How disclosure works — an owner's view
Screenshot shared by a local Ann Arbor business owner
What it is: A cross-sectional scrape of ~5 million U.S. Google Maps business listings collected in 2021–2022. Business listings include owner-elected identity tags (Black-owned, Women-owned, Veteran-owned, LGBTQ+-owned). Tags are public-facing and represent an active strategic choice by the owner.
Coverage: ~160,000 Black-owned businesses; ~280,000 Women-owned businesses; 2,681 U.S. counties.
Key variable: Binary indicator for whether a listing carries the identity tag. County-level disclosure rates computed as tagged / total (by identity type).
What it is: Longitudinal tracking of Google Maps business listings in the New York–Newark–Jersey City MSA across three observation windows (2019, 2021, 2022). Identifies when businesses added or removed identity tags.
Coverage: ~18,000 business listings tracked across observation windows. Enables difference-in-differences analysis around the 2020 Floyd protests.
Key variable: Tag adoption timing. Whether businesses that added a tag in 2021 were in high- or low-solidarity neighborhoods.
What it is: State cosmetology licensing records (from 12 states with accessible data) matched to Google Maps business listings by business name and address. Provides a subsample of businesses with verified ownership status and strong theoretical reasons to expect identity disclosure.
Coverage: ~22,000 matched cosmetology businesses. License records provide demographic data on license holders.
Key variable: Verified Black-owned status (license holder race/ethnicity, where available) vs. Google Maps tag presence. Tests whether the J₁×J₂ interaction holds in a sample with verified ownership.
What it is: Longitudinal panel of Google Maps business listings in Washtenaw County, Michigan (Ann Arbor metro) across 2019–2023. Supports the Chapter 4 interview sampling by providing a mapping of which businesses have/have not disclosed over time in the interview geography.
Coverage: ~4,200 businesses with identity tag data. Used to identify interview targets in each of the four sampling cells (disclosed/not × Black-owned/Women-owned).
What it is: 45 semi-structured interviews with Black-owned and Women-owned business owners in the Ann Arbor/Washtenaw County area. Sampling frame drawn from the Washtenaw panel. IRB active.
Current status: 12 interviews complete (as of April 2026). IRB protocol #[HUM XXXXXX] active.
Protocol: 60-minute interviews covering business history, identity and positioning, awareness of Google Maps tagging, the decision process, perceived risks and benefits. Atlas.ti for thematic coding.
Sampling cells: Disclosed / Not disclosed × Black-owned / Women-owned. Oversamples non-disclosers.
What it is: Primary and secondary archival sources documenting the history of Black business visibility in the United States, focusing on the period 1880–2020. Sources include the Chicago Defender, Pittsburgh Courier, Tulsa World, Library of Congress Prints and Photographs, and secondary historical scholarship.
Key sources: Business directories (e.g., Negro Business League records); "Buy Black" campaign materials (1960s–70s); Greenwood District commercial records (Tulsa); Paradise Valley and Bronzeville merchant records.
Analytic approach: Tracing the dual-valence structure historically — how Black business visibility has simultaneously functioned as a site of solidarity/community empowerment and as a target for economic retaliation and violence.
Dissertation Proposal Defense · April 16, 2026
Kyle McCullers
PhD Candidate · Strategy · Ross School of Business, University of Michigan
Advisor
Christopher Rider · Associate Professor of Entrepreneurial Studies
Committee
Elizabeth Armstrong · Professor of Sociology
Justin Frake · Assistant Professor of Strategy
Gerald Davis · Professor of Management and Organizations
Alford Young Jr. · Professor of Sociology & Afroamerican and African Studies
Roadmap
The Phenomenon
Should a business owner tell customers who they are?The answer is not obvious — and it is not free. The same disclosure that attracts loyal, values-aligned customers can invite discrimination, selective avoidance, and retaliation. This is the disclosure dilemma — and it is the daily reality of millions of business owners deciding whether their identity is an asset or a liability in the market they operate in.
Screenshot shared by a local Ann Arbor business owner
The Phenomenon
No disclosure
Full multi-platform disclosure
This dissertation measures Google Maps — a conservative lower bound on the phenomenon.
The Phenomenon
No disclosure
Full multi-platform disclosure
This dissertation measures Google Maps — a conservative lower bound on the phenomenon.
