From the late 1970’s through 1990 I conducted numerous public policy related studies. This post relates stories of how politics and research mixed and ended with my exile from Washington. Only a distinct minority of studies in which I was involved are the subjects of the following stories.
Research in Berkeley
Disabled in Science NSF Grant: In 1977 I got a grant from the National Science Foundation of $50,000 to study discrimination against the disabled in the sciences. The grant was administered by the Lawrence Hall of Science. The grant program was particularly interested in a prior finding from a survey of scientists that there were twice as many disabled women in science than disabled men; this in contrast with the large majority of respondents being male. I found that primary cause of this unexpected result was that given identical physical conditions, women were much more likely to self-identify as disabled. The grant program found this result most unwelcome because most of their grants went to programs to promote disabled women in science. The second year of the grant program, all research was eliminated in favor of exclusive emphasis on demonstration projects.
Berkeley Planning Associates child abuse and neglect studies. In 1978 and 79 I was the head statistician on studies of child abuse and neglect. The federal government had funded a large number of pilot projects to address the issues. The pilot projects took a wide variety of approaches; there were peer counseling groups, Freudians, feminists, support groups, mother training and so on. A condition of the grants was that they would have to submit to an evaluation by BPA. At the beginning all the projects were pleased with the evaluation and believed it would demonstrate the effectiveness and indeed superiority of their particular approach. As the project advanced individual programs reported that their caseloads were much tougher than the had assumed but nonetheless they were confident of ultimate success. Six months later the projects began to express doubts about the whole idea of evaluation. Things were not working out as planned. The projects began fudging their numbers in multiple ways. Cases of various kinds were systematically reclassified as sexual abuse as sexual abuse cases became eligible for additional funding. The ugly reality of child abuse and neglect was denied and distorted to fit existing ideologies. Most abuse and neglect was perpetrated by women but projects would ignored this fact and stuck to standard male-blaming rhetoric. Similarly class, racial, and ethnic distributions were denied in favor claims the problem was equally prevalent in all sections of society. (This latter distortion was particularly promoted by programs in upper-middle class neighborhoods to protect their funding.)
Berkeley Planning Associates evaluation of Independent Living Centers. In 1979 I was the head statistician for an evaluation of community independent living centers in California. The independent living moving had originated in Berkeley with Ed Roberts and the Berkeley Center for Independent Living. When Ed became director of the department of rehabilitation for California, he got through the state legislature funding to create ILC’s throughout the state. The legislature mandated that an evaluation of the centers be conducted. I was the head statistician for this evaluation. The centers were all staffed by young idealists who were committed to doing good for their communities. The vision of the centers was that they would take institutionalized disabled people, help them out of the institutions or parent homes to independent living. They would also get the training and support they needed to become employed and productive members of society. All staff I met in the centers deeply believed in this vision. The evaluation found that the centers provided many necessary services to disabled members of the community. They allowed people to live independently and get out of their parents homes. The centers provided important employment support services. However, we were unable to find a single case where a client had moved from an institution to independent living to gainful employment. When I raised this with center staff they were surprised, they had not noticed. The ideology was so large, hopeful, and convincing that it was never questioned. Staff everyday success in disabled services–transportation, wheelchair repair, attendant hiring, sign translation, and many others–hid the larger picture. Ed was very unhappy with the results. The final report did not mention this key result but instead emphasized the services being provided.
(Martin Trow at the Center for Studies of Higher Education. I did not participate in this study. What I know of it was reported to me by my fellow graduate student Jessie Cloud, a programmer on the study. Jessie quit graduate school declaring he was just sick of all the bullshit. Trow directed a study of discrimination in education looking at minorities and women. What Jessie told me was that the study had a major finding of discrimination that perfectly fit all their measures of bias but was never reported. The study found major discrimination in favor of Jews; they were admitted to top universities with lower grades and scores, hired as faculty with fewer qualifications and promoted faster and with fewer accomplishments than non-Jews. Such findings were not just politically unacceptable but reporting them would have been academic suicide.)
In Washington
Fraud in the Free and Reduced Price School Lunch Program . At Applied Management Sciences, the first major study for which I was the principle investigator was of the school lunch program. Reagan administration was intent on ended all welfare fraud. In fact, they largely equated welfare with fraud and used findings of the existence of fraud as a rational for ending social welfare programs. A prior study of the school lunch program had shown at many participants had underreported their income and were getting benefits for which they were not legally entitled.
Schools had great difficulty in verifying parent income. Free and reduced price lunch eligibility had to be determined in the first week of school when most were generally in a state of semi-chaos. They were overwhelmed by the task of getting all the right teachers and students in the right rooms, making sure they had the necessary text books and supplies, checking all the vaccination and health records, and a myriad of other task. Neither teachers nor administrators had the means nor skills to verify parent income.
Given this reality there was significant political press to simply end the program. There existed important evident the the free and reduced price lunch program provided nutritional support for millions of children who would have otherwise had deficient nutrition. This evidence while not ignored was not deemed sufficient grounds for continuing the program. The Food and Nutrition Service was going the mandate to either fix the fraud problem or end the program.
The study investigated 12 alternative approaches in a nationally representative sample of 13 school districts and 80,000 program applicants.
The solution I devised to the dilemma was to deter fraud in the first place so that verification of income was not necessary. Fraud was deterred by changing the application form used. The prior application was a model of simplicity; an applicant family had only to report total monthly income and number of family members. To help, the back of the form had a table showing maximum income by family size to obtain a free or reduced price lunch.
My alternative application form was quite different. The single income line was replaced by a matrix that required listing income, by type, for all household adults. This matrix approach followed from a finding that families would often not regard certain income sources as “family” income. The husband’s earnings would be family income, the wife’s earnings “her money” or grandma’s pension would not be counted. If the family obtain no income from a given source they were required to report “0” rather than simply leaving the line blank based on a finding that applicants were more likely to sin by omission than commission. The matrix produced 25% higher reported income than the single line approach. At the bottom of the form was a section that said “School Use Only” that contained check boxes for the type of income verifications the school conducted on the reported income and to what local, state, or federal authorities they reported the results. This section was included to promote the impression of vigorous verification and dire consequences of misreporting. Finally the eligibility income levels were removed from the back of the form.
