Who Files Discrimination Charges?

Donald Tomaskovic-Devey & Carly McCann

Download Full Article PDF

It has been sixty-five years since the 1964 Civil Rights Act outlawed employment discrimination and created the U.S. Equal Employment Opportunity Commission (EEOC). The political focus at the time, of course, was discrimination against African Americans. The 1964 law, however, extended these protections to all races, national origins, skin colors, religions, and genders. Shortly thereafter protections were extended to older workers, and in 1990 to people with disabilities. Although we are many decades past the moments where these bases for discrimination in employment were made illegal, we know surprisingly little about discrimination charges. 

In this article we ask who files discrimination charges, comparing charge rates on the bases of sex, race, national origin, age, and disability.1 We discover that disabled workers, followed by African Americans, are most likely to respond to experiences of discrimination by filing formal charges. We extended these comparisons to observe state variation in charge rates, identifying a set of states --Alabama, Arizona, Delaware, Indiana, the District of Columbia, Missouri, Nevada, and New Mexico-- that generate particularly high rates of discrimination charges on multiple dimensions and one --New Jersey-- that generates low levels of recourse to legal redress.

The EEOC and State Fair Employment Practice Agencies (FEPA) are the regulatory bodies authorized to receive and adjudicate discrimination complaints against employers. Under the Civil Rights Act of 1964 a person who believes they have been discriminated against in employment must file with either the EEOC or a state FEPA before pursuing legal action against their employer. While we do know from past research that most acts of discrimination do not lead to formal charges, we do not know much more. The purpose of this article is primarily descriptive. We have two main findings. First, we document for the first time variation in the rate of discrimination charges for the most prevalent forms of employment discrimination. Second, we also document that for all protected bases of employment discrimination there is substantial variation in the rate of formal charge filing between U.S. states. Although we make some initial descriptive comparisons and review some plausible explanations, we do not endeavor to explain this variation, which we leave for future research.


Comparing the bases of discrimination

While the EEOC routinely reports the number of discrimination charges it and state FEPAs receive at eeoc.gov, they do not produce any standardization to allow us to gauge how widespread discrimination complaints are. In this article we look at all of the sex, race, national origin, disability, and age discrimination charges filed with the EEOC or state FEPAs between 2012 and 2016 and standardize them by the number of employed people at risk for that form of discrimination. 

Figure 1 reports the proportion of all discrimination charges received by the EEOC or state FEPAs between 2012 and 2016, broken down by protected category. Race based charges are the most common, but sex, disability, and age charges are not far behind. Religion and color based charges are much less common.3  We can make partial comparisons to the proportions of claims in 2001 reported by Wakefield and Uggen (2004). The proportion of all claims based on race (35%), sex (30%) and age (22%) were roughly comparable in the early 2000s and more recently. In contrast disability based claims grew from 21% to 27% of all charges.

These proportions do not tell us much about the rate at which employees experience or report discrimination. When people perceive that they have been treated unfairly, they first have to make sense of why, be aware of their rights, and then take action (Miller and Sarat 1980; Festiner 1980). In 2016 the General Social Survey asked respondents if they felt that they had been discriminated against at work in the last five years. Nine percent of respondents reported having experienced discrimination. Of these age (24%), race (21%) and sex (14%) discrimination were the most common, while religion (1%) and disability (3%) were relatively uncommon.4   Compared to the distributions of complaints to the EEOC, the incidence of perceived age, race, sex and religion discrimination are quite close to their proportion of discrimination charges.5  Disability discrimination claims stand out for being much more common than the perception of disability based workplace discrimination. This suggests that the disabled have stronger rights consciousness or access to law than other groups.


Figure 1: Proportion of Charges Recieved by the EEOC and State FEPAs, 2012-2016

To understand the relative rate of filing discrimination charges we standardize by population size. Figure 2 reports the same discrimination charges to the EEOC and FEPAs broken down by protected group and standardized relative to every 100,000 employees. Discrimination charges associated with disability (429 per 100,000 workers) and being African American (195 per 100,000) are much higher than those filed by others groups. Discrimination charges filed on the basis of national origin (50 per 100,000) gender (38 per 100,000) and age (35 per 100,000) are less common. Finally, race discrimination charges filed by Asians (19 per 100,000), Hispanics (7 per 100,000) and Whites (1 per 100,000) and sex discrimination charges filed by men (5 per 100,000) are rare.

