Introduction

Abstract

This chapter provides an introduction to the two main sections of this book. The first section of the book provides a summary of the historical and social science literature about the history of higher education – and science, technology, and computing education in particular – for women, African Americans, Hispanics, and American Indians in the United States. The second section provides case studies of organizations interested in broadening participation in the science and technology disciplines in general and in computing in particular; as well as case studies about college and university departments of computer science and engineering that have had success in attracting, retaining, and advancing women in engineering and computing careers. The chapter discusses overarching themes that run through the book: exogenous forces (war, civil rights, reverse discrimination, and IT workforce needs); the conceptualization of the underrepresentation problem in terms of a pipeline instead of a pathway; solutions that involve fixing people contrasted with those that involve fixing the system; the role of nonprofit organizations and individual change agents in broadening participation in computing; and the issues surrounding intersectionality, i.e. cases in which someone belongs to two or more underrepresented groups such as being both female and African American.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

We felt like anomalies…. The women felt the difference most keenly during breaks, when they couldn’t join in the inside jokes and casual conversations into which their male colleagues seemed to fall so easily. In a profession so dependent on teamwork and learning new technology, being part of the community is not just a matter of feeling comfortable. It’s essential to being competitive. (Katherine Jarmul , co-founder of PyLadies , as quoted in Shah 2012)

This book addresses the history of underrepresentation of women and certain groups of minorities in the computing field in the United States.¹ The focus is primarily on the formal education of information workers rather than on the workforce itself. As Fig. 1.1 and Table 1.1 show, women, African Americans, Hispanics, and American Indians have been consistently underrepresented in the computing field throughout the entire era of modern computing, i.e. since 1945.² Figure 1.1 shows underrepresentation of women and minorities in receiving doctoral degrees in computing from 1977 to 2010. Table 1.1 shows similar underrepresentation at the bachelor’s level for 2014.

Fig. 1.1

Computer science doctoral degrees granted (Data Source: IPEDS Completion by Race, from https://webcaspar.nsf.gov). This figure is published as Figure 1 in http://cra.org/crn/2013/05/ expanding_the_pipeline_diversity_drives_innovation/]. (Source: McKinley and Camp 2013)

Table 1.1 Bachelor’s production in the computing fields (2014) – percentages by demographic group

There are various reasons why this underrepresentation is important. It a social equity issue that these high-paying, fulfilling, socially transformative jobs are less available to individuals from other demographic groups than they are to many White and Asian men. Shortages of skilled professionals occur every few years in the computing fields, and many scholars and policymakers believe both that these shortages have been harmful to American competitiveness and that larger participation of these underrepresented groups would go a long way towards meeting this skilled workforce need. Other scholars and policymakers (Page 2008; Barker et al. 2014) have pointed out that the work products of diverse work teams are more innovative and more likely to meet the needs of a wide range of customers than those created by a White male monoculture of technology developers.³

There is a small body of historical literature about women in computing in the United States. In addition to a number of articles and books focused on narrowly defined pieces of this history, there are three general books on this history: monographs by Janet Abbate (2012) and Nathan Ensmenger (2012), and an edited volume by Thomas Misa (2010). The literature on the history of underrepresented minorities and computing is much thinner; and in fact there are not yet any book-length general histories of this topic.

The social science literature on underrepresentation in computing is perhaps two orders of magnitude larger than the historical literature mentioned above. Like the historical literature, this social science literature is stronger and more numerous on women than on underrepresented minorities. Unfortunately, none of the major historical studies in this area have been informed in any substantial way by this social science literature, even though it has promise to offer insights into underlying causes.

This book is intended primarily as an historical study, even though it pays considerably more attention to the social science literature than do the other historical works on underrepresentation and computing mentioned above. This book is not a definitive history of underrepresentation and computing in the United States, but it provides useful background material drawn from the historical and social studies of computing, education (especially technical education), and race and gender that should help prepare some future scholar to write a more definitive history.

1.1 Overarching Themes

We will close this introduction with a brief discussion of some of the themes that run throughout this book and its companion volume – and, indeed, are likely to be present in any historical study of broadening participation in computing in the United States.

1.1.1 Four Exogenous Forces

In this section, we briefly identify four exogenous forces that have shaped efforts to broaden participation in computing in the United States since 1945. The first is the return after the Second World War of male veterans who displaced a number of women from science and engineering jobs. There were many fewer women working in jobs during the war as scientists or engineers than there were women working in manual manufacturing positions, but those who were displaced from scientific occupations were often unhappy about this change. One outcome of this dissatisfaction was the creation of the Society of Women Engineers in 1950. It was the first of a number of organizations formed from the 1950s to the 1970s with the purpose of helping to broaden participation in the STEM disciplines. This story is told in passing in Chaps. 2 and 6, and will not be discussed further here.

