The Denver Study

In 1999 I received a phone call from a lawyer in Boston representing the Congress of Hispanic Educators. Years before, this group had sued the Denver Public Schools for segregating minority students. The original desegregation lawsuit involved bussing students, but the case had evolved over time into one about language and culture. The federal court now required the Denver Public Schools to provide native language instruction to students who did not understand English. The U.S. Justice Department joined the suit as coplaintiff. Each group—the school district, the Hispanic educators, and the Justice Department—was represented by lawyers.

The court agreement specified in detail educational services the district should provide, such as teacher qualifications, criteria for enlisting and exiting students, and the size of classes. Federal court judge Richard Matsch (who presided over the Oklahoma City bombing trial of Timothy McVey) needed someone to monitor whether the school district was fulfilling its obligations. After consulting various parties, the judge appointed me court monitor based partly on previous work I had done in New York and Chicago. I envisioned a study to monitor the implementation of the educational program and report these findings to the court, the plaintiffs, and the school district. I expected considerable differences of opinion and political conflict from various stakeholder groups, that is, those groups that had the most to gain or lose in the program.

Language, Class, and Cultural Politics

Denver had a school population of seventy thousand students; fifteen thousand of these students could not speak English very well. They were overwhelmingly Spanish speakers, mostly recent immigrants from Mexico and Latin America. A few were Russian or Vietnamese speakers. Many students were illegal immigrants whose parents had come to Denver during the boom economy of the 1990s, when the Colorado population increased by 30 percent. The parents of the students built houses, cooked food, washed cars, and performed basic manual labor for the city and state. It is the policy of most American schools to accept students who appear at the school door and not question their citizenship. Denver schools followed this policy.

The city itself was dominated by an Anglo business establishment, and Anglos displayed ambivalent attitudes toward these recent immigrants. The Denver program was titled the English Language Acquisition (ELA) program to signal that its purpose was to teach English, not maintain Spanish. Bilingual instruction was a hot political issue, as it is in most of the United States. The old Latino part of town had become so crowded with the new immigrants that Latinos were moving to other parts of the city. African Americans—long-established residents—were being pushed out of their neighborhoods. Tensions between blacks and Latinos were high. Some blacks saw the Latinos as taking the available affordable housing and undercutting them for jobs. Political power was shifting as tens of thousands of Latinos moved in. When the study began, the governing school board was dominated by Anglos, but two Latinas had just been elected. As the study progressed, more Latino members joined the board.

Furthermore, many teachers and administrators in the Denver schools were Latinos who had come from southern Colorado and northern New Mexico, descendants of the old Santa Fe culture. Santa Fe, founded in 1610, is the oldest capital city in the United States. These people have a distinct cultural identity predating Anglo settlement by centuries, and they consider themselves Spanish Americans, not Mexican. Other teachers and administrators were Chicanos, United States–born descendants of Mexicans whose ancestors had come from Mexico generations before, usually as migrants to pick crops.

These two Latino groups spoke both English and Spanish and staffed many educational positions in the Denver schools. Although they identified with the new Latino immigrants, they also saw them as different. The immigrants came from poor rural villages, mostly in Mexico, and were sparsely educated in any language. Ethnic, cultural, and class differences among the Latino groups themselves generated misunderstandings. For example, some immigrants took their children out of school for weeks to return to their home villages in Mexico for fiestas, which infuriated some professional educators. The teachers and administrators saw the loss of a month of school time as a serious setback for the students.

Some immigrant parents wanted their children to go directly into English classes so they could learn enough English to quit school, find jobs, and help the family. Again, this was not what the professional educators thought best for the students. Most new immigrants wanted their children in Spanish classes first, then in English classes after three years. That was the way the program was structured, accommodating a gradual transition to English classes. According to the court agreement, parents had the choice as to what classes their children should take, but the school district had a strong preference that students be placed according to their limited English ability.

Over the years of the lawsuit (this is a famous legal case called the Keyes case), some school officials and some plaintiffs had deepened their distrust of each other. Each side considered the other suspect. Some school personnel suspected the plaintiffs wanted to build a Latino political base in Denver; some plaintiffs thought the schools did not really want to provide good educational services for these students. In my early encounters with both sides, these hostile attitudes came through forcefully. I was told that the other side was untrustworthy. Personalities rubbed each other the wrong way: such-and-such was “unprofessional”; so-and-so was “a snake in the grass.” Much stronger language was expressed. When the monitoring began, passions were inflamed in some quarters, and the bilingual program was controversial.