The Phenomenon
No disclosure
Full multi-platform disclosure
This dissertation measures Google Maps — a conservative lower bound on the phenomenon.
The Phenomenon
No disclosure
Full multi-platform disclosure
This dissertation measures Google Maps — a conservative lower bound on the phenomenon.
The Phenomenon
No disclosure
Full multi-platform disclosure
This dissertation measures Google Maps — a conservative lower bound on the phenomenon.
The Phenomenon
No disclosure
Full multi-platform disclosure
This dissertation measures Google Maps — a conservative lower bound on the phenomenon.
The Puzzle
Black-owned businesses in high-solidarity counties disclose at six times the rate of those in low-solidarity counties
Same platform · Same year · 152,794 businesses · 51 states
What determines whether a business owner chooses to be visible?
Solidarity should operate wherever co-ethnic community exists. But high Black population share in Republican counties produces less than half the disclosure rate of equivalent Democratic counties. Something attenuates the mechanism.
Portes, A., & Sensenbrenner, J. (1993). American Journal of Sociology, 98(6), 1320–1350. | Coleman, J. S. (1988). American Journal of Sociology, 94(Suppl.), S95–S120.Identity commitment and values should predict disclosure regardless of context. But the ecological pattern is too systematic for individual variation to explain — there is a community-level structure that individual accounts miss.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In The Social Psychology of Intergroup Relations. | Spence, M. (1973). Quarterly Journal of Economics, 87(3), 355–374.Neither account alone explains the full pattern. A framework that bridges solidarity structure and individual decision-making is required.
Why This Matters
Greenwood District, Tulsa, 1921: Black Wall Street — a thriving commercial district — was burned to the ground. Visible prosperity made the community a target. The discrimination side of dual valence has a material record stretching back to Reconstruction.
Buy Black campaigns, Negro Motorist Green Book, and today's Google Maps tags all represent the same logic: deliberate visibility as an appeal to a solidarity-based market. The dual-valence structure predates the digital era.
⟷ These are not parallel tracks — they are mutually constituting. The solidarity signal is made necessary by discrimination exposure; visibility invites retaliation. The Negro Motorist Green Book was both solidarity and a tool to evade violence. ⟷
Black-owned businesses · 2020
"How The Black Lives Matter Movement Boosted Local Black-Owned Businesses"
GBH · 2021 ↗"Black businesses still face systemic racism — even from well-meaning customers"
Washington Post · 2020 ↗Asian-owned businesses · COVID-19
"The Best AAPI-Owned Businesses to Shop Right Now"
NBC News Select ↗"Reports of Anti-Asian hate crimes rose nearly 150% in major U.S. cities"
CBS News · 2021 ↗Palestinian-owned businesses · 2023
"Supporting Palestinian startups and small businesses during this time"
Google Blog · 2024 ↗"Palestinian-American business owners face death threats, expletives"
ABC News · 2023 ↗Latino-owned businesses · 2024–2025
"Grassroots efforts sprout to support Lake Street businesses hurting as Latinos stay home"
Star Tribune · 2025 ↗"Hispanic businesses in Montgomery feeling 'hunted' and afraid after triple-slaying"
NBC News · 2024 ↗Same disclosure act. Opposite valences. The net outcome is community-contingent.
Explaining who discloses — and where — requires a framework. →
Stage 1 — Enabling Conditions
Owner holds a stigmatized or historically discriminated identity — race, gender, sexuality, religion, or national origin.
Owner-operator with direct agency over how the business presents itself — not a passive investor, employee, or franchisee.
Views identity as a potentially legitimate competitive dimension — no categorical objection to identity-based positioning.
Stage 2 — At Least One Trigger
Co-ethnic in-group reciprocity — community members preferentially support owners who share their identity.
Values-aligned ally support from the broader political community — non-co-ethnic customers who actively seek out businesses with that identity.
Disclosure as moral obligation regardless of economic return — identity group commitment overrides the market calculus.
Solidarity arising within a bounded co-ethnic or co-identity community. Emerges from shared circumstance and adversity; produces in-group reciprocity and preferential exchange among group members.
Mechanism: The higher the concentration of co-ethnics in a market, the higher the potential return to disclosure — more people who will preferentially patronize the business.