Figure 2: Discrimination Charge Rate Per 100,000 Employees, 2012-2016

Filing a discrimination charge is quite different from experiencing discrimination. Past research suggests that less than 1% of people who experience discrimination actually file a discrimination charge (McCann, Tomaskovic-Devey and Badgett 2018; Berrey, Nelson,and Neilsen 2017).  When employees perceive that they have been treated unfairly, they first have to make sense of why, be aware of their rights under the law, and then take action. Scholars conceptualize the process of filing a case in terms of a dispute pyramid in which many people experience grievances, some of whom come to recognize their treatment as violating the law and are also able to identify the source of that violation, and very few move on to actually make a legal claim (Miller and Sarat 1980; Festiner 1980). Individuals vary in terms of their awareness of their legal rights (Burstein and Monahan 1986; Hirsh and Kornrich 2008) and access to the resources necessary to access the legal process (Berrey et al. 2017). 

Past research also makes clear that complaining to an employer about discriminatory treatment can be quite dangerous, often leading to employer harassment, retaliation, and firing (Roscigno 2007). In addition, only a small proportion of discrimination charges result in some benefit to the employee making the claim. Those benefits tend to be small monetary payments, coupled with loss of employment and no change in workplace practices or even admission of fault by the employer (Berrey et al. 2017). Simply complaining about unfair treatment can be quite risky, often leading to job loss. Legal action can be even more difficult to contemplate and undertake.

Since individuals who file charges first need to understand their right to do so, the higher rates of charging by disabled workers and African Americans suggests that they encounter more discrimination, but also that they are potentially more aware of their rights. 


Does where you live matter?

We created a series of visualization of state level variation in discrimination charge rates with the Center for Employment Equity’s Diversity Analytics web page. Figure 3 displays heat maps of discrimination charge rates for the fifty U.S. states, plus Puerto Rico and the District of Columbia. We report maps for the four most common status bases of discrimination charges:  female, Black, disability and age. There are two striking patterns. 

First, there is a great deal of variation in discrimination charge rates across political units. Women file discrimination charges at their highest rate, 85 per 100,000 workers, in New Mexico. Their lowest rate in neighboring Utah, is 84% smaller, at 14 charges per 100,000 women employees. For African Americans the highest charge rate is in Iowa, with 719 discrimination charges filed for every 100,000 workers. The lowest rate is in New York at 65 per 100,000. Disability discrimination charges show similar variation. Disability charge rates max out in Delaware with 851 per 100,000, and are lowest in South Dakota (112 per 100,000). Age discrimination charges by workers over age 40 are at their highest in Nevada (70 per 100,000) and lowest in Vermont (9 per 100,000). This high level of variation is a surprising and previously unknown pattern. It implies either that the rate of employment discrimination varies tremendously between states, or that rights consciousness and access to legal remedies are associated with where you live, or most likely some combination of the two.

The second striking pattern in these maps is that, while there are no obvious regional patterns, certain states stand out in Figure 3 for either having high or low levels of discrimination charge rates across multiple protected bases. On these four maps New Mexico and Delaware are particularly prominent among the high discrimination charge states. Among low charge rate states, New Jersey stands out on multiple maps.


Figure 3: Heat Maps of Female, Black, Disability, and Age Discrimination Charge Rates, 2011-2016. Darker colors indicate higher discrimination charge rates, Greyed out states have fewer than 30 total charges.


Table 1: States that appear six or more times among the top and bottom discrimination charge rates for nine bases of discrimination.
Type of Discrimination Claim
  Disability Race: Black National Origin Sex: Women Age: Over 40 Race: Asian Sex: Men Race: Hispanic Race: White
Bottom 10   NJ NJ NJ NJ NJ NJ NJ  


We looked across all nine charge rates displayed in Figure 2, enumerating the top and bottom 10 states for each and we found a remarkable pattern -- a very few states consistently generate the highest discrimination charge rates across all nine bases for making a discrimination charge. Low discrimination charge rate states were much less concentrated. Table 1 displays the result.