Research Plan

My plan was to reduce distrust among these parties by involving the major stakeholders in the research study and by making my own actions transparent. I did not want any group to see me as siding with the other groups or as being duplicitous. Circumstances were ripe for misunderstanding. When I first announced that I was trying to make the monitoring study transparent, one administrator told me that was a big mistake. Why didn't I just act with the authority of the court? The other side had no choice but to accept my research findings. In any case, the opposition was not going to change.

To foster mutual understanding, I brought the representatives of the contending parties together face-to-face twice a year to discuss the research findings and to allow the parties significant input into the process. Because many participants were lawyers, adversarial by occupation (and some would say nasty by disposition), the meetings had some contentious encounters. Although I set the agenda for the meetings and chaired them, I could not anticipate what would occur when the parties met. I structured interaction around information and issues we, the court monitors, and they thought significant. In general, the sessions were cordial, whatever people said about each other privately.

At the beginning I intended to use data from the district's new management information system to identify schools that appeared deficient in implementing the program. However, the implementation of the data system fell far behind schedule. I had to do something else. People expected me to provide findings that would indicate whether the program was being implemented successfully. I constructed a checklist based on key elements of the program. By visiting schools, I could judge whether each school was in compliance. I submitted the checklist to all parties to ensure the items on the checklist were the most significant program features. People made useful suggestions, and I revised the checklist. I visited a few schools to see what collecting the data would be like. With more than one hundred schools involved, there was no way I could collect the data myself. Sampling schools—choosing some to represent the entire district—did not seem viable either because determining whether each school was in compliance was important.

I hired two retired school principals from Denver to visit and rate the schools using the checklist. It would have been easier to use graduate students, but graduate students would have had little credibility with administrators and teachers, and it would be too easy for principals to fool them. By contrast, the former principals knew how the Denver schools worked. When they were fed a suspect explanation they could sense it; they had been in similar positions, and they knew the program, the personnel, and the students. Because they were former principals from the school district, the central office trusted them. Because they were Latinas, spoke fluent Spanish, and had supported the ELA program and the Hispanic educator lawsuit from the beginning, the plaintiffs also trusted them. Without the trust from both sides, there was no sense in collecting the data.

The former principals did lack research experience, and sometimes they reverted to their former roles as helpers by acting as consultants. I had to prevent this, because this behavior would bias the findings. I held regular meetings with them to discuss their findings school by school and to remind them they were researchers now. It is surprising how well people can assume a role once the role expectations are clear. To help separate their opinions from the monitoring, I added a section to the checklist where they could provide their professional observations, with the understanding that these comments did not have to be based on the elements of the legal agreement. That helped them (and me) sort out their ratings of schools from other (often invaluable) insights. They felt better because their professional insights did not go unrecognized.

Discussing what was going on with the former principal–data collectors helped me construct an image of how the program was functioning overall and in depth. From them I had insights into what was happening and why that I could never have learned otherwise. For example, we might discover that a school principal was undercounting the number of eligible students deliberately. Why would the principal do that? I would have no idea. My colleagues might suggest that the principal was concerned about losing her expert veteran teachers who had been with her for a long time. The legal agreement stipulated that when numbers of eligible students reached a certain level, Spanish language teachers must be introduced to provide instruction. That could mean that regular teachers would have to transfer to other schools. The principal was therefore protecting her veteran teachers. Although we could not solve the problem, we could alert the school district staff to seek solutions. Alerting the district and the court to problems in particular schools was part of our job. In some research this reporting might be a breach of confidentiality, but in this study reporting problems of implementation was our basic task.

Enlisting these two former principals as coresearchers was one of the best things I did in the study. They not only could communicate with immigrant parents, teachers, and administrators but also were able to detect when things were awry when I had no idea. I could not have obtained this inside knowledge using a more traditional research approach, for example by using data collectors who had no special knowledge of the program. As a check on our site visits, I encouraged the ELA program staff to challenge our findings about particular schools when they disagreed. They were forthcoming when they thought we were wrong, and we hashed out disagreements face-to-face. Eventually the ELA staff members developed their own similar checklist so they could anticipate which schools had problems and fix them.

As the management information system improved, I also developed quantitative indicators of progress in program implementation based on district data. One indicator was the percentage of eligible students enrolled in the program. The other indicator was whether students with the least English ability were placed in appropriate classes. Again, I discussed these indicators with all parties until everyone accepted them as measures that showed the progress of the program. Higher percentages of students in appropriate places meant the program was being successfully implemented. The development of the information system was slow and tortuous, reflecting how difficult it is to obtain accurate information in such organizations. Data had to be collected at the school level, entered into the data system, and aggregated. Errors plagued the process every step of the way. It cost the district a huge effort to obtain reliable data, but school district personnel managed it over two or three years.