Portes, A., & Sensenbrenner, J. (1993). Embeddedness and immigration: Notes on the social determinants of economic action. American Journal of Sociology, 98(6), 1320–1350.Solidarity extending beyond the in-group to values-aligned allies in the broader political community. Not named in prior literature, but consistent with homophily theory: people prefer to associate with, and support, those who share their values and political worldview.
Mechanism: In politically progressive contexts, a broader pool of non-co-ethnic customers will actively seek out and support identity-owned businesses — expanding the potential return to disclosure.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415–444. | Greenberg, J., & Mollick, E. (2017). Activist choice homophily and the crowdfunding of female founders. Administrative Science Quarterly, 62(2), 341–374.Disclosure as moral obligation independent of economic calculation. Identity salience theory: when one's identity is highly central to self-concept, its expression becomes obligatory rather than instrumental.
Stryker, S. (1968). Identity salience and role performance. Journal of Marriage and the Family, 30(4), 558–564. | Follmer, K. B., Sabat, I. E., & Siuta, R. L. (2020). Disclosure of stigmatized identities at work. Journal of Organizational Behavior, 41(2), 169–184.In community political contexts where generalized solidarity is present, both pathways activate — each independently sufficient, their joint presence amplifying the effect.
Generalized solidarity (J₂) provides an additional activation pathway that amplifies disclosure beyond what in-group solidarity alone produces. Both triggers operate; the population of potential supporters is larger.
Bounded solidarity still operates — owners with marginalized identities in high-solidarity Republican contexts disclose at roughly twice the rate of those in low-solidarity contexts. J₂ is not activated; only the in-group mechanism fires.
The disclosure dilemma: the owner must assess whether the solidarity potential outweighs the discrimination exposure risk — and this calculus depends on the community context in a predictable, testable way.
(A ∧ B ∧ C) + (J₁ ∨ J₂ ∨ K) → substantially increases the probability of disclosure
The Dissertation
Coleman (1990) — macro-micro-macro linkages in social explanation
CHAPTER 1 · Archival · Historical Sociology · In Planning
"You can't study Black people without history." — Karida Brown — April 2026 · University of Michigan Sociology
What it establishes
The dual-valence structure is not new to the digital era — Black business visibility has simultaneously produced solidarity benefits and discrimination costs for over a century.
The geography of today's high-disclosure counties mirrors historical patterns of organized Black commercial districts — solidarity infrastructure has deep roots.
Without Ch. 1, the discrimination side of the dual-valence construct is asserted, not demonstrated. The history is theoretical grounding, not background.
Open questions
The chapter examines the 1961–1975 period — the Kennedy-Johnson-Nixon transition — as the founding episode of institutionalized identity disclosure in American business. The Office of Minority Business Enterprise (OMBE, 1969) and Black Enterprise magazine (founded 1970) are the central organizing examples: the public policy infrastructure and the private solidarity infrastructure that emerged simultaneously. This is the moment when disclosure first became both incentivized (AA programs, government contracting) and institutionally contested.
CHAPTER 2 · Theoretical Framework · In Progress · Proposal Submitted
What it establishes
The (A∧B∧C)∧(J∨K) two-stage logic is the minimal form required to handle dual valence — fewer conditions produce false positives or false negatives in predicting disclosure.
J₁ and J₂ are substitutes: the same disclosure decision can be reached via bounded solidarity (in-group) or generalized solidarity (ally-based) depending on community context.
This chapter is the architecture that Chapters 3, 4, and 5 test against — it must be stated before it can be tested across methods.
Status
Proposal submitted. Chapter 2 is the theoretical backbone — it must be stated before Chapters 3, 4, and 5 can test against it.
CHAPTER 3 · Quantitative · OLS/Logit · Analysis Complete
Results — H1 through H4
Disclosure rates increase with county-level Black population share (J₁).
β = 5.286*** — confirmed. ✓
Disclosure rates increase with county-level Democratic vote margin (J₂).
β = 1.411*** — confirmed. ✓
The J₁ effect is stronger in Republican contexts; J₂ attenuates reliance on J₁ in Democratic contexts.
β = −1.541*** interaction — confirmed. Q1/Q2 = 2.0×. ✓
The J₁×J₂ interaction does not predict Veteran-led disclosure (discriminant validity).
Veteran interaction null — confirmed. ✓
Data and quadrant structure
~5M Google Maps listings · 2,681 U.S. counties · ~160K Black-owned businesses
Each cell = Black-owned disclosure rate in counties where that combination of solidarity conditions is present. Q1 = both J₁ and J₂ active; Q4 = neither.
