November 28, 2022

Critical Incidents, Invisible Populations, and Public Policy: A Case of the LGBT Community

by Roddrick Colvin, Ph.D.


This exploratory research paper discusses important issues in public policy and service delivery, critical incidence analysis, and invisible communities. Using the lesbian, gay, bisexual, and transgender (LGBT) community as a case study, I raise several questions about critical issues, including: How does critical incident analysis fit into study of public policy and how do we understand invisible communities in critical incident analysis, and what next steps are needed to improve critical incidence analysis with regards to invisible communities? In order to improve our understanding of these issues, I recommend a deeper study of social construction of target populations, more systematic data gathering on invisible communities, and increased media accountability and standards.


This paper considers critical incidents and invisible communities in the context of public policy and service delivery. Specifically, it seeks to explore issues that arise when groups of people are invisible to the broader community, including policymakers and service providers for that community.  When these communities are invisible to policymakers and service providers, they are not included or considered in the policy or planning process. Furthermore, these communities can be over-looked during critical incidents and other emergency situations.  For example, hospital patients were haphazardly evacuated after Hurricane Katrina partly due to that fact that they were invisible to planners and emergency managers.   Frequently, policymakers and service providers conceptualize communities or populations in terms of majority members, without consideration of the full diversity of communities or populations. There are numerous factors that help to explain community or population invisibility.  In this paper, I focus on three factors: social taboo, community norms, and institutional barriers.

Critical Incidents in Public Policy: Finding a Common Language

A critical incident is typically defined as “a relatively brief occurrence involving injury, loss, conflict, discovery or change of significant proportion, usually unscripted and unanticipated, with the potential to alter existing societal norms. Critical incidents are usually traumatic, threatening the bonds of trust that bind communities, but may be positive, initiating historic consequences” (ACIA, 2008).

______________________________________________________________________________About the Author

Roddrick Colvin is an Associate Professor in the Department of Public Management at John Jay College of Criminal Justice. His research interests include employment policy, gay rights, hate crimes, and international human rights policies. He currently teaches courses in public administration, human resources management, and policy analysis.

There are at least two common concepts in public policy that relate to critical incidents: trigger mechanisms and potential focusing events. In both cases, the event garners the attention of policymakers and moves an issue or problem to the institutional agenda. Thus, in public policy, the outcome or residual consequence of a critical incident can be policy changes or changes in administrative operations designed to prevent the incident from re-occurring.

Trigger Mechanisms

Cobb and Elder (1983) identify several unpredictable events, or “trigger” mechanisms. Trigger mechanisms include natural disasters, unanticipated human events, technological changes, societal imbalance, and ecological changes. Such events have incredible agenda setting ability. The rare and unanticipated nature of trigger mechanisms makes them news and agenda-worthy. Once on the agenda, advocates and interests groups attempt to frame and re-frame the problem and the universe of possible solutions to the problem.

Focusing Events

Birkland (1997) builds in the idea of trigger mechanisms but develops a more precise definition called “potential focusing events (PFEs).” A PFEs is “an event that is sudden, relatively rare, can be reasonably defined as harmful … that is known to policymakers and the public virtually simultaneously” (1997, p. 22). These unexpected or surprising events draw the attention of policymakers and media.

From these definitions, we can see some common themes. First, mechanisms, incidents, and events, emphasize the unplanned nature of the occurrence. Second, awareness of the occurrence among different stakeholders happens within a very short period of time, and generally (although not necessarily the case) the incident is viewed as harmful or negative. In the case of critical incidents, the event may change relationships and/or perceptions in civil society. Based on these definitions, we can imagine a host of events as potentially critical, including:  plane crashes, oil spills, floods, and/or earthquakes.

In the discipline of public policy, it is important to note that both trigger mechanisms and focusing events have a direct connection to and influence on public policy. The effects of these occurrences include: when an issue captures the attention of policymakers; how policymakers, the media, and the public come to understand the issue; who the major and minor stakeholders are surrounding the issue; and what the “universe” of possible outcomes or policy solutions exist to address or remediate the issue. In essence, trigger events and focusing events help to frame our understanding about an incident. The lessons learned (or experience gained) during and after the event shape policy solutions pursued by government (Birkland, 2006).