The second exogenous force shaping broadening participation in computing in the United States were the civil rights and women’s rights movement s. For example, the women’s rights movement led directly to the establishment in 1980 of the Committee on Equal Opportunities in Science and Technology with the purpose of advising the Director of the National Science Foundation on matters of broadening participation across the STEM disciplines.⁴ The role of the women’s rights movement in these broadening participation activities is told in passing in Chaps. 2 and 6 and in another publication by this author (Aspray 2016), and so will not be discussed further here.

The impact of the civil rights movement is less well known, so we will take some extra space to discuss one important episode related to this particular exogenous force. The second half of the 1960s was punctuated by race riots in Rochester, NY, Harlem, and Philadelphia in 1964; in the Watts section of Los Angeles in 1965; in Cleveland and Omaha in 1966; in Newark and Plainfield, New Jersey as well as in Detroit and Minneapolis in 1967; in Chicago, Washington DC, Baltimore, and again in Cleveland in 1968; and again in Omaha in 1969. In response to the wanton destruction and lack of opportunities for minorities, in 1973 the National Academy of Engineering , together with the Commission on Education , convened a Symposium on Increasing Minority Participation in Engineering . The more than 250 attendees included government officials, industrial representatives, minority leaders, and students. The major outcome of this meeting was a call for “equitable participation” of minorities within a decade.

To understand the dimensions of the problem and operationalize the call for action at the symposium, the Alfred P. Sloan Foundation sponsored a 7-month study carried out in late 1973 and early 1974 at Stanford University . The study provides today’s historian with a good snapshot of the situation in the early 1970s. For the contemporary reader of the 1970s, the study intended to identify numerical targets for minority representation in engineering; ascertain the feasibility, steps, and costs to achieve these goals; and identify organizations that could carry out various aspects of the work. The report appeared in 1974 under the title Minorities in Engineering: A Blueprint for Action (Sloan 1974).

The report focused on four underrepresented minority groups in engineering: Black, Chicano, Puerto Rican, and American Indian (using the terms that appeared in the report). These four minorities represented 14.4 % of the population in 1970 but only 2.8 % of the engineers. Of these four minorities, Blacks had the poorest representation, with 11.1 % of the U.S. population but only 1.2 % of the engineering profession. Table 1.2 lists the report’s 13 major recommendations.

Table 1.2 Major recommendations from the 1974 Sloan Report

The report found that Black students take longer to complete a high school degree and have a higher dropout rate than White students. While Black enrollment in college had at least doubled in every decade of the twentieth century, and while there had been a particularly significant growth in Black freshmen enrollment in college between 1970 and 1973, there was still a significant gap between college entry for Blacks and Whites. A greater percentage of Blacks dropped out of college than Whites. The community colleges were a particularly important educational feeder for Blacks, with more than 40 % of full-time Black students enrolled in community colleges; but there were concerns that a two-tiered higher educational system would emerge with the community colleges becoming “dumping grounds” for minorities, where career opportunities would be limited. Until 1960, approximately two-thirds of Blacks enrolled in college were enrolled at historically Black colleges and universities, but there was concern about the financial soundness of these institutions.

The report found that Chicanos were less likely than Blacks to be successful at the elementary school, high school, and college levels. Unlike Blacks, who were widely dispersed across the nation, Chicanos were primarily concentrated in five states in the American southwest. For every 100 Chicano children entering first grade, only 60 graduated from high school, only 22 entered college, and only 6 graduated from college. This indicated that programs intended to increase Chicanos in engineering would have to focus on precollege as well as college education. The study found that precollege Chicano students are held back by language, both reading skills in English and the low number of Chicano and bilingual teachers. Family expectations that females are less likely than males to continue their education was also cited as a factor.⁵

Puerto Ricans were concentrated mostly in the states of New York and New Jersey. The report found that high school and college completion rates for Puerto Ricans were lower than for Chicanos. English illiteracy was higher for Puerto Ricans than Chicanos. Puerto Ricans were channeled more than White or other minorities into vocational rather than college preparatory public education programs.