New Jersey stands out as a low discrimination state. New Jersey is among the bottom 10 discrimination charge rate states for seven of nine bases of discrimination claims. No other state makes the top 10 for six of more discrimination bases. High discrimination charge rates are much more concentrated in just a few states. New Mexico (9), Delaware (7), Nevada (7), Alabama (6), Indiana (6), Missouri (6), and the District of Columbia (6) all generate high discrimination claims on six or more claim bases. 

All of the highest discrimination states have high levels of disability charges. This patterns is nearly as complete for discrimination claims by men, women, over 40 and immigrants. Top discrimination states associated with African American claims show less overlap with the other high discrimination claims places. Only New Mexico and Indiana make the top 10 list for African American discrimination charges.

The most obvious explanation for the high rates of discrimination charges in a handful of states is that they have worse employers, and those employers run equal opportunity discrimination workplaces. It could, however, be that rights awareness and access to legal redress is higher in those states. It might even be that the EEOC and the courts are more sympathetic and so people who have grievances are more likely to file formal charges because they are more likely to receive some benefit. 

Figure 4 compares these high rate states and New Jersey, the single low rate state, to the national average for multiple indicators associated with discrimination charges. To create Figure 4 we first calculated all of these statistics for each of the nine charge bases, than took their average within each state, and then report these averages for the nation, the eight high discrimination states, and for New Jersey. We see this exercise as highly exploratory, and do not make any strong inferential claims, other than noting where zero order relationships are absent or exist.

We first examine the proportion of charges filed with legal representation to get at access to the legal system. In less than twenty percent of discrimination charges are lawyers present at filing. Lawyers accompany charges about 3% less often in high charge rate states than in the national average state, while low discrimination New Jersey is right at the national average. Thus, workers in high charge rate states, if anything, have lower than average access to legal representation. Legal representation does not standout for New Jersey. It does not look like legal representation is what generates discrimination charges. This is not surprising as past research suggests that lawyers only take discrimination cases they are confident they can win (Berrey et al. 2017). It may signal that in high discrimination charge rate states, lawyers are more reluctant to pursue such claims. This would be consistent with a legal environment that was more favorable to employers.

Two motives Berrey et al. (2017) found for filing discrimination complaints were to receive monetary benefits and to fix discriminatory workplaces for themselves and others. Consistently, past research suggests that higher rates of racial discrimination filing helps to reduce within state racial income inequality (Maume, Mathews, and Wilson 2018). We examine here the frequency of monetary benefits and changes in workplace practices associated with filing a discrimination charge as indicators of the likelihood that charging parties will find legal redress. Both monetary and workplace change agreements are relatively infrequent. The rates at which charges lead to monetary benefits in high charge rate states are pretty equivalent to the national average. In New Jersey there is a slightly higher probability of receiving monetary benefits, but a lower probability of negotiated settlements in which the employer must change workplace practices to combat future discrimination. Benefits to the charging party do not seem to distinguish between high and low discrimination charge states.


Figure 4: Averages for the eight high discrimination (NM, DE, NV, AL, IN, MO, DC,AZ) and the one low discrimiantion (NJ) charge rate states compared to national averages

Consistent with past research (e.g. Roscigno 2007), both job loss and employer retaliation are much more common than receiving some benefit for filing a discrimination charge. The high charge rate states are at just about the national average for charges being associated with both job loss and employer retaliation. If these states have more discriminatory employers than other states it is not evident in these two statistics. In contrast, New Jersey’s employers are less likely to both fire and retaliate against employees who file formal discrimination charges. This may be evidence that New Jersey employers are better at handling discrimination charges internally, and so employees are less likely to take their grievances to the EEOC or state Fair Employment Practice Agencies.



In this article we have discovered two new social patterns about contemporary employment discrimination in the United States. 

The first is that different groups file employment discrimination charges at wildly different rates. Persons with disabilities file employment discrimination charges under the American with Disabilities Act at more than twice the rate of the next most frequent filers -- African Americans. Other groups file employment discrimination charges at much lower rates. 