In addition to the trigger mechanisms and focusing events as critical incidents-related concepts, social construction of target populations is an important concept for understanding critical incidents in the public policy universe.

Social Construction

Schneider and Ingram (1993) use a social construction framework to explain how policy benefits and burdens are distributed. The social construction of a target population refers to the recognition of shared characteristics that distinguish a target population as “socially meaningful, and the attribution of specific, valence-oriented values, symbols, and images to the characteristics” (Schneider and Ingram, 1993, p. 335). In other words, the social construction of a target population is greatly influenced by our societal constructs surrounding the particular population or community at hand.

The theory of social construction of the target populations holds that policy design and the alternative policy solutions can be broadly predicted by understanding the social construction and political power of the groups targeted by a given policy. The social construction perspective helps to explain the unequal distribution of policy benefits and burdens. For Schneider and Ingram (1993) a group’s political power and the perception of a group’s social construction range from strong to weak and from positive to negative, respectively. The convergence of political power and social construction creates four target populations: advantaged, contender, dependents and deviants.  In terms of development, Carney (2010) suggests a bi-directional relationship between policy and social construction. In some cases, state action or policy might influence the social construction of populations.  For example, a population generally perceived as deviant, might be re-cast as dependent depending on the policy debates and initiatives. Undocumented immigrants in the United States is an appropriate example.

When considering critical incidents in public policy, social construction of target populations can be an important concept. It might be especially useful when thinking about invisible populations. For example, undocumented immigrants might be categorized as dependents or deviants. They are not politically strong and have a generally negative connotation. Such a designation would help to explain some of the policies that are designed to punish or penalize this population.

Collecting Data and Information on Critical Incidents in Invisible Communities

In most cases, the methodology used to collect data about critical incidents is archival research, with the intention of observing the “incident” and the subsequent policy change. This change can be an actual change in the law or policy (from any branch of government), a change in agency rules or operation procedures, or a change in behavior by affected actors. Researchers typically tracked an issue via the legislative, executive, and judicial routes to better understand the role of the event or incident have on the policy, planning or service delivery within government (Colvin & Riccucci, 2002; Mintrom, 2001; Mooney, 2001). Some researcher have also gathered data via the public administration (rule, procedural, and regulatory change), in an effort to determine the role of incidents (Kerwin, 2003; O’Leary, Yandle, and Moore, 1999).

While most of the research using these data have been qualitative, a number of studies have employed quantitative analysis, namely, event history analysis ((Berry & Berry, 1990; Hays & Glick, 1997; Klawitter & Hammer, 1999). Event history analysis (EHA) is the study of events, the duration of time between events, and the probability of events occurring at selected points in time or under certain sets of conditions (Barton and Pillai, 1995). The goal of event history analysis is to explain a qualitative change – an “event” – that occurs in the behavior of an individual or group at a particular point in time (Berry and Berry, 1990). This methodology allows us to estimate the probability of policy adoption in any given period of time, depending on a number of factors, including adoptions in previous periods. In terms of policy, event history analysis can help to predict the likelihood of an incident in a population.

Event history analysis is an ideal method for analysis when there is a visible community and a plethora of data that can be analyzed. While regression considers just one ‘point-in-time’ for analysis, EHA can consider variations in the measures and their changes over time. However, if the community under consideration is not accessible, then the analytical power of EHA is lost. That is to say, if you cannot observe that community, then you cannot observe an event within that community.

Topography of Invisible Populations and Communities

Whether providing social, health, or police services, the organizations delivering those services must know something about the community to be served. They must be able to identify and communicate with individuals and groups inside of the community. No doubt, service delivery to some communities is much easier than service delivery to other communities. Oftentimes, the ability and level of service delivery is based on the ‘visibility’ of that community and its members. The most difficult communities are communities that are invisible to providers.

There are several factors that can contribute to a communities’ visibility, or lack thereof. These factors fall into three general categories: social taboo, community norms, and institutional barriers. These three factors tend to contribute to community invisibility. In some cases, communities engage in active efforts to remain invisible. For example, individuals and communities who engage in activities considered to be socially taboo and/or illegal might attempt to remain invisible. This might include: closeted lesbian and gay people, undocumented immigrants, drug addicts, sex workers and gang members. In each of these cases, the members of the community are engaging in activities that isolate then from the larger society. Since their activities are considered taboo (and in some cases are illegal) by most in the majority of the population, there is incentive to stay away from the police and other public service providers.