The fraud control form, when tested, was shown to substantially reduce fraud. When adopted it produced an estimated $169 million in savings, without any measurable effect on providing needed nutritional support to at risk children. For better or worse, the form is still in use today with very few changes . (Click on the following link school-lunch)
The school lunch study was the beginning of a long career in fraud control Before I left Washington my group dominated the fraud control business creating control systems for Food Stamps, AFDC, Medicare, Medicaid, Student Financial Aid, the IRS, and many others. Collectively these systems produced billions of dollars in savings each year. (I have no notion of how many others are still in existence today and if they are still functioning in any meaningful sense.) The study also originated by belief that the key to fraud control was deterrence rather than detection.
Sexual Sterilization Regulation Evaluation. Arriving at AMS I took over the Sexual Sterilization Evaluation. The project was mandated by Congress to review associated federal regulations. I first envisioned the project as centered around a set of surveys of affected groups. However, political considerations soon changed everything. The election of 1980 brought into HHS a new conservative administration. Richard Schweiker was appointed Secretary of HHS, the nominal liberal of the administration. Carolyne Davis was appointed head HFCA. Davis was a strong conservative ideologue. Davis quickly announced opposition to my project as part of a larger attack on Schweiker. In a memo she wrote that the project was a Carter era effort to impose burdensome regulations on the health care industry and involved family planning, to which she was deeply opposed. Schweiker was on verge of canceling the evaluation. In response I a ghost wrote a memo in Schweiker’s name saying that he was promoting the project as a review of possibly burdensome regulations and unnecessary federal intrusion into individual lives. If Davis was comfortable with just continuing Carter’s regulations without review, he was not. And so the evaluation continued.
Almost immediately a second political barrier emerged. Any survey conducted by the federal government was subject to review by OMB under the Paperwork Reduction Act. The Paperwork Reduction Act that no survey with more than nine respondents could be imposed on the citizenry without OMB approval. OMB, which was headed by David Stockman made the same ideological objections to evaluation as Davis. So again the project, and my job, appeared to be at an end. As I solution I proposed a large number of small surveys with nine respondents each of affected subgroups. We visited Indian Health Clinic across the country, traced the migrant farm worker trail and visited their clinics. We interviewed advocates for the mentally disabled, poor minorities, doctor groups, and so on. We conducted an exhaustive literature review and interviewed all leading experts on the subject.
Liberal advocacy groups strongly weighed in. The National Organization of Women (NOW) was conflicted. They recognized the advantages for women of a required 30 day waiting period between signing a consent form and the sterilization operation. The waiting period prevented unconsidered or coerced consent during birth and helped minimize subsequent regret. However the required waiting period might be seen as limiting a woman’s right to choose, a NOW slogan. As a result, NOW would only back the regulation if it also applied to men. The regulation made little sense for men; they did not give birth so could not be coerced then. Men experience much less post-sterilization regret. Nonetheless, resulting regulations contained a waiting period for men to satisfy the political agenda of NOW.
Ralph Nader’s health care group was also involved. They had written an expose on sterilization of the mentally retarded that showed multiple examples of abuse. The Nader report contributed to the regulations prohibiting such sterilizations without full informed consent of the patient. Many severely retarded individuals could not give such consent and so sterilization was effectively prohibited. From my research I concluded this solution was inadequate. I came to believe that such a ban harmed my retarded individuals. Many were fully capable to leading independent lives but not of being parents. Among this group contraceptive failure was common. Some severely retarded women would go into major panic every month when they had their period.
Looking a state-specific regulations, I concluded there existed alternatives that gave access to sterilization for the retarded while protecting their rights and interests. However, the Nader group was categorically opposed to anything but a full ban. I became quite disgusted with the Nader gang because they appeared to me totally ignorant of the on-the-ground realities, and were much more concerned with their own moral posturing (and fund raising) than the needs of the retarded. The Nader group prevailed and the ban was perpetuated.
Responding to the ban parents and caregivers of the mentally have formed underground networks of providers who will perform the operation outside any public scrutiny. The outcome is that the rights of the retarded are routinely abused rather than protected.
Once the study was complete, the Public Health Service sent it to the Centers for Disease Control for expert review. I was very nervous. Although we had done our best, without large-scale surveys, I could not produce definitive numbers on the health impact of the regulations. To my surprise, CDC loved it. All the interviews with impacted individuals gave the study a rich human texture that the epidemiologists found lacking in their statistics.
The study produced a report to Congress which really means a report to congressional committee staff. The time congressional staff was dominated by the majority Democrats and deeply skeptical of any Republican administration efforts to modify health care regulation. As a last political maneuver I leaked word to the relevant Democratic congressional staff that the study had been conducted by a Berkeley political lefty (me) and not to make any noise so we could slip our civil rights agenda past the oblivious Republicans. To the Republicans in HHS I conveyed the exact opposite message; we had thwarted the leftist agenda to control all health care but cloaked the report in bland technical jargon. The truth was neither. The needs of Americans for access to sexual sterilization services have nothing to do with the agendas to the left and right. Since the evaluation the federal regulations remain essentially unchanged.
Loan Standards at Treasury
AMS held a contract to provided statistical support to the Treasury Department on an as needed basis. Overtime I became the technical lead for this contract. One project was to develop a set of financial standards that would apply to all federal loan programs. The project originated at the Office of the President (OMB) in 1983. However, because OMB did not have contracting authority, Treasury took the lead and I directed all technical aspects of the project. (I have attached our statement of work and the project description. The actual work performed evolved over time. loan-program-standards )
At the time the federal government had over 150 loan. These programs were run under the authority a diverse set of departments and agencies such as HUD, Veteran Affairs, and the Departments of Agriculture, State, and Education. The total guaranteed debt exceeded $300 billion. The individuals running these programs often had little background in finance. Problems with growing loan default rates in several programs provoked concern at OMB and Treasury.
