AEJ99JQFinal

Differential Employment Rates in the Journalism and Mass Communication Labor Force Based on Gender, Race and Ethnicity:
Exploring the Impact of Affirmative Action

James M. Cox Jr. Center for International Mass Communication Training and Research
College of Journalism and Mass Communication
University of Georgia
Athens, GA 30602
Tel. 706 542-5023
Fax 706 542-5036
E-mail: lbbecker@arches.uga.edu

Lee B. Becker is professor and director of the James M. Cox Jr. Center for International Mass Communication Training and Research in the Henry W. Grady College of Journalism and Mass Communication at the University of Georgia. Edmund Lauf, a visiting scholar in the Cox Center in the autumn of 1998, is a research associate in the Amsterdam School of Communication Research at the University of Amsterdam. Wilson Lowrey is a research assistant in the Cox Center and a doctoral student in the Grady College at the University of Georgia.

* This article is copyrighted by the Association of Education in Journalism and Mass Communication
and is presented on this web site with the permission of AEJMC.

This paper examines whether gender, race and ethnicity are associated with employment in the journalism and mass communication labor market and--if discrepancies in employment exist--what explanations might be offered for them. The data show strong evidence that race and ethnicity are associated with lower levels of employment among journalism and mass communication graduates. These discrepancies in success in the job market are explainable in highly specified situations by factors normally associated with hiring, such as type of training, type of institution offering the training or qualifications such as internship experience and level of performance in the classroom.

Anti-discrimination laws and affirmative action procedures theoretically have leveled the playing field in employment. Race, gender and ethnicity are supposed to no longer have a role in hiring, and, as a consequence, employment rates for the powerful and less powerful groups in society should be indistinguishable.

Because of the supposition that race, gender and ethnicity no longer play a significant role in hiring, affirmative action--though not anti-discrimination legislation--has come under severe attack in recent years. Affirmative action is often termed a form of reverse discrimination that ought to be outlawed by the anti-discrimination legislation and policy that preceded--and in many ways--spawned it.

Debates within the communication industries have mirrored those in society at large. Media industries are considering the need for continued affirmative action in a period of high employment overall and supposed equality of access to jobs in the economy.

Does discrimination continue to exist in the journalism and mass communication labor force? Is there evidence that anti-discrimination legislation and affirmative action policies on the part of media and related communication employment sectors have eliminated race, ethnicity and gender as criteria in hiring? This paper examines that question in detail by documenting, first, the rate of employment of those seeking entry-level jobs in the field of journalism and mass communication and then attempting to explain why discovered gaps in employment rates exist.

Until the U.S. Congress passed the Civil Rights Act in 1964, it was legal to make hiring decisions based on race, ethnicity and sex. At that time, it was common to use these characteristics in selecting employees and in assigning them to work tasks once they were hired. Such decisions might have made short-term economic sense, given past practices and existing prejudices in the society. They certainly were accepted practice in U.S. industries, including those in the communication sector.

Title VII of the 1964 Civil Rights Act made race and sex based segregation in the workplace illegal. Affirmative action resulted from the recognition that outlawing discrimination in hiring would not, of itself, result in equal access for women and members of racial or ethnic minorities to all segments of the labor force. Affirmative action requires employers to do more than refrain from making decisions based on race and ethnicity. It requires pro-active activities to promote equal employment opportunities to groups traditionally discriminated against in employment. Hiring is one of the important pro-active activities. Reskin, in her review of the effects of affirmative action on employment, says that the 1964 Federal law banning discrimination in employment curtailed the most blatant forms of discrimination but had little effect on the discrimination that stemmed from the ways that organizations went about recruiting, screening and evaluating workers.⁽¹⁾ She continues:

Reskin said that eliminating forms of habitual discrimination requires employers to actively modify their personnel practices. These modifications, designed to ensure race- and gender-neutral employment practices, are a fundamental part of affirmative action.