Second, individuals and communities who engage in activities outside of community norms can also be often rendered invisible. For example, homeless people, alcohol abusers, and teen mothers. In terms of community norms, the activities of these members are not necessarily ‘negative’ or illegal, but are often considered outside of the range of acceptable behaviors. As Schneider and Ingram (1993) note, the social construction of populations can affect the level of service delivery for various populations. In this case, the ‘less-than-positive’ social construction would adversely affect service delivery for these populations.

Finally, institutional barriers might render individuals and communities as invisible. There are institutional barriers in policies or procedures that restrict information and data about some population. For example, there are no known representative public surveys that ask about the sexual orientation of the respondent, thus we have very little data about the lives of gay and lesbian people.

Layered on top of invisibility are other factors which may add to the difficulty of effective delivery of public goods and services.  One factor is the level of community permeability. Some communities might be considered ‘closed’ and others more ‘open.’ Closed communities are an added challenge for service providers because non-members of the community are not able to easily enter and exit.  This might be due to language barriers, divisions in culture or belief, or geographic isolation. For example, immigrant communities, separatists’ communities, or religious sects could be considered closed, and thus a hinder to proper service delivery. Open communities are more easily accessible by non-members. In contrast, while largely invisible, it fairly easy for non-members and service providers to enter and exit the lesbian, gay, bisexual, and transgender (LGBT) community’s common institution (i.e., community centers, restaurants, bars, other businesses, and churches).

The second factor that can add to the difficulty of effective service provision is the level of vulnerability. Many communities lack political influence, financial resources, and/or numerical size, which are often needed to extract public services at an appropriate level. These communities may have a history of being denied public support, which many have also contributed to their current vulnerability. For example, people with mental or physical disabilities were neglected in terms of policy-making, planning, and service delivery for much of modern American history. Only with the interventions of the courts and the passage of the Americans with Disabilities Act did tangible and measurable services flow to these communities.

From a public perspective, delivering services to invisible, closed, and vulnerable communities would be the most challenging. Foreign-language immigrants without proper documentation are exemplar of such a community. The language barrier makes the community closed. Their immigration status makes them invisible (both as a social taboo and institutional barrier). Since they also lack political influence, resources, or a measurable population (undocumented workers might have numerical size, but there is no reliable way to measure), they are also vulnerable.
Table One: Community Factors Related to Invisibility

LGBT People as a Case Study

In order to better understand the challenges faced when researching critical incidents and invisible communities, we can consider the lesbian, gay, bisexual, and transgender (LGBT) community as a case study. Unlike members of a racial or ethnic community, LGBT people are not necessarily identifiable upon sight. Unlike religious enclaves or immigrant communities, LGBT people do not necessarily live in one community or geographic area. LGBT people are integrated into every part of broader society. They can be found among the wealthy and affluent, among the racial, ethnic, and religious in our society, among the elderly and youth, and among the poor and destitute. Some LGBT people are out, known, and visible as sexual minorities. Others are closeted, unknown, and invisible within the LGBT and larger community. Overall, we can classify the LGBT community as largely invisible, open, and somewhat vulnerable.

Past research of this community and critical incidents analysis has been built on the premise that all the members of the community were ‘out’ or open about their sexual orientation or gender identity. For example, if we consider the following common critical incidents: hate crimes, civil rights discrimination (in employment, accommodations, housing, credit, education, etc.), bullying in schools, and sexual harassment, we would discover the common assumption that the “victims” were visible and targeted base on that visibility. For examples, in an anti-gay hate crime, the sexual orientation of the victim has to be assumed by the perpetrator, witness, responding officer or the victim. Without this assumption, the likelihood that the crime would be categorized as a hate crime is low.  Complicating this situation, a victim may decide to not report an incident because of social, economic or privacy concerns.

We have an example of how this community as well as policymakers and other stakeholders react to critical incidents when the victim is visible in the community: the death of Matthew Shepard in 1998. Shepard was a gay student at the University of Wyoming who was attacked and killed near Laramie. His murder was determined to be based on his gay sexual orientation, and as a result, brought national attention to the issue of hate crimes.