Q1/Q4 = 6.3× total premium · Q1/Q2 = 2.0× J₂ effect
CHAPTER 4 · Qualitative Interviews · 12/35 Complete · IRB Active
What it establishes
What drives within-context variation: two owners in the same high-solidarity county making different choices — a pattern Ch. 3 cannot explain.
Enabling condition C (legitimacy beliefs): not observable in administrative data — only accessible through interviews.
Activation trigger K (identity commitment): when and why K fires independently of J — the trigger that operates outside the market calculus.
Design decisions
CHAPTER 5 · Qualitative Interviews · 34 Interviews Complete · Fieldwork Spring 2026
What it establishes
What the single-identity model cannot handle: competing solidarity claims when J₁ for race conflicts with J₁ for sexuality or gender.
When the enabling/trigger framework produces contradictory predictions for different identity dimensions held simultaneously — the intersectional extension required.
How owners resolve solidarity tensions in real time — observable only through in-situ ethnographic fieldwork.
Open questions
The current plan is to expand Chapter 4 into two contrasting interview studies — a Washtenaw County geography-constant sample and a national Black bookstore owners sample. This chapter contributes the 34 existing interviews as the intersectionality component. The broader Elevate Ten project requires a formal data access agreement and is planned as a post-dissertation extension.
Addressing Common Objections
Isn't there less disclosure simply because there are fewer businesses in certain geographies?
No. Disclosure rates are computed as disclosed businesses divided by total businesses in each identity category per county. Sparse counties with high solidarity still show elevated rates. The variation is in the propensity to disclose among those eligible, not in the raw count.
Does Black-owned disclosure even matter in neighborhoods that are already predominantly Black?
Yes — and this is exactly what the bounded solidarity hypothesis tests. In high–Black-population-share counties, disclosure rates are significantly higher (13.2% vs. 2.1% in low-share counties), consistent with co-ethnic solidarity as an activation trigger even where visibility is less novel.
Additional robustness: the pattern holds controlling for urban/rural, education, income, and industry composition.
Documents that the discrimination side of dual valence has a 150-year material record — visibility has invited retaliation since the Reconstruction era.
Constructs the enabling condition + trigger architecture that frames all subsequent empirical tests.
Tests the community-level predictions at national scale across 150,000+ businesses.
Reveals the within-context individual variation that ecological data structurally cannot see.
Observes the intersectional solidarity tensions that single-identity models miss — in real time, in situ.
The Contribution
The same attribute simultaneously attracts solidarity-seeking customers and invites discrimination. This structure is not captured by existing competitive signaling or social capital accounts.
A two-stage logic separating structural prerequisites (A∧B∧C) from activation conditions (J∨K) — showing that J₁ and J₂ are alternative but reinforcing pathways.
Breadth: ~160K Black-owned businesses across 2,681 counties. Confirms the J₁×J₂ substitution interaction and rules out education, urbanicity, and pre-existing political trends.
Depth: In-depth interviews with Black business owners documenting the deliberate calculus behind disclosure decisions — the individual mechanism behind the aggregate pattern.
Five chapters where each contribution depends on what the others establish. Not parallel descriptions of the same phenomenon at different scales — each chapter's validity requires the others.
We opened with a geographic puzzle — why do Black-owned disclosure rates range from 2% to 13% by county? This dissertation adds the conceptual, theoretical, and empirical apparatus to answer it.
The Research Plan
Chapters 1–3 are fixed. The open question is how to configure Chapters 4 and 5.
Ch. 4 becomes a comparative qualitative design. Addresses denominator problem directly. Ch. 5 extended study moves post-dissertation.
5-chapter structure as proposed. Individual variation held constant in Ch. 4; intersectional in Ch. 5.
Swaps geographic constraint for industry/identity focus. Tests framework portability across markets.
The dissertation as a whole
The Dissertation
Chapter 3 → Job market paper
"Strategic Identity Disclosure at Scale" — quantitative analysis with supplementary interview support from Chapter 4. First paper to model the disclosure decision rather than estimate performance effects.
Chapters 4 + 5 → Individual mechanisms paper
Qualitative study of individual strategic orientations — the within-context variation the quantitative chapter structurally cannot see. Speaks to interpretive sociology and strategy audiences simultaneously.