According to Reskin, four different types of affirmative action exist. The first type results from presidential and gubernatorial executive orders requiring action by government contractors and subcontractors and actually affects only a very small number of employers and employees. The second type is the consequence of regulations on government itself in its capacity as employer. The third type is based on court decisions in anti-discrimination cases. The final type stems from voluntary human resource policies of employers.

There is no doubt a link between the third and fourth types of affirmative action. The 1972 Equal Employment Opportunity Act amended Title VII of the 1964 Civil Rights Act to allow federal courts to include affirmative action among the remedies they can require of firms found guilty of discrimination and use affirmative action plans as part of a consent decree in settlement of such cases. Employers might well have initiated affirmation action programs voluntarily with the knowledge of the power of the courts to mandate such programs should a legal complaint be adjudicated.

Voluntary affirmative action programs have been found to have common characteristics.⁽²⁾ They promote the integration of jobs in organizations and racial and gender neutrality in decision-making. Advertisements that say the organization is an equal-opportunity employer are a common component. Some organizations train and reward managers for their affirmative action performance.⁽³⁾

Estimates of the percentage of firms implementing voluntary affirmative action policies vary. Edelman found that 71 percent of the organizations surveyed nationally had affirmative action plans, almost one in five had an Equal Employment Opportunity or affirmative action office, and most had created structures and rules to foster affirmative action.⁽⁴⁾ Holzer and Neumark, however, found that less than half of employers in Los Angeles, Atlanta, Boston, and Detroit said they took into account either equal employment opportunities or affirmative action in hiring workers.⁽⁵⁾ Miller found that only four in 10 of the large firms in the New York-New Jersey-Connecticut metropolitan area surveyed reported some form of affirmative action to recruit minorities, less than three in 10 did so to recruit women, about three in 10 employed affirmative action in promoting minorities, and one in four did so in promoting women.⁽⁶⁾ Fewer than half of workers have been found to believe their employer practices affirmative action.⁽⁷⁾

The American Society of Newspaper Editors, the National Association of Broadcasters, the Newspaper Association of America, the Radio-Television News Directors Association and other professional mass communication groups have been outspoken in promoting affirmative action among their members.⁽⁸⁾ The Associated Press in the early 1980s settled a complaint by the Federal Equal Employment Opportunity and seven female, former AP employees requiring the news service to, among other things, develop an affirmative action policy for the organization.⁽⁹⁾ The Federal Communication Commission historically has required stations applying for license renewal to present an affirmative action plan for hiring and promotion of women and minorities.⁽¹⁰⁾

Despite these efforts at affirmative action and the existence of anti-discrimination laws, evidence of discrimination continues to exist in the U.S. This conclusion is based on evidence of differential employment outcomes, on reports of employers to researchers describing how race and sex affect their hiring decisions, and on the prevalence of organizational practices that have been shown to have discriminatory consequences.⁽¹¹⁾ All show evidence that discrimination continues to be a prevalent part of American employment. This is true despite mounting evidence of potential effectiveness of affirmative action procedures. Reskin, in her review of a research on this topic, found that affirmative action programs required by presidential executive orders have reduced discrimination against minority and female workers, as have court-ordered affirmative action and even some voluntary affirmative action plans.⁽¹²⁾ Bowen and Bok's study of the performance of minority students admitted to selective colleges through race-sensitive admission programs similarly documents the positive consequences of that decison.⁽¹³⁾

Weaver and Wilhoit have found little evidence that American newsrooms have become much more diverse across time.⁽¹⁴⁾ Women made up 20% of the journalistic workforce in 1971, 34% in 1982-83, and 34% in 1992. Minorities increased from about 5% of the journalistic workforce in 1971 to 8.2% in 1992. The American Society of Newspaper Editors has found that minorities make up less than 12% of the newsrooms of daily newspapers in the U.S.--roughly the same percentage as a year earlier and up from 8% in 1990 and 4% in 1978.⁽¹⁵⁾ Women made up 37% of the daily newsrooms in 1998the first year that figure was tracked by the organization. In 1997, 27% of the U.S. population was minority; the figure is projected to increase a percentage point by 2000.⁽¹⁶⁾