The programs were operated without any uniform standards for credit granting, risk analysis, write-off policies and loan sales. The projected aimed to develop such standards. The project was confidential and internal to the government.
Overseeing the project was a committee of assistant secretaries from the departments with the largest loan programs. I quickly learned that it was assistant secretaries that yielded the real day-to-day power of government. The political appointees created policies. The assistant secretaries translated these policies into action, or inaction, as they saw fit. Meetings were held in the “cash room” of main Treasury.
During the project I learned how liquid and shapeless was bureaucratic power. Administrative support for the committee was provided by Susan Brita. Susan was a mother who had recently entered the labor force once here children entered college. She held a low administrative support job ranking. Her office job was to schedule meetings, let members know when and where, arrange pastries and coffee, take notes and so on. And take notes she did. Further, she would call committee members in advance of each meeting and remind them of commitments they had made in prior meetings and ask for updates on their progress. Now this is not behavior one expects from a low-level support person. Soon, all committee members came to assume Susan was the real representative of the administration. Once an Assistant Secretary, during one of Susan’s update calls, reported that he had been unable to find funding to support project request. Susan replied that perhaps she could be of assistance. She would arrange a meeting with the Department’s desk at OMB where they might be able to find the necessary funds. A Department’s OMB desk is a very frightening place for government bureaucrats. It is the point were their funding is controlled. All Departments limit access to their OMB desks very strictly. Not surprisingly, the committee discovered he could find the necessary funding without Susan’s assistance.
Another job Susan held was scheduling of department limousines. Again the intent was purely functional and again Susan found access to power. The Treasury is right next door to the White House. There are underground tunnels between the two buildings. However, some senior Treasury official, for some unknown reason, where unable to get between the two buildings without a limousine. Limousine access was a prime topic of concern at Treasury’s highest levels. At the time Treasury was developing a new information system valuable at over $500 million. The internal executive memo system contained almost 30 memos relating to this system. There were over 80 relating to limousines. Susan explained those officials who were helpful for Susan’s projects somehow always had access to the limousines. Those not so helpful, for purely objective and verifiable reasons, often had problems getting limo.
The Director of Administration for Treasury was John Garmat. Garmat was a classic wheeler-and-dealer. He wanted me to guarantee to him that certain project audits and demonstrations would came out the way he wanted. In particular, he desired that a demonstration of D&B loan analysis products should prove successful in managing small business administration loans. I would not guarantee that outcome only a fair test (although I had great hopes for success.) D&B was very aggressive suggesting to me my actions could improve or end my career. I went to Susan with this problem. A week later Garmat resigned from Treasury. I found out later Garmat had told the President of AMS that I was naive as to the ways of power and should be fired. And I would have been fired if Garmat had not gone first.
A last story about Susan. In 1984 Terence Golden was named Assistant Secretary of Treasury for Administration. Susan was assigned to showing him around the building. Golden had no prior experience in government. Susan claimed all he knowe was that he voted every four years. The notion was she would show him where the executive washroom was, the security codes, etc. As she showed him around the building she explained how it all worked. Very soon Susan was his chief of staff.
Finally, about me. When I got this assignment I was a quite competent statistician. However, I knew nothing about finance nor running loan programs. My first step was to buy some expensive pinstripe suits. I sat up nights reading finance texts. I soon discovered the power of a title. Everyone on the project assumed Treasury had selected me as a leading financial expert. Being Dr. Finnegan and speaking the language of statistical risk analysis was enough to get me through without ever being questioned. I listened to everyone and learned fast.
Student Loan Fraud Control
Food Stamp Audit Control System.
The Temporary Emergency Food Assistance Program (TEFAP)
In 1987 Quality Planning saw its downfall in Washington consulting with our evaluation of the Temporary Emergency Food Assistance Program (TEFAP) . It was a large, year-long project. TEFAP was a program to distribute food purchased as part the Department of Agriculture commodity price supports. The surplus food was distributed through various private social assistance programs such food banks and senior centers. Large, five-pound blocks of cheddar cheese was then the primary commodity TEFAP was then distributing.
The study was designed to address three primary questions, who was getting TEFAP benefits, their income and participation in Food Stamps, and the nutritional value of the distributed food.
During the study our teams visited food distribution centers throughout the country and interviewed individual receipts including a complete nutritional and income analysis of recipient households.
Before, study results were released, USDA announced they would be eliminating the program because 1) Benefits were not well targeted to at need families, 2) Those families at need were already getting food stamps, and 3) TEFAP did not provide meaningful nutritional support to recipient families. (We indirectly learned that the decision to eliminate the program was made under pressure from Kraft Foods who believe that TEFAP was reducing the demand for their cheese products. More about this below.)
The conclusions of USDA were reasonable on their face: the program was not well targeted and had few controls to assure only the poor participated; the Food Stamp program did overlap with TEFAP; and the package of foods distributed by TEFAP was constructed on the basis of what USDA had purchased for price supports and not to match the needs of recipients.
To our surprise, the evaluation found TEFAP was, after all, a very effective program. We found that almost all recipients were quite poor (who else would get up early in the morning, take a bus to a food bank in a bad part of town, and often wait hours of a box of food not of their own choosing?). We found that although almost all recipients were eligible for Food Stamps, only about 25 percent of these were actually getting Food Stamps. Finally, our biggest surprise was that TEFAP appeared well-tailored to the nutritional needs of recipients. The largest single group of recipients were elderly women who needed the calcium and protein provided by the cheese.
When we submitted our report to USDA Food and Nutrition Service, they were not pleased. Various changes were demanded—but never in writing. I responded that as Congress had mandated the study, I could not ethically make most of the requested changes. One area USDA was particularly insistent was our finding of nutritional impact. While refusing to change the findings based what I saw as political pressure, I did agree a compromise suggest by my team of nutritionist. We would make changes if a leading nutritionist on the staff of USDA believed there were errors in our analysis.
