert_2023_8934) at London Business School.

📧 Contact & Feedback

For questions, suggestions, or to report errors, please contact the authors:

We welcome feedback on:

  • Suggestions for new measures to include
  • Corrections or updates to existing entries
  • Additional data sources
  • Improvements to the interactive resource
📋 Updates & Version History

Version 1.0 - Initial release

Future updates will be documented here