The recent literature on employment helps to explain why discrimination continues, despite legal prohibition of it. Much of this literature is a departure from the classic economic perspective on hiring, in which it is assumed that individuals looking for jobs, or managers looking for employees, are rational actors and follow the dictates of labor market theory. Employers are the buyers of labor, and employees are the sellers. Supply and demand operate to regulate wages and establish equilibrium, and within a strict interpretation of this view, unemployment should not exist.⁽¹⁷⁾ It is also assumed that individuals can possess complete information about the "market".⁽¹⁸⁾

Granovetter, however, has resituated the concept of job search from the rational actor approach of economics to the realm of sociology.⁽¹⁹⁾ Job-finding behavior, he says, "is heavily embedded in other social processes that closely constrain and determine its course and results." Granovetter begins with the idea that information is imperfectly distributed in the job-search process. Many top-level jobs are filled by those privy to special information, and he found that 57 percent of jobs are filled through informal contacts and non-rational avenues. In similar studies other researchers have found percentages ranging from the 30s to the 50s.⁽²⁰⁾

According to Granovetter, individuals are most likely to receive new job information from those with whom they have "weak ties."⁽²¹⁾ They are less likely to receive new information from those with whom they have strong ties (family and close friends) because they are likely to already have this information. Common examples of weak ties include friends and family of former co-workers and schoolmates, and contacts made from networking at professional conferences. Weak ties are occupational rather than social, and those who have fewer ties tend to have more trouble in the job market. Historically, blacks have been at a disadvantage because they have been underrepresented in the occupational structure itself and therefore have been less privy to informal channels of opportunity.

Formality of search is also a major dimension in the literature on hiring theory. According to Marsden and Campbell, informal recruitment (e.g., employees recruiting acquaintances) facilitates the hiring of persons who are already well informed about a workplace.⁽²²⁾ For this reason, informal recruitment can lower turnover, but it may also lead to a lack of diversity in the applicant pool, because those associated with the firm are likely to convey information to socially similar persons. Coverdill and Finlay, in a study of headhunters and white-collar recruitment, found that the assessment of a job candidate's likely social fit is a highly important factor in the hiring of white-collar employees. In short, "like hires like."⁽²³⁾

According to Marsden, formal channels of recruitment--including advertising in the mass media, visiting campuses and using placement agencies--are more suitable for reaching geographically dispersed and more demographically diverse candidates.⁽²⁴⁾ Also, larger organizations are more likely to use formal channels of recruitment because they have more resources for formal searches and their needs may exceed the skill resources of the local community. Barron and Bishop found that economic conditions may also have an effect on the level of formality and on how intensive the search is. When faced with a smaller applicant pool, employers tend to screen potential hires less carefully.⁽²⁵⁾

Pfeffer and Cohen found the existence of internal labor markets in organizations to be an indicator of formalized recruitment methods.⁽²⁶⁾ At organizations with internal labor markets, i.e. formal promotion ladders with few ports of entry at the bottom, employees are expected to stay with the organization for a lifetime and management therefore has a higher stake in their quality. Similarly, Barron and Bishop found that organizations use more intensive formal methods of recruiting when it is necessary for the organization to invest more training in the position.⁽²⁷⁾

Little systematic work has been done focusing on hiring and its effects of the journalism and mass communication labor market. Becker, Fruit and Caudill did study how media organizations make hiring decisions, but they did not examine how those decisions were affected by gender and race or ethnicity.⁽²⁸⁾ Becker, Kosicki, Engleman and Viswanath examined the individual predictors of success in the journalism and mass communication labor market, though the analysis dealt only tangentially with gender and race.⁽²⁹⁾ Neither variable predicted success of the applicants in finding a job in the field. Claussen, Lowrey, Anderson and Becker found evidence that newspapers, at least, do not have elaborate internal labor markets--that is, policy that leads to clear paths of internal advancement--and rely heavily on external markets of the type most likely to be influenced by informal hiring methods.⁽³⁰⁾

Research in the sociology of organizations has found that a strong external or "craft" labor market is an indicator of more informal recruitment channels. In a craft market, skills are transferrable between organizations and job openings are often communicated informally between occupational members. News organizations have been viewed as existing in an external "craft" labor market.⁽³¹⁾

Two different types of literature have been reviewed to this point. The first is macro-level, looking at social policy and its outcome in producing a labor force that is not differentiated by race or sex. The second is organizational, looking at factors that come into place as employers make decisions on whom to hire and whom to promote. The social policy can dictate organizational behavior, and only if the organizational behavior is consistent with social policy will the outcome of employment undifferentiated by race and gender be possible. At least that is the assumption of the social policy.