The report concurred with a Presidential message to Congress that American Indians are “the most deprived and most isolated minority group” in the nation. (Sloan 1974, p. 53) Many American Indians received their public school education through special Bureau of Indian Affairs-run schools or Church-run mission schools. American Indians experienced elevated high-school dropout rates, and of those who did graduate from high school, they were half as likely as White high-school graduates to matriculate in college. Once matriculated in college, American Indian students had attrition rates not much different from other underrepresented minorities. The report argued that close ties to family and tribe, as well as differences between American Indians and Whites regarding cultural values about individual versus collaborative action, placed stress on American Indian students. There were even more acute problems in finding bilingual and bicultural teachers for American Indian children than there were for Spanish-speaking children.

Responding to this situation described in the 1974 Sloan report, a number of new organizations were formed to support the advancement of ethnic and racial minorities in the science and engineering disciplines. These included the American Indian Higher Education Consortium (AIHEC) in 1972, Society for Advancement of Chicanos and Native Americans in Science (SACNAS ) in 1973, the National Action Council for Minorities in Engineering (NACME ) and the Society of Hispanic Professional Engineers (SHPE ) in 1974, the National Society of Black Engineer s (NSBE ) in 1975, and the American Indian Science and Engineering Society (AISES) in 1977. Background material on this exogenous force is covered in passing in Chaps. 3, 4, and 5, and the stories of these support organizations are told in Chap. 7.⁶

The third important exogenous force shaping the efforts to broaden participation in computing is the reverse discrimination environment that developed in the United States in the 1990s and that is still in force to some degree today. It was a conservative reaction to civil rights legislation passed between the 1960s and the 1980s; and it tried to undo preferential policies that had been given on the basis of race or gender through affirmative action programs, e.g. relating to employment or admission to higher education institutions. This issue is discussed in connection with access to higher education for African Americans and Hispanics in Chaps. 3 and 4 and in connection to computer science re-entry programs for women in Chap. 10. This author has also discussed the impact the reverse discrimination environment of the 1990s had on the programs of the National Science Foundation to broaden participation in the STEM disciplines in Aspray (2016).

The fourth exogenous force that we mention here as shaping efforts to broaden participation in computing is IT workforce needs. The demand of employers for IT workers has inexorably increased in the United States over the past 75 years as information and communication technologies have become embedded in ever more aspects of work. However, the demand has not increased at a steady or predictable rate; instead, IT workforce demand is characterized by cycles of demand and glut, of uneven duration and ferocity. A good example is the late 1990s and early 2000s, when there was an insatiable demand for workers to resolve Y2K problems of legacy computer systems and fuel the dot-com boom – only to be followed in 2002 with a dot-com bust that led to numerous IT worker layoffs. One might think that a demand for IT workers would be a good situation for women and minorities who were underrepresented in the workforce. However, employers generally preferred to fill open IT positions with foreign workers rather than with Americans from underrepresented groups. This was the story of the fierce battles around H-1B visas around 2000 and over outsourcing of IT work 5 years later. A second way in which this exogenous force played out is that higher education enrollments in computer science increased rapidly when industry increased demand for IT workers. The higher education system is not very flexible in its ability to meet rapidly changing enrollments. In the face of too many students, many computer science departments introduced weed-out courses that had the impact of weeding out many students with weak high school education or low self-confidence. These courses had the inadvertent effect of disproportionately weeding out women and underrepresented minorities. They sometimes also introduced higher entrance standards that also often affected women and underrepresented minorities in a disproportionate manner. This exogenous force is discussed in passing in Chaps. 8 and 9.

1.1.2 Pipeline Versus Pathway

Many analyses of the issues related to broadening participation in computing use the metaphor of a pipeline: if you do not take the right preparatory courses in middle school and high school, you cannot be admitted into an undergraduate major in one of the computing disciplines; if you do not have receive an undergraduate degree in a computing discipline, you cannot be admitted into a graduate program in a computing discipline; if you do not receive a graduate degree in a computing discipline, you cannot obtain a position in a high-level computing occupation, such as professor or senior researcher in industry. The iconic embodiment of this argument is a paper that appeared in Communications of the ACM in 1997, written by the computer science professor Tracy Camp and entitled “The Incredible Shrinking Pipeline.” (Camp 1997) Camp’s paper described a pipeline that leads from high school to a good computing career, but one that is leaky in many places, losing people from the pipeline at each of these transitions; so that only a few – too few – individuals were able to achieve the desired job at the end of the pipeline.