Full dissertation → Book
The multi-method architecture — historical, theoretical, quantitative, qualitative — is designed for a book-length treatment. The individual papers are the entry points; the book is the synthesis.
Questions for the Committee
Chapter 4 sampling design
I have three plausible paths: (a) complete the Washtenaw County sample as designed; (b) replace it with a national sample of Black bookstore owners; or (c) given Elevate 10's re-engagement for Chapter 5, do both the local sample and the bookstore interviews as Chapter 4, and expand Chapter 5 into a standalone future project. Which design best serves the theoretical goals of the dissertation?
Path to publication
I envision at least two publications from this dissertation: a paper drawing primarily from Chapter 3, and a second drawing from Chapters 4 and 5 together. I also see a longer book-length treatment that synthesizes the full multi-method argument. I'd like the committee's guidance on how to sequence and frame these — and what's the one investment that would have the highest return for the job market.
Kyle McCullers · Ross School of Business · University of Michigan · April 16, 2026
Coming Up
Archival analysis of how Black business visibility has historically functioned as both a site of community empowerment and a target for violence — grounding the dual-valence structure in historical experience.
Historical Sociology · Archival
Black business visibility has been simultaneously a site of community empowerment and a target for violence. This history is not prologue — it constitutes the discrimination side of the dual-valence structure that the theoretical framework requires.
The dual-valence structure in historical form
Greenwood District, Tulsa — June 1921. Visible prosperity made the community a target. Wikimedia Commons / public domain.
What Chapter 1 establishes
Black business visibility has historically produced both solidarity benefits and discrimination costs — this dual-valence structure is not new to the digital era.
Ellsworth, S. (1992). Death in a Promised Land: The Tulsa Race Riot of 1921. LSU Press. | Hahn, S. (2003). A Nation Under Our Feet. Belknap/Harvard.The geographic clustering of high disclosure today mirrors historical patterns of organized Black commercial districts — solidarity infrastructure has deep roots that carry forward.
Walker, J. E. K. (1998). The History of Black Business in America. Twayne/Macmillan. | Butler, J. S. (1991). Entrepreneurship and Self-Help Among Black Americans. SUNY Press.Without Chapter 1, the discrimination side of the dual-valence construct is asserted rather than demonstrated. The history is not background — it is theoretical grounding.
Tilly, C. (1998). Durable Inequality. UC Press. | Seamster, L., & Ray, V. (2018). Against teleology in the study of race. Sociological Theory, 36(4), 315–342.Coming Up
Develops the two-stage (A∧B∧C)∧(J∨K) framework, names the dual-valence construct, and positions it relative to social capital, stigma management, and competitive signaling theories.
Theoretical Framework · Building to the Two-Stage Model
Three prior streams — each essential, none sufficient. The dual-valence structure requires synthesizing all three.
Stigmatized actors manage information strategically — passing, covering, or disclosing depending on audience and context. Disclosure is a form of identity management with social consequences.
Goffman (1963). Stigma. | Yoshino (2006). Covering.Co-ethnic networks generate solidarity resources — trust, referrals, preferential exchange. Disclosure signals group membership, activating these benefits.
Portes & Sensenbrenner (1993). AJS, 98(6). | Coleman (1988). AJS, 94(Suppl.).Firms disclose attributes when credible signals attract value-maximizing exchange partners. Identity tags are low-cost, verifiable signals to specific audiences.
Barney (1991). J. Management, 17(1). | Spence (1973). QJE, 87(3).None of these accounts handles an attribute that simultaneously attracts solidarity and invites discrimination — where net value depends on community composition and political context.
What Chapter 2 establishes
The two-stage (A∧B∧C)∧(J∨K) structure is the minimal logical form required to handle dual valence — fewer conditions produce false positives or false negatives in predicting disclosure.
J₁ and J₂ are substitutes: the same disclosure decision can be reached via bounded solidarity (in-group) or generalized solidarity (ally-based), depending on community context.
The framework must be stated before it can be tested. Chapter 2 is the architecture that Chapters 3, 4, and 5 test against.
Coming Up
~5 million Google Maps business listings. 2,681 U.S. counties. Tests H1–H4 of the enabling/trigger framework using county-level variation in solidarity conditions. Analysis complete.
Quantitative · Google Maps · ~160K businesses
Analysis of ~5 million Google Maps listings across 2,681 U.S. counties. County-level variation in Black population share (J₁ proxy) and Democratic vote margin (J₂ proxy) predicts Black-owned disclosure rates.