The the review was under way USDA rewrote the executive summary of the report, the only part of these reports anyone ever reads, to downplay, obscure, and hide our key findings. We never saw the agreed to technical review. When I pressured USDA for a company I was assured that the internal nutritionist had found numerous errors and the results had to be adjusted.
Not long after the report was publicly released I got a call from FRAC, the Food Action Resource Council. FRAC was following up on rumors we had hidden our findings to fit the political agenda of the Bush administration. I assured them this was not the case. The only findings changed were in the nutritional analysis and this was based on expert peer review. We were experts in our field but not infallible and therefore peer review was justified. FRAC responded this is not the story they got from inside USDA. I therefore followed with my USAD contacts. The nutritionist who supposedly had conducted the review had never heard of the study and definitely had not review it. The set a copy of the study to the chairs of the nutrition department of Berkeley and Harvard and ask what they thought. In both cases they praised the methodology as far better than the usual USDA work. In particular we were praised for evaluating what recipients actually ate, taking account of waste and spoilage rather then simply what food was available.
Again I followed up with USDA and told them what I found. The new story was more telling. Why USDA was particularly upset with our methodology was it called into question of the nutritional analysis USDA used to set the level of Food Stamp benefits. USDA is required by law to set benefits at a level to meet the nutritional needs of recipients. Their method of doing this was to determine the lowest price a trained nutritionist could buy foods necessary to meet minimal nutritional standards. This method did not look at what foods recipients bought and fact and how those foods were consumed. If USDA had employed a method similar to ours, based on reality and not ideal theory, Food Stamp benefits would have had to be greatly increased. This definitely was not acceptable.
I reported my new discoveries to FRAC. Not long after I got a call from CBS News. I told them the above story. They sent out a film crew to interview me. and included the story as one instance of how the Bush Administration was suppressing science. The story was national news.
The shit hit the fan. QPC’s government contracts were audited. Our retirement plan was audited by the Department of Labor. Problems were found with every bill we submitted to the Federal Government. Payments were delayed or denied. Word was passed to us we would never again win a government contract. We fought all negative findings from the audits and won every dispute. However, the damage was done. Almost all our income was from Federal Contracts and QPC came within an inch of going bankrupt.
The upside of this story, was that Quality Planning changed its business model to concentrate on fraud in auto insurance. While the loss of federal contracts cost me millions, the move to insurance work made me tens of millions.
(A closing note on this story about Kraft Cheese: This whole dispute appeared to have been set off by Kraft Foods’ belief that TEFAP was suppressing the demand for Kraft products. While this was not part of our study, in the course of our interviews we did learn a lot on the topic. The reason USDA had so much cheddar cheese to give away was they had converted massive milk purchases into cheese rather than just poor the milk down the drain which would have been a very bad public image. To minimize storage costs, USDA made cheese that was almost pure cream. This cheese was then stored in caves under Kansas. Thus, the cheese distributed was very high cream and well-aged. Program recipients loved it. A particular advantage was that you could cut off a slice, put it in a frying pan for a couple minutes and have a great cheese sauce. Many interviewees reported TEFAP made them come to love cheeses. Between TEFAP distributions, these recipients would go to their local grocery stores to attempt to buy similar cheese. The problem was the quality of Kraft and other cheese products available. They were not nearly as good.)
The American’s with Disabilities Act
In 1988 both presidential candidates, Vice President Bush and Governor Dukakis, endorsed broad civil rights protections for people with disabilities. When George Bush was elected president, the possibility opened for passage of the Americans with Disabilities Acts (ADA). Bush had appointed Evan Kemp, who had been his roommate at Yale and was disabled, as head of the Office for Civil Rights. Bob Dole was majority leader of the Senate, and partially disabled during the Korean War, signaled his provisional support for the act.
Pat Wright of DREDF in Berkeley, an organization with which I had deep ties, went to Washington to coordinate the lobbying effort on behalf of the disabled community. While the bill had substantial support issues relating to costs threatened the prospects for passage. The New York Times ran an editorial questioning costs. Bod Dole made his support contingent on reasonable costs.
As a result of this pressure, I was contacted by Edward Kennedy’s staff to conduct a cos-benefit analysis. Because of the short time-frame, it was judged better to hire me as part of the Senate staff than to let a contract for the work. My work with the Bush and Reagan administrations made me acceptable to the Republican side and may work with DREDF and participation in the disability rights movement recommended me to the Democrats. It was a very different Washington then when issues could be rationally discussed, evidence counted, and political differences did not outweigh the common interest.
I reached out to various experts and lobbyist to obtain the necessary information. Ron Mace provided needed architectural information. For data on telephone relay system costs I contacted ATT. I then ran their cost data by various advocacy groups and conducted comparative analysis of costs of existing relay systems. ATT had made high cost estimates in hope of being awarded a profitable contract. When confronted with data that called these high costs into question, they modified their estimates to a more reasonable level.
Similarly bus systems, particularly Greyhound Bus made very high estimates of the costs of making their systems wheelchair accessible. On the other side, providers of wheelchair bus lifts, made much lower cost estimates in hope of greatly expanding their bus. A back-and-forth dialog I created between the bus systems and providers of wheelchair bus lifts resulting in substantial convergence and a more realistic estimate.
When I submitted my preliminary report, advocates for ADA where concerned my cost estimates were too low to be creditable. Going over the details, I could not find any strong justifications for substantially higher costs but did add notes emphasizing areas of uncertainty.
Looking back, there is no doubt I was correct overall, the ADA did not result in the ruinous costs opponents projected. I believe I mostly correctly estimated ADA costs of public transportation access, architectural access and relay services. I overestimated economic gains of disabled employment. I did not anticipate legal costs for small business access.
Now both Bush and Dole claim the ADA as the greatest accomplishment of their long political careers.