Unknown in the field of journalism and mass communication is the extent to which social policy of equal opportunity and anti-discrimination has led to employment patterns that are not differentiated by race and gender. The evidence at hand has focused only on segments of the industry and only on the cumulative effects of hiring, not on the immediate hiring decisions themselves.

If the social policy of affirmative action and the laws of anti-discrimination have been effective, there should be no difference in employment rates for women and men, or for minorities and those not labeled as minorities. If the policy has failed in the field of journalism and mass communication--as seems likely given the evidence of its failure in the labor force generally--women and men should show differential employment rates, as should minorities and nonminorities. We hypothesize this second outcome.

Hypothesis 1: Across time, differential employment rates of women and men and of minorities and nonminorities in the field of journalism and mass communication will continue to exist.

We are not expecting blatant discrimination to continue to play a significant role. The law has made it difficult for employers to refuse to hire qualified applicants or to openly follow policies that make it impossible for women or minorities to gain jobs. Gender, race and ethnicity, however, are often linked to other characteristics that play a role in the hiring decision. Minorities are less likely to have had the same educational experiences, the same opportunities, and the same guidance as their colleagues. The same can be said for women in comparison with men. The result should be that much of the evidence of discrimination should be attributable to the legacy of inequality. What isn't attributable to these should result from what we have termed the informal hiring practices of media organizations--the practices that can be addressed by renewed efforts at affirmative action. As noted above, the informal hiring procedures often used make it likely that people will hire people like themselves--and those making hiring decisions are not often female or nonwhite. We hypothesize that not all of the variance in employment rates will be explainable by the legacy of inequality of experience. We speculate that such a finding will provide evidence of the effects of informal hiring practices.

Hypothesis 2: Even after controlling for the inequality of experiences of women and men and of minorities and nonminorities, women were be less likely to be hired than men, and minorities will be less likely to be hired than nonminorities.

Data gathered as part of the Annual Survey of Journalism & Mass Communication Graduates are appropriate for a test of these hypotheses. The Annual Survey of Journalism & Mass Communication Graduates is designed to monitor the employment rates of graduates of journalism and mass communication programs in the United States, including Puerto Rico. In addition, the survey tracks the curricular activities of those graduates while in college. The survey has employed a comparable methodology since 1987. During that time, more than 24,200 graduates of journalism and mass communication programs around the country participated in the survey.

At present, the Annual Survey of Journalism & Mass Communication Graduates is conducted at the Henry W. Grady College of Journalism and Mass Communication at the University of Georgia. From 1987-1996, the survey was conducted at The Ohio State University.

As a first step in the survey, a sample of schools is drawn from those listed in the Journalism and Mass Communication Directory, published annually by the Association for Education in Journalism and Mass Communication, and The Journalist's Road to Success: A Career and Scholarship Guide, published each year by the Dow Jones Newspaper Fund, Inc. The sample of schools is modified each year to reflect changes in these two directories. Sample selection is probabilistic, so that schools selected represent the population of schools listed in the two directories. In 1997, as an example, 92 schools were drawn from the 451 unique entries of four-year programs in the U.S. (including Puerto Rico) in the two directories.

Administrators at the selected schools are asked to provide the names and addresses of their spring bachelor's and master's degree recipients. As the second step in the survey, a questionnaire is mailed in November or December to all spring graduates receiving either a bachelor's or a master's degree from the selected programs. A second questionnaire is sent to nonrespondents in January or February.