The response to the Shepard death – in many ways – represents a classic example of agenda setting and critical incident reporting. The horrific nature of Shepard’s death, the apparent gay and anti-gay undertones, internal community outrage, and the high level of media attention, helped to put this incident into the public discourse, and resulted in policy changes.

More interesting than the attention that the Shepard incident generated is the attention other similar incidents did not generate. In 1993, five years before Shepard’s death, Brandon Teena, a transgender man, was murdered in a hate crime. Few people had heard of Brandon Teena before the 1999 film about his life. In 2003, Sakia Gunn, a 15-year old lesbian was fatally stabbed in Newark by a man for rebuking his advances. Even fewer people have heard of Sakia Gunn. In 2003, a transgender woman, Bella Evangelina, was denied medical care by emergency responders after a car accident in Washington, DC.  In 2002, two young lesbians, Stephanie Thomas and Ukea Davis, were beaten and stabbed in Newark after being seen holding hands in a public park.  A Lexis-Nexis review of articles published about recent sexual orientation and gender identity-related murders highlights the disparity in reporting about these crimes.

Table Two: Newspaper Coverage of LGBT Murders

Name Year of Incident Number of Articles

Matthew Shepard



Brandon Teena 1993 969
Elvys Perez (Bella Evangelista) 2003 104
Sakia Gunn 2003 30
Stephanie Thomas/Ukea Davis 2002 5

Even accounting for the accumulation of articles over time, the number of published articles for each murder suggests a large disparity.

Equally interesting are the (low) numbers of known LBGT related deaths reported in prior and subsequent years in the United States. According to the National Coalition Anti-Violence Programs (NCAVP) 2008 Annual Hate Violence Report, consistent and reliable data about gay-related is incomplete at best (Sklar, 2008).

Table Three: Known LGBT-related Murders

The disproportional number of newspaper article about Matthew Shepard, and the very low numbers of known gay-related murders over time creates some interesting conundrums. First, why did hate-related murders similar to Shepard’s receive very little media attention or invoke any policy changes? Second, given that these incidents occurred to visible and invisible members of the LGBT community, what mechanisms are needed to gather data on the invisible members of the community? Third, what roles do race, socio-economic status and social construction explain the varied public responses? Finally, do communities self-regulate in incidents and influence whether they become ‘critical’?

Does Social Construction Offer An Explanation?

As noted earlier, a social construction framework can be used to help explain how policy benefits and burdens are distributed. Some authors have applied this understanding to target populations to help explain media attention and action (or lack thereof) by policymakers. In this case, I believe it can be used to help us understand the variations in media attention and service delivery in the LGBT community. For example, social constructionists might argue that Shepard was of a different social construction than the other victims. Shepard’s identity as a young, white gay college student in a rural community would have given him a generally positive social construction (Orr & Aoki, 2002). In contrast, Elvys Perez’s identity as a transgender person of color casts her in a negative social construction or ‘deviant.’ Both inside and outside of the community, social construction can be applied. From the outside, media and policymakers might be hesitant to take on the cause of a transgender woman. From the inside, leaders and community members might be hesitant to highlight a member to the community who is not positively constructed. Thus, it is possible that forces inside and outside of the LGBT community contributed to the invisibility of some members, but not others.

Methodological Issues and Sexual Orientation

There are two main challenges that have hindered critical incident analysis in LGBT-related incidents: data sources and disclosure of sexual orientation.

Paucity of Data and Information About LGBT people

First, because there is so little reliable information about gay people in general, the data sources continue to be inadequate. Very few representative samples that capture sexual orientation information exist. Many institutions gather data from populations within the larger LGBT community. For example, there are representative studies that focus on LGBT youth or LGBT smokers, but not a broad representative sample from the entire LGBT population. Currently, there are two common sources of data: the US Census and the General Social Survey (GSS). However, neither the US Census, nor the GSS gather data in ways that would be helpful for critical incidence analysis.

