Appendix A - License to Pour: Access & Equity in Massachusetts Liquor Licensing Report

Research Methods

This appendix explains how this study was conducted: what it set out to answer, what evidence it assembled and in what order, how the evidence streams were checked against one another, and what limitations attach to the study. Method-specific detail appears in the appendices that follow—quantitative methods in Appendix B, survey and interview methods in Appendix C, and the legal context in Appendix E. 

A.1 Study scope and objectives

Purpose. This is a municipal impact study conducted under MGL c. 11, § 6B, which directs DLM to examine the effect of state law on municipalities. The question it asks is whether the population-based quota on retail liquor licenses in MGL c. 138, § 17, produces outcomes consistent with its purpose, and what those outcomes mean for municipalities and for people trying to obtain a license.

Principal research questions. (1) Does the population-based quota create supply and demand mismatches? (2) What practical difference does the local-control exemption make for the 26 municipalities that hold it? (3) Does license access vary with municipal wealth? (4) How do the home rule/special legislation and secondary-market pathways work in practice, and for whom? (5) Can prospective licensees and the public find consistent process and availability information?

Licenses examined. This study covers on-premises retail licenses issued under § 12—the licenses held by restaurants, hotels, taverns, and similar establishments—in both the All-Alcohol and Wine-and-Malt tracks. Off-premises licenses under § 15 are outside of the scope. They serve a different market, are allocated under a separate formula, and would not be comparable using any of the measures used here.

Geographic and temporal scope.This study covers all 351 Massachusetts municipalities. Licensing data contains the ABCC fiscal year 2024 annual quota reports, which supply each municipality’s annual report of license counts and active licenses as of the fiscal year 2024 ABCC active-license extract. For all population-dependent analyses in this report—licenses per 10,000 residents, Finding 1 constrained-population totals, Finding 2 rates and the hypothetical quota allocation, Finding 3 access rates, and Table 1 classification populations—population is the 2023 municipal series in DLM’s dataset (Department of Revenue/Division of Local Services estimates based on US Census Bureau population estimates). The § 17 statute itself keys allotments to decennial census populations; historical reconstruction of statutory allocations uses decennial census counts back to 1930. Municipal wealth is fiscal year 2026 EQV, aligned with 2023 populations. Survey fieldwork ran from December 2025 through February 2026; interviews were conducted in 2026.

Boundaries. Three boundaries matter for reading the findings. The licensing data is cross-sectional, so this study documents association and does not estimate change over time or causal effect. This study measures access at the municipal level and does not measure how licenses are distributed among applicants within a municipality. And this study does not audit ABCC or any municipal licensing authority; it takes the administrative records as given and says so where that assumption matters.

A.2 Research Design

The study uses a mixed-method research design. The administrative data establishes how many licenses exist, where they are, and how they correlate with population and wealth. This data cannot show what it costs in time or money to obtain one, why a petition failed, or what a prospective licensee was able to find out before applying. Six evidence streams were used, each answering what the others could not.

1. Statutory, regulatory, and legislative research.The § 17 allocation was reconstructed from the statutory text, rather than taken from any secondary description, because secondary sources frequently misstate it. Session laws from 1933 forward were reviewed to identify special legislation. ABCC advisories and decisions, the 2017 Task Force Report, and the legislative history of the 1978–1984 local-control window supplied the regulatory and historical context. This work defined what the quantitative analysis had to measure. See Appendix D.
2. ABCC licensing data.The ABCC fiscal year 2024 annual quota reports supply each municipality’s annual license counts.The ABCC fiscal year 2024 active‑license extract provides record-level license counts by municipality, type, class, and premises address, together with the flag identifying licenses obtained by special act. It is the source for every issued-license figure in the report.
3. Municipal and socioeconomic data.Population, EQV, and related fiscal/socioeconomic fields are used for wealth and access measures. The Division of Local Services provided socioeconomic data. The Department of Revenue EQV (fiscal year 2026) supplies the wealth measure; 2023 Department of Revenue/Division of Local Services municipal population estimates in the DLM dataset supply the normalization base for licenses per 10,000 and EQV per capita. Administered § 17 allotments used for utilization are the DLM dataset allowed fields. County Business Patterns supplies establishment counts for the interstate comparison.
4. Quantitative and statistical analysis.Descriptive and rank-based association testing, described in Appendix B, establishes the distributional patterns on which the findings rest and tests which candidate mechanism carries the wealth gradient.
5. Municipal and stakeholder survey.A survey of municipal officials, business owners, and state officials’ documents cost, delay, and process experience, which the administrative records do not capture, and supplies the barrier ratings and open-ended commentary used throughout the findings. See Appendix C.
6. Interviews and stakeholder engagement.Semi-structured interviews with license holders, municipal officials, legislators, and industry participants supply the process narratives—how a home rule petition moves, what a transfer negotiation looks like—that neither the data nor the survey can reconstruct.

These streams were used together rather than in parallel. Where a claim rests on more than one, the report cites both. Where it rests on one, the report says which. 

A.3 Analytical Approach

A.3   Analytical Approach 

This study proceeded in a fixed sequence, and the order matters because each stage conditions the next.

1. Legal and historical review established what the statute allocates, to whom, and with what exceptions, which is what made an independent reconstruction of the quota possible.
2. Data collection and construction assembled the municipality-level dataset and the record-level license extract and validated them against each other.
3. Descriptive analysis reconstructed the § 17 allocation for every municipality, compared it with licenses issued, and identified where the formula binds and where it does not.
4. Comparative and statistical analysis compared local-control with quota municipalities, tested whether license access varies with municipal wealth, and tested the three candidate mechanisms (allocation, utilization, and special legislation) to determine which one carries the gradient.
5. Survey research documented cost, delay, information availability, and process experience across all three stakeholder groups.
6. Interviews followed the survey, drawing partly on respondents who indicated willingness to speak further, and pursued the questions the survey surfaced.
7. Validation and synthesis reconciled the streams, resolved discrepancies against source records, and fixed the language of each finding to what the evidence supports.

A.4 Limitations

The study’s principal limitations are summarized below. Limitations specific to the quantitative analysis appear in B.5 and those specific to the stakeholder research in C.6.

1. The design is observational and cross-sectional. The report documents association. It does not establish that the quota causes any outcome it describes, and no part of it should be read as claiming so.
2. Selection into local control. Eligibility for the 1978–1984 window turned on previous seasonal licensing, so local-control municipalities are disproportionately resort and tourism economies. Their higher issuance cannot be read as an estimate of what local control would produce elsewhere.
3. No demand measure. There is no direct measure of latent demand for licenses. Utilization is used as a proxy for constraint and is imperfect: A municipality may issue few licenses because demand is low rather than because it chooses not to.
4. Transfer prices are self-reported. Thirteen business owners reported what they paid. These are not market surveys and were not independently verified. No public record of transfer prices exists.
5. Survey composition. Municipal officials are substantially over-represented relative to business owners and state officials, and respondents self-selected. Results describe the respondent pool and are not weighted to the population of municipalities.

Date published:	October 13, 2026