Oct 29 1984
My Washington Education: The Politics of Research
From the late 1970’s through 1990 I conducted numerous public policy related studies. This post relates stories of how politics and research mixed and ended with my exile from Washington. Only a distinct minority of studies in which I was involved are the subjects of the following stories.
Research in Berkeley
Disabled in Science NSF Grant: In 1977 I got a grant from the National Science Foundation of $50,000 to study discrimination against the disabled in the sciences. The grant was administered by the Lawrence Hall of Science. The grant program was particularly interested in a prior finding from a survey of scientists that there were twice as many disabled women in science than disabled men; this in contrast with the large majority of respondents being male. I found that primary cause of this unexpected result was that given identical physical conditions, women were much more likely to self-identify as disabled. The grant program found this result most unwelcome because most of their grants went to programs to promote disabled women in science. The second year of the grant program, all research was eliminated in favor of exclusive emphasis on demonstration projects.
Berkeley Planning Associates child abuse and neglect studies. In 1978 and 79 I was the head statistician on studies of child abuse and neglect. The federal government had funded a large number of pilot projects to address the issues. The pilot projects took a wide variety of approaches; there were peer counseling groups, Freudians, feminists, support groups, mother training and so on. A condition of the grants was that they would have to submit to an evaluation by BPA. At the beginning all the projects were pleased with the evaluation and believed it would demonstrate the effectiveness and indeed superiority of their particular approach. As the project advanced individual programs reported that their caseloads were much tougher than the had assumed but nonetheless they were confident of ultimate success. Six months later the projects began to express doubts about the whole idea of evaluation. Things were not working out as planned. The projects began fudging their numbers in multiple ways. Cases of various kinds were systematically reclassified as sexual abuse as sexual abuse cases became eligible for additional funding. The ugly reality of child abuse and neglect was denied and distorted to fit existing ideologies. Most abuse and neglect was perpetrated by women but projects would ignored this fact and stuck to standard male-blaming rhetoric. Similarly class, racial, and ethnic distributions were denied in favor claims the problem was equally prevalent in all sections of society. (This latter distortion was particularly promoted by programs in upper-middle class neighborhoods to protect their funding.)
Berkeley Planning Associates evaluation of Independent Living Centers. In 1979 I was the head statistician for an evaluation of community independent living centers in California. The independent living moving had originated in Berkeley with Ed Roberts and the Berkeley Center for Independent Living. When Ed became director of the department of rehabilitation for California, he got through the state legislature funding to create ILC’s throughout the state. The legislature mandated that an evaluation of the centers be conducted. I was the head statistician for this evaluation. The centers were all staffed by young idealists who were committed to doing good for their communities. The vision of the centers was that they would take institutionalized disabled people, help them out of the institutions or parent homes to independent living. They would also get the training and support they needed to become employed and productive members of society. All staff I met in the centers deeply believed in this vision. The evaluation found that the centers provided many necessary services to disabled members of the community. They allowed people to live independently and get out of their parents homes. The centers provided important employment support services. However, we were unable to find a single case where a client had moved from an institution to independent living to gainful employment. When I raised this with center staff they were surprised, they had not noticed. The ideology was so large, hopeful, and convincing that it was never questioned. Staff everyday success in disabled services–transportation, wheelchair repair, attendant hiring, sign translation, and many others–hid the larger picture. Ed was very unhappy with the results. The final report did not mention this key result but instead emphasized the services being provided.
(Martin Trow at the Center for Studies of Higher Education. I did not participate in this study. What I know of it was reported to me by my fellow graduate student Jessie Cloud, a programmer on the study. Jessie quit graduate school declaring he was just sick of all the bullshit. Trow directed a study of discrimination in education looking at minorities and women. What Jessie told me was that the study had a major finding of discrimination that perfectly fit all their measures of bias but was never reported. The study found major discrimination in favor of Jews; they were admitted to top universities with lower grades and scores, hired as faculty with fewer qualifications and promoted faster and with fewer accomplishments than non-Jews. Such findings were not just politically unacceptable but reporting them would have been academic suicide.)
In Washington
Fraud in the Free and Reduced Price School Lunch Program . At Applied Management Sciences, the first major study for which I was the principle investigator was of the school lunch program. Reagan administration was intent on ended all welfare fraud. In fact, they largely equated welfare with fraud and used findings of the existence of fraud as a rational for ending social welfare programs. A prior study of the school lunch program had shown at many participants had underreported their income and were getting benefits for which they were not legally entitled.
Schools had great difficulty in verifying parent income. Free and reduced price lunch eligibility had to be determined in the first week of school when most were generally in a state of semi-chaos. They were overwhelmed by the task of getting all the right teachers and students in the right rooms, making sure they had the necessary text books and supplies, checking all the vaccination and health records, and a myriad of other task. Neither teachers nor administrators had the means nor skills to verify parent income.
Given this reality there was significant political press to simply end the program. There existed important evident the the free and reduced price lunch program provided nutritional support for millions of children who would have otherwise had deficient nutrition. This evidence while not ignored was not deemed sufficient grounds for continuing the program. The Food and Nutrition Service was going the mandate to either fix the fraud problem or end the program.
The study investigated 12 alternative approaches in a nationally representative sample of 13 school districts and 80,000 program applicants.
The solution I devised to the dilemma was to deter fraud in the first place so that verification of income was not necessary. Fraud was deterred by changing the application form used. The prior application was a model of simplicity; an applicant family had only to report total monthly income and number of family members. To help, the back of the form had a table showing maximum income by family size to obtain a free or reduced price lunch.
My alternative application form was quite different. The single income line was replaced by a matrix that required listing income, by type, for all household adults. This matrix approach followed from a finding that families would often not regard certain income sources as “family” income. The husband’s earnings would be family income, the wife’s earnings “her money” or grandma’s pension would not be counted. If the family obtain no income from a given source they were required to report “0” rather than simply leaving the line blank based on a finding that applicants were more likely to sin by omission than commission. The matrix produced 25% higher reported income than the single line approach. At the bottom of the form was a section that said “School Use Only” that contained check boxes for the type of income verifications the school conducted on the reported income and to what local, state, or federal authorities they reported the results. This section was included to promote the impression of vigorous verification and dire consequences of misreporting. Finally the eligibility income levels were removed from the back of the form.