Camp is by no means the only person to use this metaphor; indeed it is commonly found in discussions of broadening participation in computing and also in efforts to craft solutions: for example, how do we make the pipeline less leaky so that more people can make the transition from high school to the undergraduate computer science major? The metaphor perhaps makes the most sense in the case of the computing occupation that is of primary interest to Camp, becoming a professor of computer science in a research-intensive university. It is not a perfect metaphor even in this occupational setting, however, because there are people who take a different pathway to their computing professorship, for example through the reentry program discussed in Chap. 9 that operated at Berkeley (to enable people who majored in a non-computing subject as an undergraduate to prepare for graduate student in computer science) or by training in a traditional way for the professoriate in a different, computing-intensive field, such as physics or economics, and then being hired into a computer science faculty position.

In point of fact, there are many different computing occupations; and these occupations have widely varying educational requirements.⁷ It is hard to count the numbers accurately, but perhaps only a quarter of the people who hold positions in computing occupations have any formal education in computing. (See Freeman and Aspray 1999.) In 2004, the Committee on Equal Opportunity in Science and Engineering, the esteemed advisory body created by the U.S. Congress in 1980 to advise the Director of the National Science Foundation on issues of broadening participation in the various science and engineering disciplines, specifically renounced the pipeline metaphor and advocated replacing it with a pathways metaphor that emphasizes the many different possible pathways to a computing career.⁸

The 2004 CEOSE report to Congress included a review of NSF ’s broadening participation efforts. Analyzing NSF programs since 1980, the Executive Summary of this report highlighted the need for a changing framework for understanding the nature of the process by which one prepared for and entered a STEM occupation. This is where they discussed moving from the pipeline metaphor to the pathway metaphor⁹:

Early efforts to broaden participation focused primarily on encouraging individuals from underrepresented segments of the population to enter STEM disciplines. This “pipeline” metaphor is a way of looking at the persistence of women, minorities, and persons with disabilities in STEM statistically. It emphasizes attracting students into the STEM “pipeline” when they are young, and spotlights the points at which “leaks” occur, differentially draining away individuals from underrepresented groups. Today, many efforts to make science and engineering more inclusive are paying attention instead to the multiplicity of “pathways” by which persons from underrepresented groups can enter and progress through STEM careers. Creating viable pathways requires addressing the tough issues related to what invites children to learn science (attraction), what causes young people to choose to keep learning mathematics and science (retention), and what then leads students to graduate (persistence) and continue into STEM careers (attachment). ( CEOSE 2004)

This emphasis on pathways rather than pipeline is probably even more appropriate today, given the many informal pathways to a computing education through online courses, computing boot camps, and hackathons.¹⁰

Several scholars studying broadening participation in computing have criticized the pipeline metaphor. Jolene Jesse , a program officer at NSF who studies issues of women in science, is one. ( Jesse 2006) An earlier study in which she participated, jointly conducted by the American Association for the Advancement of Science and the Committee on Professionals in Science and Technology , examined nontraditional pathways into the computing workforce.

A ‘nontraditional pathway’ is defined as the path taken by a nontraditional student, i.e., someone who: delays enrollment at least three years after graduating from high school or earning a GED; attends college mostly part-time; takes longer than six years to complete a degree; is employed full-time during most of their studies; or has dependents while attending college. (MIT Press, Scholarship Online, http://mitpress.universitypressscholarship.com/view/10.7551/mitpress/9780262033459.001.0001/upso-9780262033459-chapter-8)

These pathways are completely missed by the pipeline metaphor.

Mark Guzdial , a computer scientist at Georgia Tech who is a leading scholar in computer science education, writes in his computer education blog:

By using the “leaky pipeline” metaphor, we stigmatize and discount the achievements of people (women, in particular in this article) who take their technical knowledge and apply it in non-computing domains. Sure, we want more women in computing, but we ought not to blame the women who leave for the low numbers. … [N]ew research of which I am the coauthor shows this pervasive leaky pipeline metaphor is wrong for nearly all postsecondary pathways in science and engineering. It also devalues students who want to use their technical training to make important societal contributions elsewhere. (Guzdial 2015)

Lecia Barker , a professor in Department of Information Science at the University of Colorado Boulder and a senior research scientist at NCWIT , has noted that the pipeline metaphor can be counterproductive when trying to broaden participation in computing because it accepts as lost those individuals who leaked out of the pipeline at an earlier stage rather than finding a way to prepare them for a computing career. (Private communication to author, November 2015)

Historian of computing Thomas Haigh has also criticized the pipeline metaphor, but his criticism is that the pipeline metaphor focuses too much on the formal educational system instead of on the workplace. Haigh (2010) writes:

History broadens our perspectives. The literature on women in computing is dominated by discussion of computer science education. Fixing computer science is equated with fixing computing. This is justified by the metaphor of the pipeline carrying women from specialist education into IT work. Yet we saw that the gender dynamics of data processing were well formed by the 1960s, before undergraduate computer science education was an appreciable factor. Gender dynamics were shaped instead by the specific historical legacy of data processing work and the broader gender politics of corporate society. So to understand gender segmentation in the workforce, we must study the workplace as well as the classroom.