Filing a discrimination charge requires both experiencing discrimination, labeling it as such, and access to the legal system (Miller and Sarat 1980; Festiner 1980). Legal representation for disability charges is at about the same level as for other bases, but less common for African Americans. Thus we can rule out access to the legal system as an explanation for the higher rates of discrimination filings by the disabled and African Americans. This leaves two plausible explanations, these two groups experience more discrimination and/or have higher rights consciousness. Berrey et al. (2017) document that people who file discrimination lawsuits are initially motived by goals of fixing their workplaces for themselves and those who come after them. Assuming this is the case here, it suggests that the disabled and African Americans experience both more discrimination and have stronger identity based goals of fixing broken workplaces for other employees.

Our second discovery is that states vary widely in their rates of discrimination charges for all bases of claims. For every bases of the nine we examined the highest charge rate state received charges at ten times the rate of the lowest. There were seven states that stood out as having high discrimination charge rates for at least six of the nine bases we examined. These are New Mexico, Delaware, Nevada, Alabama, Indiana, Missouri, and the District of Columbia. These high discrimination charge rate states did not stand out for having higher access to legal representation or higher benefits to charging parties. It is difficult to imagine that for some reason they also share a higher rights consciousness than other states. It seems possible, that they have worse, and perhaps more equal opportunity discriminatory employers. 

Only New Jersey stood out as producing low levels of discrimination charges across multiple claims bases. Here the evidence points toward employers being on average better at resolving discrimination complaints internally, and less likely to react with the firing and retaliation so common in discrimination cases (Roscigno 2007). Why and if, New Jersey might have a better employer culture are, of course, important questions in their own right.

Of course, discovering these two social facts is not the same as providing a strong explanation for their variation. We think that the rights consciousness framework (Miller and Sarat 1980; Festiner 1980) is a good place to start further exploring this variation. Other promising approaches might be to examine variation in the role of group threat in producing discrimination (Maume et al. 2018), local political context (Wakefield and Uggen 2004), cultural variation in tolerance for discrimination (Skaggs 2009), and the legal system’s receptiveness to discrimination claims (Skaggs 2008).



Our core data are workplace discrimination charges from the Equal Employment Opportunity Commission (EEOC). These charges were filed directly with the EEOC or with one of the state or local Fair Employment Practices Agencies (FEPAs) that have agreements with the EEOC to share the processing of charges. The data include all workplace charges filed between fiscal years 2012 and 2016.2

Yearly charge population rates are calculated using the American Community Survey (ACS) employment estimates for states and industries. The ACS is the core survey used by the U.S. Bureau of the Census to describe local area populations. It is collected yearly from random samples of people living in the United States. Five years of ACS date are required to produce statistically accurate estimates of population size for sub-national units such as states or industries. We use the 2011-2015 five year ACS to estimate local employed workers potentially at risk for employer discrimination. We defined our population baselines as the civilian population currently employed at work. In addition, we restricted our population to individuals aged 16 and above. From these data, we estimate the civilian employed at work population at risk for each type of discrimination at the state and national level. These populations are described in more detail in the definitions section below.


Types of Discrimination

basis is the legally protected category that the plaintiff claims was the basis of discrimination. Under Title VII of the 1964 Civil Rights Act and its extensions these protected categories include race, sex, color, religion, disability, age, and national origin. We focus on five discrimination charge bases, and for sex and race the status characteristic of the charging party. 

Age Discrimination: includes all Age Discrimination in Employment Act (ADEA) charges filed by a charging parties 40 years of age or older that contain age as a basis of the charge.

National Origin Discrimination: includes all Title VII charges that contain national origin as a basis of the charge.

Race Discrimination: includes all Title VII charges that contain race as a basis of the charge.

Sex Discrimination: includes all Title VII and Equal Pay Act charges that contain sex as a basis of the charge. This includes charges made on the basis of sexual orientation or gender identity.

Disability Discrimination: includes all Americans with Disabilities Act (ADA) charges that contain disability as a basis of the charge.