In terms of the US Census, sexual orientation is not a separate question (like, race or sex), instead it is an interconnected question related to householder and his/her relationship to other adults. That is to say, instead of just asking census-takers to identify their sexual orientation, they ask about the adults and the relationships between them. Of course, such a system is flawed because it renders all single, not coupled gay men and lesbians invisible.

In terms of the GSS, sexual orientation is based on the sex of the sexual partners, not a self-identification question. For researchers using this dataset, he or she must make a determination about how many same-sex sexual partners are needed to determine sexual orientation. In other words, if a male GSS survey-taker has two heterosexual and homosexual encounters, he could be labeled has gay, straight, or bisexual, depending on the researcher’s determination. The determination by the researcher can have real implications the pool of data. For example a more restrictive determination would limit the size of your population.

Managing Disclosure

Unlike gender and race, sexual orientation is not necessarily an observable characteristic. This presents research challenges that often do not exist for other populations. For example, in the workplace, employees would have to disclose their sexual orientation in order for it to have an influence on earnings, discrimination or other work-related factors. Without such disclosure, we have little evidence to assume relations between sexual orientation and employer actions. Of course, disclosure could be influenced by a number of factors as well, including; career choice, length of time in career, future career options, socio-economic status, industry, and overall workplace climate. Thus, with the option to disclose sexual orientation, workers may be able to better manage the discriminatory actions in the workplace. From a research perspective, measuring discrimination or its effects becomes more difficult due to many variations in the population. Some workers may determine that disclosure of sexual orientation will have a positive discriminatory effect on earnings, whereas another employee may determine that the disclosure would have a negative effect. Furthermore, some researchers have suggested the sexual orientation itself might influence career decisions. Such workers would “pre-sort” along occupations and work environments.

We can extrapolate the disclosure issues to other environments as well. Some students will manage their identities in order to minimize bullying in school. Some transgender people will relocate and hide their transition in an effort to reduce the likelihood of hate crime victimization. Alternatively, many people many choose to be open about sexual orientation or gender identity because they perceive a more receptive or safe environment. Regardless of the situation, individual management and disclosure of information make gathering and analyzing about sexual orientation very difficult. Because of these challenges, the population remains largely invisible and inaccessible.


Incentives for Systematic Data Gathering on Invisible Minorities

CIA depends on the availability of data and information. With LGBT and other populations, there is often little or no reliable data about the community. Organizations that collect representative data should be encouraged to expand the universe of questions asked. For example, a question on the US Census about sexual orientation at the individual level would provide a rich source of data about LGBT people, including; socio-economic, race, education, income, and housing status. Furthermore, incentives should be used to encourage participation from members of these invisible communities. For example, gay and lesbian people might be more willing to disclose information about hate crimes (including murders), if they could be assured a minimum level confidentially.  Moreover, sex workers might be more willing to participate in data-gathering efforts, if financial incentives are part of the recruitment effort. Monetary and non-monetary incentives have been a proven technique in other invisible communities. For example, health departments across the United States instituted needle exchange programs in order to improve the health outcomes for intravenous drug users. In order to find out how to best serve this community, health officials talked to them in strict confidence about their needle-sharing habits and other challenges they faced. The health officials had to put aside the moral or ethical issues association with illicit drug use, and instead focused on building trust, which lead to better access and improved service.

Media Accountability and Standards

Central to CIA is the role of the media in informing the public and policymakers about events and incidents. When the media uses random or arbitrary mechanisms for deciding what is newsworthy, many items that the public cares about are not reported. More consistent standards would open the now-closed media system to a broader array issue of issues and communities. For example, a basic standard that hate crimes will be reported with emphasis on protecting the identity of the victim would help to bring invisible communities’ issues in to the light. If people were reminded that anti-black, anti-gay, and anti-Semitic hate crimes occur every day and are three most common areas, perceptions about such crimes would change. One area where this has been the case, is in the reporting of soldier deaths in the wars in Afghanistan and Iraq. In this case, most of the major newspapers keep a running tally of the names of fallen soldiers. The effect this does not necessarily create a critical incident, but it does keep this in the mind of the reader and, it creates a source of data that can be archival and reviewed for future searches and critical analysis.