The questionnaire asks about the respondent's experiences both while as a student and in the months since graduation. Included are questions about university experiences, job-seeking and employment, and salary and benefits.

Employment rates for each of the survey years are based on responses to a question on employment at the time of completion of the interview. Respondents could indicate if they were employed full-time, part-time, enrolled in school, or unemployed. Respondents enrolled in school were eliminated from computation of employment rates.

Respondents classified themselves according to sex and race/ethnicity, allowing for computation of employment rates by these social strata.

Additionally, respondents indicated if they had sought jobs in the field of journalism and mass communication and if they had received job offers in the field. For those who sought such a job, receiving an offer is an indicant of job-seeking success, whether or not the job was actually taken.

Respondents indicate the type of specialization within journalism and mass communication they completed, the number of internships they completed, in which campus media they participated, and their grade point average.

Respondents were classified as having attended an accredited journalism program or not and, based on the percentage of students applying to their college and university who were admitted, as having attended a selective university or not.

Return rate for the 1997 survey, computed as the number returned divided by the number mailed minus the bad addresses, was 54.5%. Of the usable questionnaires, 2,169 were from bachelor's degree recipients and 145 were from those who received a master's degree. Earlier surveys produced similar return rate statistics.

Most graduates of journalism and mass communication programs receive an undergraduate, rather than graduate degree, and the graduate survey reflects this fact. The design of the survey was changed in 1989 only to incorporate master's degree recipients. The analyses that follow are based only on the responses of bachelor's degree recipients from 1987-1997, two-thirds of whom were female and 15 percent of whom were classified as minorities--that is, African American, Hispanic, Native American, Asian American, Pacific Islander American, or Other.

Tables 1 and 2 show the employment rates for bachelor's degree recipients who did not return to school in the year following the completion of their undergraduate programs. Table 1 reports the rates separately for women and men. Table 2 reports the employment rates for Minorities and Nonminorities.

In general, there are only small differences between women and men in terms of level of full-time employment six to eight months after graduation from a journalism and mass communication program. The figures were nearly identical in 1987, and they vary only slightly in the years that follow. Across all years, the difference in full-time employment rates is 2.7%a difference that is statistically significant (Pearson Chi-Square=26.28, df=2, p<.001). The difference is in favor of women. Women have a slightly better chance of being employed full-time six to eight months after graduation than do men, who are slightly more likely to end up in part-time positions.

The discrepancies between minorities and nonminorities, however, is more striking. In every year except one since 1987, minorities report lower levels of full-time employment six to eight months after graduation than do nonminorities. The exception is 1988, when the number of minorities surveyed was quite small, and, as a consequence, the power of the estimation low. To be sure, in two years, 1991 and 1992, the differences in level of full-time employment was slight. Across all years, however, the gap in level of full-time employment six to eight months after graduation was 5.2 percentage points, with minorities significantly less likely to report full-time employment than were nonminorities (Pearson Chi-Square=41.19, df=2, p<.001).

Hypothesis 1 predicted discrepancies in the labor market--as reflected in employment rates--for women versus men and for minorities versus nonminorities, despite the existence of affirmative action initiatives. Evidence of discrepancies for women are slight and suggest that women have fared just slightly better than men. The discrepancies are more striking when minority graduates are compared with nonminority graduates. This discrepancy actually increased over time and supports the hypothesis of continued inequalities, based on race and ethnicity, in the labor market.

The employment rates are macro-level indicators of potential discrimination in the labor market. A better indicator of bias in hiring, however, comes from an analysis of any discrepancy between job seeking and the receipt of a job offer. The graduate survey asked respondents to indicate from which types of mass communication employers they sought jobs and from which ones they received offers. These survey questions test hiring discrimination at the individual level and are therefore appropriate for use in a test of Hypothesis 2.

Table 3 reports the results of a logistic regression analysis in which the receipt of a job offer is the dependent variable. Only those graduates who actually sought work with a communication employer are included.⁽³²⁾ The dependent variable is scored as dichotomous, with receipt of an offer coded as 1 and no receipt coded as 0.