Charge Rates: We follow Wakefield and Uggen (2004) in using Census Bureau employment estimates to calculate the discrimination charge rates: the yearly average charge rate per 100,000 employed persons in the state (or nationally). That is,


The at risk employed population is calculated using the American Community Survey (ACS) for the civilian employed at work population. Because our charges include charges filed in a five-year time frame, we multiply the employed population estimates by 5 in order to obtain a yearly charge rate. The populations are defined as:

  • For sex based discrimination, this includes the civilian employed at work population. Rates are calculated separately for each sex to produce sex specific charge rates and then combined to produce total charge rate. For example, the female charge rate is calculated as the number of charges filed by women over the number of employed women at work in the state or nationally.
  • For race based discrimination, this includes the civilian employed at work population in the state and nationally. Rates are calculated separately for each race and combined as above. For example, the black charge rate is calculated as the number of race based discrimination charges filed by a black charging party over the number of employed black at work employees.
  • For age discrimination, this includes the civilian employed at work population over the age of 40
  • For national origin discrimination, this includes the civilian employed at work population born outside of the United States and Puerto Rico.
  • For disability discrimination, this includes the civilian employed at work population identified as having a disability. The ACS asks respondents about six different disabilities: hearing difficulty, vision difficulty, cognitive difficulty, ambulatory difficulty, self-care difficulty, and independent living difficulty. Respondents who report any one of these difficulties is considered to have a disability. For more information see: https://www.census.gov/topics/health/disability/guidance/data-collection-acs.html

Lost Job: Percent of all charges filed which alleged job loss. Job loss includes constructive discharge, layoffs, and suspension, in addition to firing. Constructive discharge refers to an employer created hostile workplace that forces the employee to quit.

Faced Employer Retaliation: Percent of all charges filed that alleged retaliation. Almost all retaliation happens after an employee raises a complaint in the workplace, but prior to filing a formal charge with the EEOC or state FEPA.

Received Monetary Benefit: Percent of all charges closed for non-administrative reasons that received a monetary benefit. Administrative closures include charges closed for reasons such as failure to locate charging party, lack of jurisdiction, or charging party withdraws the charge. For more information, see:  https://www.eeoc.gov/eeoc/statistics/enforcement/definitions.cfm

Change in Employer Practice: Percent of all charges closed for non-administrative reasons that received a nonmonetary benefit. For more information, see:  https://www.eeoc.gov/eeoc/statistics/enforcement/definitions.cfm

Represented by Legal Counsel: Percent of all charges in which the charging party was represented by legal counsel. 



1 Berrey, Ellen, Robert J. Nelson, and Laura Beth Neilsen, 2017, Rights on Trial: How Workplace Discrimination Law Perpetuates Inequality.  University of Chicago Press.

2 Burstein, Paul, and Kathleen Monaghan. 1986. “Equal Employment Opportunity and the Mobilization of Law. Law and Society Review 20:356–88.

3 Felstiner, William LF, Richard L. Abel, and Austin Sarat. 1980. "The emergence and transformation of disputes: Naming, blaming, claiming..." Law and Society Review 15: 631-654.

4 Hirsh, C. Elizabeth, and Sabino Kornrich. 2008. “The Context of Discrimination: Workplace Conditions, Institutional Environments, and Sex and Race Discrimination Charges. American Journal of Sociology 113:1394–492.

5 Maume, David, Ervin Matthew, and George Wilson. 2018. "Minority Threat, Worker Power, and Discriminatory Complaints: State-level Effects on Racial Income Inequality among Men." Social Currents 5: 215-227.

6 McCann, Carly, Donald Tomaskovic-Devey, and Lee Badgett. 2018. "Employers' Responses to Sexual Harassment." Center for Employment Equity. University of Massachusetts. https://www.umass.edu/employmentequity/employers-responses-sexual-harass...

7 Miller Richard E. and Austin Sarat. 1980. “Grievances, Claims, and Disputes: Assessing the Adversary Culture” Law and Society Review. 15:525-566.

8 Roscigno, Vincent J. 2007. The face of discrimination: How race and gender impact work and home lives. Rowman & Littlefield Publishers.

9 Skaggs, Sheryl. 2008. "Producing change or bagging opportunity? The effects of discrimination litigation on women in supermarket management." American Journal of Sociology 113: 1148-1182.

10 Skaggs, Sheryl. 2009. "Legal-political pressures and African American access to managerial jobs." American Sociological Review 74: 225-244.

11 Wakefield, Sara and Christopher Uggen. 1994. “The Declining Significance of Race in Federal Civil Rights Law: The Social Structure of Employment Discrimination Claims.” Sociological Inquiry. 74: 128–157.