The fraud control form, when tested, was shown to substantially reduce fraud. When adopted it produced an estimated $169 million in savings, without any measurable effect on providing needed nutritional support to at risk children. For better or worse, the form is still in use today with very few changes . (Click on the following link school-lunch)
The school lunch study was the beginning of a long career in fraud control Before I left Washington my group dominated the fraud control business creating control systems for Food Stamps, AFDC, Medicare, Medicaid, Student Financial Aid, the IRS, and many others. Collectively these systems produced billions of dollars in savings each year. (I have no notion of how many others are still in existence today and if they are still functioning in any meaningful sense.) The study also originated by belief that the key to fraud control was deterrence rather than detection.
Sexual Sterilization Regulation Evaluation. Arriving at AMS I took over the Sexual Sterilization Evaluation. The project was mandated by Congress to review associated federal regulations. I first envisioned the project as centered around a set of surveys of affected groups. However, political considerations soon changed everything. The election of 1980 brought into HHS a new conservative administration. Richard Schweiker was appointed Secretary of HHS, the nominal liberal of the administration. Carolyne Davis was appointed head HFCA. Davis was a strong conservative ideologue. Davis quickly announced opposition to my project as part of a larger attack on Schweiker. In a memo she wrote that the project was a Carter era effort to impose burdensome regulations on the health care industry and involved family planning, to which she was deeply opposed. Schweiker was on verge of canceling the evaluation. In response I a ghost wrote a memo in Schweiker’s name saying that he was promoting the project as a review of possibly burdensome regulations and unnecessary federal intrusion into individual lives. If Davis was comfortable with just continuing Carter’s regulations without review, he was not. And so the evaluation continued.
Almost immediately a second political barrier emerged. Any survey conducted by the federal government was subject to review by OMB under the Paperwork Reduction Act. The Paperwork Reduction Act that no survey with more than nine respondents could be imposed on the citizenry without OMB approval. OMB, which was headed by David Stockman made the same ideological objections to evaluation as Davis. So again the project, and my job, appeared to be at an end. As I solution I proposed a large number of small surveys with nine respondents each of affected subgroups. We visited Indian Health Clinic across the country, traced the migrant farm worker trail and visited their clinics. We interviewed advocates for the mentally disabled, poor minorities, doctor groups, and so on. We conducted an exhaustive literature review and interviewed all leading experts on the subject.
Liberal advocacy groups strongly weighed in. The National Organization of Women (NOW) was conflicted. They recognized the advantages for women of a required 30 day waiting period between signing a consent form and the sterilization operation. The waiting period prevented unconsidered or coerced consent during birth and helped minimize subsequent regret. However the required waiting period might be seen as limiting a woman’s right to choose, a NOW slogan. As a result, NOW would only back the regulation if it also applied to men. The regulation made little sense for men; they did not give birth so could not be coerced then. Men experience much less post-sterilization regret. Nonetheless, resulting regulations contained a waiting period for men to satisfy the political agenda of NOW.
Ralph Nader’s health care group was also involved. They had written an expose on sterilization of the mentally retarded that showed multiple examples of abuse. The Nader report contributed to the regulations prohibiting such sterilizations without full informed consent of the patient. Many severely retarded individuals could not give such consent and so sterilization was effectively prohibited. From my research I concluded this solution was inadequate. I came to believe that such a ban harmed my retarded individuals. Many were fully capable to leading independent lives but not of being parents. Among this group contraceptive failure was common. Some severely retarded women would go into major panic every month when they had their period.
Looking a state-specific regulations, I concluded there existed alternatives that gave access to sterilization for the retarded while protecting their rights and interests. However, the Nader group was categorically opposed to anything but a full ban. I became quite disgusted with the Nader gang because they appeared to me totally ignorant of the on-the-ground realities, and were much more concerned with their own moral posturing (and fund raising) than the needs of the retarded. The Nader group prevailed and the ban was perpetuated.
Responding to the ban parents and caregivers of the mentally have formed underground networks of providers who will perform the operation outside any public scrutiny. The outcome is that the rights of the retarded are routinely abused rather than protected.
Once the study was complete, the Public Health Service sent it to the Centers for Disease Control for expert review. I was very nervous. Although we had done our best, without large-scale surveys, I could not produce definitive numbers on the health impact of the regulations. To my surprise, CDC loved it. All the interviews with impacted individuals gave the study a rich human texture that the epidemiologists found lacking in their statistics.
The study produced a report to Congress which really means a report to congressional committee staff. The time congressional staff was dominated by the majority Democrats and deeply skeptical of any Republican administration efforts to modify health care regulation. As a last political maneuver I leaked word to the relevant Democratic congressional staff that the study had been conducted by a Berkeley political lefty (me) and not to make any noise so we could slip our civil rights agenda past the oblivious Republicans. To the Republicans in HHS I conveyed the exact opposite message; we had thwarted the leftist agenda to control all health care but cloaked the report in bland technical jargon. The truth was neither. The needs of Americans for access to sexual sterilization services have nothing to do with the agendas to the left and right. Since the evaluation the federal regulations remain essentially unchanged.
Loan Standards at Treasury
AMS held a contract to provided statistical support to the Treasury Department on an as needed basis. Overtime I became the technical lead for this contract. One project was to develop a set of financial standards that would apply to all federal loan programs. The project originated at the Office of the President (OMB) in 1983. However, because OMB did not have contracting authority, Treasury took the lead and I directed all technical aspects of the project. (I have attached our statement of work and the project description. The actual work performed evolved over time. loan-program-standards )
At the time the federal government had over 150 loan. These programs were run under the authority a diverse set of departments and agencies such as HUD, Veteran Affairs, and the Departments of Agriculture, State, and Education. The total guaranteed debt exceeded $300 billion. The individuals running these programs often had little background in finance. Problems with growing loan default rates in several programs provoked concern at OMB and Treasury.