Independent variables used in the analysis were minority status (minority scored high), reported college grade point average, accreditation status of the program at which the graduate studied, selectivity of that program, sequence specialization (with dummy variables for journalism, RTV and advertising), number of internships worked by student applicants, number of college media for which the student worked, and gender (female scored high).

Table 3 shows the results of these regression analyses only back through 1990, because similar measures of all of the key variables were not available in earlier years. In each year, the variables entered into the equation explained a small but significant amount of variance in the receipt of at least one job offer. The R square ranges from a low of .031 to a high of .080.

Consistent with the findings from the analysis of employment rates, gender contributes little to an understanding of the dependent variable, receiving a job offer. Only in 1996 does gender explain a significant amount of variance. In each year, the standardized logistic regression coefficient is positive.⁽³³⁾

Race/ethnicity, however, is a different matter. Being a member of a minority is negatively associated with receipt of a job offer in four of the eight years analyzed in Table 3, and in each of these four years the standardized regression coefficient is statistically significant. What is striking is that these are the four most recent years. Minority status appears to be becoming increasingly important in understanding hiring and increasingly negative in its contribution to the hiring outcome.

Other variables that explain a significant amount of variance are grade point average, accreditation status, number of internships received and number of campus media in which the student participated. Having a newspaper major and having an advertising major are also significant predictors. None of these are consistent across all years. What is consistent is that gender matters little. Selectivity of the university also matters little, controlling for other factors.

In a separate analysis (not shown), race/ethnicity was scored as a dummy variable, with African-American, Hispanic and Asian-American status entered into the equation. The analysis showed that being Black in every year was negatively associated with getting a job offer in the field of journalism and mass communication--among those who sought such a job. In recent years, being Hispanic also produced negative outcomes. Being an Asian-American is not associated with receipt of a job offer.

To further examine the effects of race/ethnicity on hiring, a more refined model was used in which, across each year from 1990 to 1997, the dependent variable was receipt of a job in a specific communication employment sector for those who sought work in that sector. The predictors, in addition to race/ethnicity and gender were grade point average, accreditation status of the graduate's journalism program, and the selective of the graduate's university. In addition, the graduate was scored for having specialized in the area of the job sought, having completed an internship in that employment sector, and having worked for a campus medium linked to that employment sector. The results of this analysis are summarized in Table 4. In the equation shown for each sector, year is entered as one of the predictor variables, although this final equation was computed only after the separate year-by-year analysis were examined.

Notable first are the increases in predictive power of the models. Only for weekly newspapers and cable television do the R square figures remain particularly small. Also notable is the strong and consistent predictive power of an internship. Clearly completing an internship in the employment sector of the job sought is a very important predictor of success in actually finding work in that sector. For daily newspaper jobs, for television jobs, and for jobs in advertising, having specialized in college in those areas of study also is extremely important, controlling for other influences.

As in Table 3, gender is a relatively unimportant predictor in Table 4. In the cases of public relations and advertising, women are more likely to get a job offer than are men, other things being equal, but otherwise there is no effect of gender. Unlike in Table 3, however, race/ethnicity also is not a particularly important predictor of getting a job offer in the analysis shown in Table 4. Only where weekly newspaper and adverting jobs are concerned is race/ethnicity a significant--and negative--predictor of job market success. Even here the effects are small.

In general, there is considerable consistency for the equations across the 1990 to 1997 period. For race/ethnicity, for example, the standardized regression coefficient with receiving a job offer is .00 each year. Because the job market improved year-to-year after 1990, as Table 3 shows (% offered job), year is generally a significant predictor of getting a job in the summary equations shown in Table 4.

In sum, while Table 3 offers support for the hypothesis that race/ethnicity remains a predictor of job market success even after controlling for formal qualifications of the applicants, Table 4 generally does not. Neither table offers support for the hypothesis that gender remains a predictor of job market success after controlling for formal qualifications of the applicants.