Four hypotheses, one framework
Bounded solidarity (J₁): Black-owned disclosure rates increase with county-level Black population share, holding political context constant.
Generalized solidarity (J₂): Black-owned disclosure rates increase with county-level Democratic vote margin, holding Black population share constant.
Substitution: The J₁ effect is amplified in Democratic counties — generalized solidarity provides an additional activation pathway beyond what bounded solidarity alone produces.
Discriminant validity: The J₁×J₂ interaction does not predict Veteran-led disclosure — Veterans follow a different solidarity logic, confirming the racial solidarity mechanism is specific.
From framework to observation
The data confirm this ordering exactly. The substitution logic makes this prediction: J₁ and J₂ are alternatives — having both produces more than either alone.
| Variable | Model 1 | Model 2 | Model 3 | Model 4 |
|---|---|---|---|---|
| Black population share | 5.286*** | 5.016*** | 6.614*** | 6.949*** |
| (0.157) | (0.186) | (0.434) | (0.470) | |
| Democratic vote margin | — | 1.411*** | — | 0.271*** |
| (0.208) | (0.096) | |||
| Democratic county (binary) | — | — | 0.340*** | 0.406*** |
| (0.070) | (0.072) | |||
| Black pop. share × Democratic county | — | — | −1.541*** | −2.191*** |
| (0.452) | (0.426) | |||
| County FE · State FE · Controls: education, income, urbanicity, veteran share. *** p < 0.001. | ||||
Robustness and discriminant validity
Coming Up
Holding community conditions approximately constant (within-metro sampling), what drives within-context variation? Interviews document the decision process, the subjective experience of the disclosure dilemma, and what K (identity commitment) looks like in practice.
Qualitative Interviews · 12/35 complete
Two owners in the same high-solidarity county make different choices. The county-level pattern cannot explain individual variation. Interviews document enabling condition C (legitimacy beliefs) and activation trigger K (identity commitment) directly.
What Chapter 3 cannot tell us
Chapter 3 establishes that community conditions predict aggregate disclosure rates. It cannot explain why two owners in the same high-solidarity Democratic county make different choices.
Personal identity commitment (K): some owners disclose as moral obligation, regardless of perceived solidarity context. Interviews document when and why K fires independently of J — the trigger that operates outside the market calculus.
Legitimacy beliefs (C): what enables an owner to view identity as a legitimate competitive dimension? Enabling condition C is not observable in administrative data. Semi-structured interviews are the appropriate method for capturing this orientation.
Methods: Small & Cook (2023). Sociological Methods & Research, 52(4). | Patton, M. Q. (2015). Qualitative Research & Evaluation Methods (4th ed.). SAGE. | Miles, Huberman & Saldaña (2014). Qualitative Data Analysis (3rd ed.). SAGE.
Interview voices — Ann Arbor
"I didn't realize how radical it was to say, I'm creating a black space for black people by black people for these very explicit uses. I didn't see that as activism at the time, but now what I see it as — my being here, even being a successful entrepreneur, was an act of resistance." — Entrepreneurship educator · Detroit Metro, MI · K: identity as resistance
"I'm going to drive that angle. But then there's also the dangerous side — I'm Jewish, but I'm not going to drive on that. I have a very unpopular opinion, and I care more about humans first than I do my religion. I've been working to bury that somewhat, because everybody who hears Jewish likes to label you as Zionist, or whatever it's going to be, and that can be dangerous for business." — Boutique owner · Ann Arbor, MI · Same person, opposite disclosure across two identities
"I want to win the blind taste test every time. I don't want people to be like, 'Yeah, I liked hers because she was a white girl from Michigan and I identify with that.' I don't care about that. I think that is a crutch to me." — Food & beverage owner · Ann Arbor, MI · C: identity as competitive dimension — rejected
"When we opened we were only 26 — we weren't too much older than all the Koreans in grad and undergrad, and I think they really looked up to Tom and me... Koreans would come in from the suburbs. It was like a cult of party Koreans. And it did play a big role." — Restaurant owner · Ann Arbor, MI · J₁: bounded solidarity through informal co-ethnic visibility
Coming Up
What happens when owners hold multiple simultaneously marginalized identities? The single-identity model cannot handle solidarity tensions — when J₁ for race conflicts with J₁ for sexuality, for instance.