The programs were operated without any uniform standards for credit granting, risk analysis, write-off policies and loan sales. The projected aimed to develop such standards. The project was confidential and internal to the government.
Overseeing the project was a committee of assistant secretaries from the departments with the largest loan programs. I quickly learned that it was assistant secretaries that yielded the real day-to-day power of government. The political appointees created policies. The assistant secretaries translated these policies into action, or inaction, as they saw fit. Meetings were held in the “cash room” of main Treasury.
During the project I learned how liquid and shapeless was bureaucratic power. Administrative support for the committee was provided by Susan Brita. Susan was a mother who had recently entered the labor force once here children entered college. She held a low administrative support job ranking. Her office job was to schedule meetings, let members know when and where, arrange pastries and coffee, take notes and so on. And take notes she did. Further, she would call committee members in advance of each meeting and remind them of commitments they had made in prior meetings and ask for updates on their progress. Now this is not behavior one expects from a low-level support person. Soon, all committee members came to assume Susan was the real representative of the administration. Once an Assistant Secretary, during one of Susan’s update calls, reported that he had been unable to find funding to support project request. Susan replied that perhaps she could be of assistance. She would arrange a meeting with the Department’s desk at OMB where they might be able to find the necessary funds. A Department’s OMB desk is a very frightening place for government bureaucrats. It is the point were their funding is controlled. All Departments limit access to their OMB desks very strictly. Not surprisingly, the committee discovered he could find the necessary funding without Susan’s assistance.
Another job Susan held was scheduling of department limousines. Again the intent was purely functional and again Susan found access to power. The Treasury is right next door to the White House. There are underground tunnels between the two buildings. However, some senior Treasury official, for some unknown reason, where unable to get between the two buildings without a limousine. Limousine access was a prime topic of concern at Treasury’s highest levels. At the time Treasury was developing a new information system valuable at over $500 million. The internal executive memo system contained almost 30 memos relating to this system. There were over 80 relating to limousines. Susan explained those officials who were helpful for Susan’s projects somehow always had access to the limousines. Those not so helpful, for purely objective and verifiable reasons, often had problems getting limo.
The Director of Administration for Treasury was John Garmat. Garmat was a classic wheeler-and-dealer. He wanted me to guarantee to him that certain project audits and demonstrations would came out the way he wanted. In particular, he desired that a demonstration of D&B loan analysis products should prove successful in managing small business administration loans. I would not guarantee that outcome only a fair test (although I had great hopes for success.) D&B was very aggressive suggesting to me my actions could improve or end my career. I went to Susan with this problem. A week later Garmat resigned from Treasury. I found out later Garmat had told the President of AMS that I was naive as to the ways of power and should be fired. And I would have been fired if Garmat had not gone first.
A last story about Susan. In 1984 Terence Golden was named Assistant Secretary of Treasury for Administration. Susan was assigned to showing him around the building. Golden had no prior experience in government. Susan claimed all he knowe was that he voted every four years. The notion was she would show him where the executive washroom was, the security codes, etc. As she showed him around the building she explained how it all worked. Very soon Susan was his chief of staff.
Finally, about me. When I got this assignment I was a quite competent statistician. However, I knew nothing about finance nor running loan programs. My first step was to buy some expensive pinstripe suits. I sat up nights reading finance texts. I soon discovered the power of a title. Everyone on the project assumed Treasury had selected me as a leading financial expert. Being Dr. Finnegan and speaking the language of statistical risk analysis was enough to get me through without ever being questioned. I listened to everyone and learned fast.
Student Loan Fraud Control
Food Stamp Audit Control System.
The Temporary Emergency Food Assistance Program (TEFAP)
In 1987 Quality Planning saw its downfall in Washington consulting with our evaluation of the Temporary Emergency Food Assistance Program (TEFAP) . It was a large, year-long project. TEFAP was a program to distribute food purchased as part the Department of Agriculture commodity price supports. The surplus food was distributed through various private social assistance programs such food banks and senior centers. Large, five-pound blocks of cheddar cheese was then the primary commodity TEFAP was then distributing.
The study was designed to address three primary questions, who was getting TEFAP benefits, their income and participation in Food Stamps, and the nutritional value of the distributed food.
During the study our teams visited food distribution centers throughout the country and interviewed individual receipts including a complete nutritional and income analysis of recipient households.
Before, study results were released, USDA announced they would be eliminating the program because 1) Benefits were not well targeted to at need families, 2) Those families at need were already getting food stamps, and 3) TEFAP did not provide meaningful nutritional support to recipient families. (We indirectly learned that the decision to eliminate the program was made under pressure from Kraft Foods who believe that TEFAP was reducing the demand for their cheese products. More about this below.)
The conclusions of USDA were reasonable on their face: the program was not well targeted and had few controls to assure only the poor participated; the Food Stamp program did overlap with TEFAP; and the package of foods distributed by TEFAP was constructed on the basis of what USDA had purchased for price supports and not to match the needs of recipients.
To our surprise, the evaluation found TEFAP was, after all, a very effective program. We found that almost all recipients were quite poor (who else would get up early in the morning, take a bus to a food bank in a bad part of town, and often wait hours of a box of food not of their own choosing?). We found that although almost all recipients were eligible for Food Stamps, only about 25 percent of these were actually getting Food Stamps. Finally, our biggest surprise was that TEFAP appeared well-tailored to the nutritional needs of recipients. The largest single group of recipients were elderly women who needed the calcium and protein provided by the cheese.
When we submitted our report to USDA Food and Nutrition Service, they were not pleased. Various changes were demanded—but never in writing. I responded that as Congress had mandated the study, I could not ethically make most of the requested changes. One area USDA was particularly insistent was our finding of nutritional impact. While refusing to change the findings based what I saw as political pressure, I did agree a compromise suggest by my team of nutritionist. We would make changes if a leading nutritionist on the staff of USDA believed there were errors in our analysis.