The analyses reported here were intended to serve two purposes. The first is an assessment of social policy--or of the effects of social policy. The second was the development of social and organizational theory. This article's origins lie in the recent debate about the need for the continuation of affirmative action policy in the United States as a means of addressing social inequalities in the labor market. It has attempted to inform that debate through an understanding of the labor market and of how organizations actually make decisions that have impact on that labor market.

A case might be made for discontinuing affirmative action policies in hiring in the field of journalism and mass communication if there were no evidence of continued inequalities in the labor market attributable to the social classifications of gender and race/ethnicity. The data presented here suggest the situation is complex, and such action might be premature. While women have fared well in recent years, allowing them to make at least slight gains that will help offset--but clearly not overcome--biases from the past, minorities have not done well. Being a minority is associated with lower levels of employment, and the situation became worse in the most recent years studied. This evidence argues that policies that are designed to offset biases in the labor market have not yet been completely effective.

Yet a closer examination of the situation shows that--at least in the specific market sectors analyzed--race is not a predictor of job searching success if the applicants have prepared themselves with employer specific internship training, campus media experience and curricular specialization. The conclusion is limited to the specific labor market segments for which adequate data for analysis exist (as shown in Table 4). And it does not negate the general findings (in Table 3) that summing across labor market segments and using non-market specific measures of internships and campus media experience, race/ethnicity is a slight, negative predictor of job market success.

We speculate that these remaining biases in the labor market are attributable to what we have termed informal hiring methods, that is, methods not based on strict rules but rather stemming from social contacts and interaction. The contrasting predictive power of the equations where general measures of formal characteristics were used (Table 3) and where specific measures of formal characteristics were employed (Table 4) argue that formal characteristics are important. To the extent informal methods remain in play, affirmative action seems to be called for.

Employers in media organizations should be aware that, if informal hiring procedures lead to a less diverse staff, this may in turn lead to less diversity in the content generated and processed by this staff. While news content may be constrained by organizational routines, the organization's culture and the individuals in the organizations play a role as well.

The policy implications of the paper seem clear enough. Increased attention to the development of hiring procedures that are designed to provide fair access to the labor market is needed. In addition, minorities need to be adequately advised of the importance of specializing in their education so that they posses the formal qualifications that are so essential for success in the job market. The theoretical implications also seem clear. This study suggests that informal hiring procedures may play an important role in the journalism and mass communication labor market. A more formal test of this relationship in which direct measures of informal hiring procedures are employed is now needed.

1. Barbara F. Reskin, The Realities of Affirmative Action in Employment (Washington D.C.: The American Sociological Association, 1998), 6-7.

2. Susan Vernon-Gerstenfeld and Edmund Burke, "Affirmative Action in Nine Large Companies: A Field Study," Personnel 62 (April, 1985): 54-60.

3. M.V. Lee Badgett, and Heidi Hartmann, "The Effectiveness of Equal Employment Opportunity Policies," in Economic Perspectives on Affirmative Action, ed. M.C. Simms (Washington, D.C.: Joint Center for Political and Economic Studies, 1995), 55-97; Lauren B. Edelman, "Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law," American Journal of Sociology 97 (May, 1992): 1531-76.

4. Edelman, "Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law."

5. Harry J. Holtzer and David Neumark, "Are Affirmative Action Hires Less Qualified? Evidence from Employer-Employee Data on New Hires." Journal of Labor Economics 17 (July, 1999): 534-69.

6. Joanne Miller, Corporate Responses to Diversity: A Benchmark Study (New York: Center for the New American Workforce, Queens College, City University of New York, 1994).

7. Davis, James A. and Tom W. Smith, General Social Survey [MRDF] (Chicago, Ill.: National Opinion Research Center. 1994).

8. The American Society of Newspaper Editors in 1998, acknowledging that newspapes would fall short of achieving racial parity in newsrooms with the population by 2000, modified its mission statement on newsroom diversity to say newsrooms must reflect racial diversity of American society by 2025 or sooner. See "ASNE Sets New Vision for Newsroom Diversity Beyond 2000," News Release Oct. 20, 1998 (Reston, VA: American Society of Newspaper Editors).