Some of the more recent social science research on underrepresentation is beginning to adopt a pathway metaphor. See, for example, Fox and Kline (2016), which discusses women faculty in computing from a pathways perspective.

1.1.3 Fixing the People Versus Fixing the System

Many of the interventions employed in broadening participation in computing principally address what is sometimes called “fixing the women”. These interventions include, for example, women’s support groups, special summer-before-college bridging programs for minority students, special sections of introductory computer science for people with limited computing experience, special financial aid programs for minorities, special mentoring opportunities for underrepresented groups, and conferences such as the Grace Hopper and Tapia celebrations, which target women and minority populations, respectively. These interventions can help to rectify shortcomings an individual has experienced, such as coming from a poor family or having a poor public school education – and some of these activities, such as the Grace Hopper conference – have fiercely loyal supporters.

However, there is also a negative side to these interventions: they can stigmatize underrepresented groups and reinforce stereotypes and unconscious bias about who should be doing computer science. Moreover, some social scientists believe that the track record of such interventions show that they are not effective at making the wholesale change of broadening participation in computing sought at the national level. Some organizations such as the National Center for Women & IT, and some individual researchers such as Mary Frank Fox and Gerhard Sonnert , believe that substantial change comes only through fixing the system rather than fixing the individuals. Systemic change is difficult, and it is unsettling to those who have existed in the established system. It is not surprising that, for many years, technology companies were more engaged in recruiting new women and minority hires rather than changing the practices and environments within their organizations so as to improve retention and advancement of women and underrepresented minorities. Chapter 10 discusses some of the issues associated with broadening participation in computer science departments; and Chaps. 8 and 9 discuss this issue in passing as various organizations, including NCWIT, are discussed.

1.1.4 Nonprofit Organizations and Individual Change Agents

Much of this book (Chaps. 6, 7, 8 and 9) focuses on nonprofit organizations. Many of these organizations had their origins in the concerns of a few scientists, engineers, and computer scientists who wanted the STEM disciplines in general or the computing discipline in particular to be equally open to all people, regardless of race or gender. These organizations can achieve what the concerned individuals – who often want to devote their main efforts to their scientific or engineering career – cannot do alone. Organizations have staff members who can handle the management of programs and events, provide effective communication to multiple stakeholders, and conduct business with safeguards that enable these organizations to act professionally and ethically. All of these organizations have strong advisory committees populated by members of the scientific professional community – many of whom volunteer large amounts of time and have strong identification with the organization. As nonprofit organizations, the most common funding model is project grants from either government (especially the National Science Foundation) or corporate entities. The most successful of these broadening-participation organizations focused on computing at this time are the CRA Committee on Women in Computing Research (CRA-W) and the National Center for Women & IT (NCWIT).

Equally important in this story of broadening participation in computing are individuals who serve as change agents – people who have served to transform their organizations or their profession through their individual actions to make them more effective at carrying out this broadening participation mission. Some of them are known for their roles in specific organizations, e.g. Jan Cuny ’s work in creating the Broadening Participation in Computing and the CS10K programs at the National Science Foundation; Anita Borg ’s work in creating Systers, the Grace Hopper conference, and what is now called the Anita Borg Institute ; Lucy Sanders in creating NCWIT; and Richard Tapia in his inspirational work at Rice University . Others, such as the ethnographer Jane Margolis who studied both Carnegie Mellon University and the Los Angeles School District , have had a reach that has had more importance for its national dissemination of ideas than its contributions to a single institution.¹¹

1.1.5 Intersectionality

This book is organized with chapters specifically about women or about a particular racial group (African Americans, Hispanics, and American Indians). The world is not nearly so tidy a place as this chapter organization suggests. There are many variations across these individual racial categories, e.g. tribal differences among American Indians, or to take another example the geographical, racial, language, and cultural differences between Chicanos living in the western and southwestern United States and Puerto Ricans living in the eastern United States. Many individuals are of more than one race. In addition, there are many other groups that are underrepresented that are not discussed in detail in this book, e.g. as organized by age, various sexual orientations, or various disabilities. People who are underrepresented in computing by belonging to more than one of these underrepresented groups often have multiple and simultaneous issues at play. This is an important topic, but it is not addressed except briefly in passing several times in this book.¹²