The the review was under way USDA rewrote the executive summary of the report, the only part of these reports anyone ever reads, to downplay, obscure, and hide our key findings. We never saw the agreed to technical review. When I pressured USDA for a company I was assured that the internal nutritionist had found numerous errors and the results had to be adjusted.
Not long after the report was publicly released I got a call from FRAC, the Food Action Resource Council. FRAC was following up on rumors we had hidden our findings to fit the political agenda of the Bush administration. I assured them this was not the case. The only findings changed were in the nutritional analysis and this was based on expert peer review. We were experts in our field but not infallible and therefore peer review was justified. FRAC responded this is not the story they got from inside USDA. I therefore followed with my USAD contacts. The nutritionist who supposedly had conducted the review had never heard of the study and definitely had not review it. The set a copy of the study to the chairs of the nutrition department of Berkeley and Harvard and ask what they thought. In both cases they praised the methodology as far better than the usual USDA work. In particular we were praised for evaluating what recipients actually ate, taking account of waste and spoilage rather then simply what food was available.
Again I followed up with USDA and told them what I found. The new story was more telling. Why USDA was particularly upset with our methodology was it called into question of the nutritional analysis USDA used to set the level of Food Stamp benefits. USDA is required by law to set benefits at a level to meet the nutritional needs of recipients. Their method of doing this was to determine the lowest price a trained nutritionist could buy foods necessary to meet minimal nutritional standards. This method did not look at what foods recipients bought and fact and how those foods were consumed. If USDA had employed a method similar to ours, based on reality and not ideal theory, Food Stamp benefits would have had to be greatly increased. This definitely was not acceptable.
I reported my new discoveries to FRAC. Not long after I got a call from CBS News. I told them the above story. They sent out a film crew to interview me. and included the story as one instance of how the Bush Administration was suppressing science. The story was national news.
The shit hit the fan. QPC’s government contracts were audited. Our retirement plan was audited by the Department of Labor. Problems were found with every bill we submitted to the Federal Government. Payments were delayed or denied. Word was passed to us we would never again win a government contract. We fought all negative findings from the audits and won every dispute. However, the damage was done. Almost all our income was from Federal Contracts and QPC came within an inch of going bankrupt.
The upside of this story, was that Quality Planning changed its business model to concentrate on fraud in auto insurance. While the loss of federal contracts cost me millions, the move to insurance work made me tens of millions.
(A closing note on this story about Kraft Cheese: This whole dispute appeared to have been set off by Kraft Foods’ belief that TEFAP was suppressing the demand for Kraft products. While this was not part of our study, in the course of our interviews we did learn a lot on the topic. The reason USDA had so much cheddar cheese to give away was they had converted massive milk purchases into cheese rather than just poor the milk down the drain which would have been a very bad public image. To minimize storage costs, USDA made cheese that was almost pure cream. This cheese was then stored in caves under Kansas. Thus, the cheese distributed was very high cream and well-aged. Program recipients loved it. A particular advantage was that you could cut off a slice, put it in a frying pan for a couple minutes and have a great cheese sauce. Many interviewees reported TEFAP made them come to love cheeses. Between TEFAP distributions, these recipients would go to their local grocery stores to attempt to buy similar cheese. The problem was the quality of Kraft and other cheese products available. They were not nearly as good.)
The American’s with Disabilities Act
In 1988 both presidential candidates, Vice President Bush and Governor Dukakis, endorsed broad civil rights protections for people with disabilities. When George Bush was elected president, the possibility opened for passage of the Americans with Disabilities Acts (ADA). Bush had appointed Evan Kemp, who had been his roommate at Yale and was disabled, as head of the Office for Civil Rights. Bob Dole was majority leader of the Senate, and partially disabled during the Korean War, signaled his provisional support for the act.
Pat Wright of DREDF in Berkeley, an organization with which I had deep ties, went to Washington to coordinate the lobbying effort on behalf of the disabled community. While the bill had substantial support issues relating to costs threatened the prospects for passage. The New York Times ran an editorial questioning costs. Bod Dole made his support contingent on reasonable costs.
As a result of this pressure, I was contacted by Edward Kennedy’s staff to conduct a cos-benefit analysis. Because of the short time-frame, it was judged better to hire me as part of the Senate staff than to let a contract for the work. My work with the Bush and Reagan administrations made me acceptable to the Republican side and may work with DREDF and participation in the disability rights movement recommended me to the Democrats. It was a very different Washington then when issues could be rationally discussed, evidence counted, and political differences did not outweigh the common interest.
I reached out to various experts and lobbyist to obtain the necessary information. Ron Mace provided needed architectural information. For data on telephone relay system costs I contacted ATT. I then ran their cost data by various advocacy groups and conducted comparative analysis of costs of existing relay systems. ATT had made high cost estimates in hope of being awarded a profitable contract. When confronted with data that called these high costs into question, they modified their estimates to a more reasonable level.
Similarly bus systems, particularly Greyhound Bus made very high estimates of the costs of making their systems wheelchair accessible. On the other side, providers of wheelchair bus lifts, made much lower cost estimates in hope of greatly expanding their bus. A back-and-forth dialog I created between the bus systems and providers of wheelchair bus lifts resulting in substantial convergence and a more realistic estimate.
When I submitted my preliminary report, advocates for ADA where concerned my cost estimates were too low to be creditable. Going over the details, I could not find any strong justifications for substantially higher costs but did add notes emphasizing areas of uncertainty.
Looking back, there is no doubt I was correct overall, the ADA did not result in the ruinous costs opponents projected. I believe I mostly correctly estimated ADA costs of public transportation access, architectural access and relay services. I overestimated economic gains of disabled employment. I did not anticipate legal costs for small business access.
Now both Bush and Dole claim the ADA as the greatest accomplishment of their long political careers.
By Daniel • Washington, Work