Ethnography · 34 Interviews Complete
Ethnographic fieldwork and in-depth interviews documenting how owners who hold multiple marginalized identities navigate competing solidarity claims — when the mechanisms that activate disclosure for one identity conflict with those for another.
Crenshaw (1989, 1991). | Collins, P. H. (2000). Black Feminist Thought. Routledge; (2019). Intersectionality as Critical Social Theory. Duke. | Wingfield, A. H. (2008). Doing Business with Beauty. Rowman & Littlefield. | McCall (2005). Signs, 30(3).
Post-Dissertation Study · Contingent on IRB + Access Agreement
If IRB approval and a company access agreement are secured, the Elevate 10 site offers a far richer window into the intersectional disclosure dynamics — enabling a multi-source organizational study that extends well beyond the dissertation.
Data sources — extended study
Podcast interview corpus — public-facing content from Elevate 10 entrepreneurial community
Social media analysis — identity disclosure across platforms, co-disclosure patterns
Continued participant observation — deeper field access beyond initial 50+ hours
Additional interviews
Elevate 10 management team — organizational perspective on community identity norms
Entrepreneurs met through the program — following up on initial 34 interviews with deeper access
This scope exceeds the dissertation. The base chapter uses the 34 existing interviews. This extended study is the post-dissertation research agenda.
What Chapters 3 and 4 cannot tell us
Chapters 3 and 4 treat identity tags as discrete and independent. Many owners hold multiple marginalized identities — Black and woman, Black and queer, woman and veteran. The solidarity mechanisms may conflict.
Solidarity tensions: when J₁ for race and J₁ for sexuality point in opposite directions (e.g., a religious Black community with anti-queer norms), the enabling/trigger framework requires an intersectionality extension that handles competing triggers.
This ethnography provided the opportunity to study these decisions in real time. The fieldwork setting — a multi-identity entrepreneurial community — makes intersectional disclosure conflicts observable in context.
Field voices
[Placeholder — field observation: a moment when an owner navigates competing solidarity claims in real time, e.g., a community event where racial solidarity and queer solidarity norms conflict around a disclosure decision.] — Field note, Spring 2026
[Placeholder — interview: owner describing the experience of holding multiple identity tags and which ones they chose to disclose and why.] — Interview, Spring 2026 · Multiple-identity owner
[Placeholder — community member or peer entrepreneur commenting on another owner's disclosure choices. Illustrates the social enforcement of solidarity norms from outside.] — Field note, Spring 2026
[Placeholder — owner describing a situation where disclosing one identity required concealing another. Illustrates the tension between competing legitimacy frameworks.] — Interview, Spring 2026
[Placeholder — owner who disclosed multiple identities simultaneously and the response they received from different segments of their customer base.] — Interview, Spring 2026 · Dual-tag discloser
[Placeholder — advisor or mentor in the entrepreneurial community describing how they counsel owners on disclosure when identities conflict.] — Interview, Spring 2026 · Community leader
Building to the framework
Stigmatized actors manage information strategically — passing, covering, or disclosing depending on audience and context. Disclosure is a form of identity management with social consequences.
Goffman, E. (1963). Stigma: Notes on the Management of Spoiled Identity. Prentice-Hall. | Yoshino, K. (2006). Covering. Random House.Co-ethnic networks generate solidarity resources — trust, referrals, preferential exchange — for in-group members. Disclosure signals group membership, activating these benefits.
Portes, A., & Sensenbrenner, J. (1993). American Journal of Sociology, 98(6), 1320–1350. | Coleman, J. S. (1988). American Journal of Sociology, 94(Suppl.), S95–S120.Firms disclose attributes when credible signals attract value-maximizing exchange partners. Identity tags are low-cost, verifiable signals to specific audiences.
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120. | Spence, M. (1973). Job market signaling. Quarterly Journal of Economics, 87(3), 355–374.None of these accounts handles an attribute that simultaneously attracts solidarity and invites discrimination — where the net value depends on community composition and political context. The framework synthesizes all three streams to address this.