When the data were reasonably accurate, our quantitative indicators showed gradual improvement year by year. Improvement was slower than anticipated, as is often the case in new programs. The indicators also showed which particular schools were in trouble. By combining our on-site checklists with the school-by-school indicators, we had a cross-check on where things stood. When schools looked bad, we revisited them to look again, and the district sent staff members to these schools to tackle problems. Finally, we could combine the checklist findings and school indicators to provide a summary of where the implementation of the ELA program stood at any given time. Good profiles of individual schools obtained by our data collectors were expected to be matched with increases in the percentages of students in the program and in the appropriate classes within the program.

Constant Change

Constant change in the school district was a complicating disturbance. School principals were retiring, resigning, and being replaced or promoted constantly. New principals meant a new situation, and we revisited schools that had new principals. Students dropped out, moved mid-term, went back to Mexico, or disappeared from the school rolls altogether. Some schools had more than 100 percent student turnover. Of course, student turnover was not unexpected.

More surprising was the turnover in school superintendents. In the six years of the monitoring study there were five different superintendents in charge of the Denver school district. Principals were accustomed to running their buildings without much interference, and each new superintendent disrupted long-standing patterns of behavior. Each superintendent had different goals for the district, and each reorganized the district administration. I had to establish a professional relationship with each new superintendent, and I had to give superintendents time to understand my role and figure out how their plans would affect the monitoring. The ELA program directors also changed. There were four during the study. I admired these people and the difficult challenges they faced. They were responsible for implementing the court agreement, yet had no line authority within the school district. They could persuade others or refer problems up the chain of command, but they could not order principals or teachers to do anything. Yet when things went wrong they were held personally accountable. Some lasted a long time; some not. Establishing a working relationship with them was critical for me and my two colleagues.

Power Shifts

Meanwhile, I met with interested groups in the community, including the militants, both those opposed to the program and those wanting Spanish in all schools. I listened, responded to their concerns, and included some of their ideas in my investigations. I followed up on information these groups provided. I turned down no one wanting to offer views, though I did not accept their information at face value. For example, the most militant Latino group wanted Spanish language classes to be maintained in all schools along with English classes. I met with the leader in a café that served as political headquarters and listened to her concerns. There was little I could do about continued Spanish classes for all students because the court agreement precluded it. That was beyond the purview of the study. However, I did investigate certain practices reinforcing her view that the district was insincere in providing services to students in the program.

My periodic, written summary reports went to the court. As court documents, the reports were public information the media seized on. I asked the school district and plaintiffs how they thought I should handle the media. Bilingual education was such a hot topic I knew the reporters would be after comments. All parties preferred that I not talk to the media. In their view, it would inflame the situation and make implementation of the program more difficult. I took their advice, referred inquiries to the parties themselves, and made no comments outside my written reports. The media accepted this stance reluctantly and quoted my reports in the newspapers.

After six years of monitoring, the program was almost fully implemented. The conflict seemed defused, at least for the time being. The opposing parties could meet in a room without casting insults at each other. I am not saying the groups loved each other, but they could manage their business together rationally. The strife and distrust was much less than when we started. The school district had established its own monitoring system. Even so, the plaintiffs were reluctant to let the district out of the court ruling.

District politics had shifted, with more Latinos having been elected to the school board. In fact, the daughter of the man who had sponsored the original lawsuit became chair of the school governing board, and the last superintendent, a lawyer who had been the mayor's chief of staff, adopted a strong pro-Latino attitude. (He was appointed a U.S. senator in 2009.) Under these circumstances, one plaintiff lawyer and the district lawyer thought the monitoring was no longer necessary. The plaintiff lawyer had long disagreed with us about whether the district was forcing students into mainstream classes inappropriately. We could find no evidence for that, but he thought we did not look in the right places. It seemed advisable to end the study because one side had gained the upper hand politically, and we had already been evaluating for three years longer than originally planned.

I had a mixed reaction to ending the study. I wanted to resolve the lawsuit in court, but that wasn't my assignment, and at the time the district had no motivation to take the case back to court. On a more positive note, however, the district and plaintiffs had reached a new level of understanding and cooperation. They could work things out without a third party. Time to quit.