9. Fink, Conrad C. Fink, Strategic Newspaper Management (New York: Random House, Inc., 1988).

10. Sherman, Barry L. Sherman, Telecommunications Management: The Broadcast and Cable Industries (New York: McGraw-Hill Book Company, 1987).

13. William G. Bowen and Derek Bok, The Shape of the River: Long-term Consequences of Considering Race in College and University Admissions (Princeton, N.J.: Princeton University Press, 1998).

14. David H. Weaver and G. Cleveland Wilhoit, The American Journalist in the 1990s: U.S. News People at the End of an Era (Mahwah, N.J.: Lawrence Erlbaum Associates, 1996).

15. American Society of Newspaper Editors, ASNE's 1999 Newsroom Census: Minority Employment Inches Up in Daily Newspapers (Reston, VA: American Society of Newspaper Editors, 1999).

16. U.S. Bureau of the Census, Statistical Abstract of the United States: 1998, 118^th edition (Washington: U.S. Government Printing Office, 1998).

17. Mark Granovetter, Getting a Job: A Study of Contacts and Careers (Chicago: University of Chicago Press 1995).

18. George Stigler, "The Economics of Information," Journal of Political Economy, 69 (June, 1961): 213-225.

20. Karen Campbell and Rachel Rosenfeld, "Job Search and Job Mobility: Sex and Race Differences," Research in the Sociology of Work, 3 (1985): 147-74; Susan Hanson and Geraldine Pratt, "Job Search and the Occupational Segregation of Women," Annals of the Association of American Geographers, 81 (June 1991): 229-53.

22. Peter V. Marsden and Karen E. Campbell, "Recruitment and selection processes: the organizational side of job searches," in Social Mobility and Social Structure, ed. Ronald L. Breiger (New York: Cambridge University Press, 1990), 59-79.

23. James E. Coverdill and William Finlay, "Fit and Skill in Employee selection: Insights from a Study of Headhunters," Qualitative Sociology, 21 (summer 1998): 105-27.

24. Peter V. Marsden, "The Staffing Process," in Organizations in America , ed. Arne L. Kalleberg, David Knoke, Peter V. Marsden and Joe L. Spaeth (Thousand Oaks, Calif.: Sage Publications, Inc., 1996), 133-56.

25. John M. Barron and John Bishop, "Extensive Search, Intensive Search and Hiring Costs: New Evidence on Employer Hiring Activity," Economic Inquiry 23 (July,1985): 363-82.

26. Jeffrey Pfeffer and Yinon Cohen, "Determinants of Internal Labor Markets in Organizations," Administrative Science Quarterly, 29 (December, 1984): 550-572.

27. John M. Barron and John Bishop, "Extensive Search, Intensive Search and Hiring Costs: New Evidence on Employer Hiring Activity."

28. Lee B. Becker, Jeffrey W. Fruit and Susan L. Caudill, The Training and Hiring of Journalists (Norwood, N.J.: Ablex Publishing Corp., 1987).

29. Lee B. Becker, Gerald M. Kosicki, Thomas Engleman and K. Viswanath, "Finding Work and Getting Paid: Predictors of Success in the Mass communications Job Market," Journalism Quarterly 70(winter 1993): 919-33.

30. Dane Claussen, Wilson Lowrey, William B. Anderson and Lee B. Becker, "Newsroom Structure: Why Does the Beat Go On?" (Paper presented at the annual meeting of the Southeast Colloquium of the Association for Education in Journalism and Mass Communication, Lexington, KY (1999).

31. James N. Baron, Alison Davis-Blake and William T. Bielby, "The Structure of Opportunity: How Promotion ladders vary within and among organizations," Administrative Science Quarterly 31 (June, 1986): 248-73.

32. We found no systematic differences in the characteristics of minorities and nonminorities who do not seek communication employment.

33. The standardized logistic regression coefficient can be interpreted as roughly equivalent to a partial correlation coefficient.

* This article is copyrighted by the Association of Education in Journalism and Mass Communication and is presented on this web site with the permission of AEJMC.