Analysis of Social Construction in Critical Incidents

Understanding the social construction of populations will provide valuable insights into CIA. Critical incidents do not just happen to “advantaged” population in society, but these populations can dominate the research agenda. However, the other populations are worthy of similar research attention too. This includes ‘deviants’ where most of the invisible individuals and communities reside. Some of this work involves trying to remove some of the social taboos associated with the ‘deviant’ communities. Communities must also be aware of social construction and how members within their community are perceived internally and externally. Individuals and members who are invisible to the internal community and invisible to the external community are especially vulnerable to being overlooked in CIA and other analytic processes. Furthermore, CIA can begin to analysis and understand the role of government in crafting or aid in crafting social construction, especially among the most disenfranchised groups.


This exploratory research starts the conversation about CIA and invisible populations. From a public policy and service perspective, it is difficult to develop appropriate responses to critical incidents, if valid and reliable information about the community is not available. For example, without information about undocumented immigrants, sex workers, or drug abusers, it is impossible to create serious health, safety or protective policies to support community members.

Using sexual orientation as a case study, at least three major issues were identified, including: lack of systematic data collection about gay people, disclosure of sexual orientation among individuals, and social construction inside and outside of the LGBT community. Finally, recommendations for better information and data gather about LGBT and other invisible communities, including: investment in systematic data collection, holding media accountable for variations in coverage, and additional research into social construction and critical incidence analysis.


Barton, Thomas R. and Pillai, Vijayan K. (1995). “Using an Event History Analysis to Evaluate

Social Welfare Programs.” Research on Social Work Practice, 5 (2), 176-194.

Berry, Frances Stokes and Berry, William D (1990). “State Lottery Adoption as Policy

Innovation: An Event History Analysis.” American Political Science Review. 84 (2), 395-415.

Birkland, Thomas (1997). After Disaster: Agenda Setting, Public Policy, and Focusing Events.

Washington, DC: Georgetown University Press.

Birkland, Thomas (2006). Lessons of Disaster: Policy Change After Catastrophic Events.

Washington, DC: Georgetown University Press.

Carney, Gemma (2010). “Citizenship and structured dependency: the implications of policy

design for senior political power.” Aging & Society, 30 (2), 229-251.

Cobb, Roger. W. & Elder, Charles (1983). Participation in American Politics: The Dynamics of

Agenda-Building. 2nd ed. Boston: Allyn & Bacon.

Colvin, Roddrick & Riccucci, Norma (2002). “Employment Nondiscrimination Policies:

Assessing Implementation and Measuring Effectiveness.” International Journal of Public Administration, 25(1), 95-108.

Hays, Scott P. and Henry R. Glick (1997). “The Role of Agenda Setting in Policy Innovation: An

Event History Analysis of Living-Will Laws.” American Politics Quarterly, 25, 497-516.

Klawitter, Marieka & Hammer, Brian (1999). Spatial and Temporal Diffusion of Local

Antidiscrimination Policies for Sexual Orientation in Gays and Lesbians in the

Democratic Process (Ellen Riggle and Barry Tadlock, editors). New York: Columbia University Press.

Kerwin, Cornelius (2003). Rulemaking: How Government Agencies Write Law and Make Policy.

3rd Edition. Washington, DC: CQ Press.

Mintrom, Michael (2001). “Policy Design for Local Innovation: The Effects of Competition in

Public Schooling.” State Politics and Policy Quarterly, 1(4), 343-363.

Mooney, Christopher (2001). “Modeling Regional Effects on State Policy Diffusion.” Political

Research Quarterly, 54(1), 103-124.

O’Leary, Rosemary, Yandle, Tracy & Tamilyn Moore (1999). “The State of the States in

Environmental Dispute Resolution.” Ohio State Journal on Dispute Resolution, 14(2), 515-614.

Ott, Brian & Aoki, Eric (2002) “The Politics of Negotiating Public Tragedy: Media Framing of

the Matthew Shepard Murder.” Rhetoric & Public Affairs, 5(3), 483-505

Schneider, Anne & Ingram, Helen (1993). “Social Construction of Target Populations.”

American Political Science Review, 87(2), 334-347.

Sklar, Roberta (2008). NYC AVP reports increase in severity of anti LGBT violence in latest

Hate Violence Report. National Coalition Anti-Violence Programs (NCAVP). Annual Hate Violence Report. Access on October 23, 2009.