Backup — Framework Detail
Disclosure ↔ ( A ∧ B ∧ C ) ∧ ( [J₁ ∨ J₂] ∨ K )
| Condition/Trigger | Definition | Observable proxy (Ch. 3) | |
|---|---|---|---|
| A | Marginalized identity | Identity historically stigmatized or discriminated against | Google Maps identity tag options |
| B | Positioning discretion | Owner-operator with brand agency | Owner-operated vs. franchise |
| C | Legitimacy beliefs | Views identity as legitimate competitive dimension | Interview-only (Ch. 4) |
| J₁ | Bounded solidarity | Co-ethnic in-group reciprocity | County Black population share |
| J₂ | Generalized solidarity | Values-aligned ally support | County Democratic vote margin |
| K | Identity commitment | Moral obligation regardless of market return | Interview-only (Ch. 4) |
Backup — Data Sources
Business listings with identity tags scraped 2021–2022. ~160K Black-owned businesses across 2,681 U.S. counties. Tags are owner-elected and public-facing.
New York–Newark MSA listings tracked across three years. Identifies timing of tag adoption. Enables temporal analysis around the 2020 Floyd protests.
State licensing records matched to Google Maps listings. Provides verified ownership status and a high-theory-relevance subsample. Tests whether J₁×J₂ holds with verified race.
ACS 5-year estimates (education, income, urbanicity, racial composition). MIT Election Lab (2016, 2020 presidential vote). SBA county business patterns (industry controls).
Backup — H3 Interpretation
Holding Black population share constant (both high-Black counties), the political context doubles disclosure. Direct test of generalized solidarity (J₂) as amplifier — bounded solidarity is held constant.
Bounded solidarity still operates in Republican counties — it is attenuated by the absence of J₂, not eliminated. High-Black-share Republican counties disclose at 3.2× the rate of low-Black-share Republican counties.
The negative interaction (−1.5 to −2.2) reflects that the marginal return to Black population share is lower in Democratic counties — because J₂ already provides a disclosure-enabling baseline. This is the substitution story: having J₂ reduces (but does not eliminate) the additional lift from J₁.
Backup — Chapter 4 Methods
Within-metro sampling in Washtenaw County — holding community conditions approximately constant to isolate individual-level variation. 35 interviews total; 12 complete as of April 2026.
Purposive sample: 4 cells crossing disclosure status (disclosed / not) × identity type (Black-owned / Women-owned). Oversamples non-disclosers — harder to reach, analytically more informative.
60-minute semi-structured interviews. Covers: business history, identity and positioning, awareness of Google Maps tagging, decision process, perceived risks and benefits. IRB active.
Thematic coding for enabling conditions (A, B, C) and activation triggers (J₁, J₂, K). Secondary coding for disclosure dilemma framing and solidarity orientation. Atlas.ti.
Backup — Women-Owned Pattern
Women-owned disclosure shows higher rates in urban cores across political contexts, less sensitive to political vote margin. This supports the identity-specificity of the solidarity framework: gender solidarity operates differently than racial solidarity, and J₂ (generalized solidarity) does not map cleanly onto partisan politics for gender in the way it does for race.
Backup — Chapter Status Table
| Chapter | Title | Method | Status |
|---|---|---|---|
| Ch. 1 | Historical Roots | Archival | In Planning |
| Ch. 2 | Theory | Theoretical | In Progress — proposal submitted |
| Ch. 3 | Quantitative | OLS/Logit | ✓ Complete — manuscript in revision |
| Ch. 4 | Interviews | Qualitative | 12/35 complete; IRB active |
| Ch. 5 | Ethnography | Ethnographic | Round 1 interviews complete; fieldwork ongoing |
This is the proposal defense — the committee is approving a design, not five finished chapters. Ch. 3 provides the empirical anchor; Chs. 1, 2, 4, and 5 are at varying stages of completion with clear feasibility paths.
Backup — Limitations
Selection: The denominator is disclosed businesses on Google Maps — not all Black-owned businesses. Owners who chose to use Google Maps at all are not a random sample of Black entrepreneurs.
Endogeneity of Black population share: High concentration may be endogenous to the same political-economic conditions that predict disclosure. IV strategy under development using historical railroad routes (Sequeira et al. 2020).
Platform specificity: Google Maps tags are a particular form of digital disclosure. Platform-specific effects cannot be fully separated from the solidarity mechanisms. Future work: Yelp replication.
Enabling condition C unobserved: Legitimacy beliefs are not observable in administrative data. Ch. 4 interviews fill this gap — but Ch. 3 assumes C is approximately distributed uniformly across communities.