Lies and Statistics
February 20, 2002
by Jack Rabbit

"There are three kinds of lies: lies, damned lies and statistics." - Attributed by Mark Twain to Benjamin Disraeli

Welfare reform will soon be up for review. There will be congressional hearing to determine how well or how poorly the present system has worked. Experts representing either side will come before Congress and give their arguments supported by pretty charts and graphs, previous studies and fancy academic language used to mask the fact that they represent a partisan point of view and don't really care about America's poor.

The Personal Responsibility and Work Opportunity Reconciliation Act was a product of Newt Gingrich's Contract with America; it was passed in 1996 and signed into law by President Clinton, who heralded the act as "the end of welfare as we know it." It replaced the old Aid to Families with Dependent Children (AFDC) program by a new Temporary Assistance to Needy Families (TANF) program. ADFC was a very bad example of a welfare program and was for thirty years an irresistible target for those who genuinely had the best interests of the poor at heart and yuppie fascists, who do not.

In many ways, AFDC was every bit as bad as its critics on the Right said. It was an inflexible system that discouraged work (earning income would result in having benefits reduced) and tore families apart (a single mother was able to draw more benefits by remaining single and married people separated to get benefits). However, TANF is really little more than AFDC with time limits and work requirements. The yuppie fascist line is that the poor are dependent on welfare and if they were threatened with being cut off after a time, they will get off the dole and better themselves.

According to an article in the Washington Post (see http://www.washingtonpost.com/wp-dyn/articles/A35857-2002Feb6.html), Mr. Bush will propose that states operate welfare programs with $16.5 billion, the same level of federal finding they received six years ago, when the welfare reform act was passed into law. Republicans will also point to plummeting welfare rolls as a reason for maintaining the funding at such levels. Democrats will assert that now that the nation is in a recession, funding should be increased to meet greater need.

The hearings will begin. We will see a parade of witnesses. Democrats will call witnesses who will assert the program could do more to fight poverty if the government would pour more money into programs that do more to prepare the poor for the workplace. Congressman Benjamin Cardin of Maryland, the main sponsor of a Democratic bill, believes that at least an additional $25 million will be needed above what the administration proposes to make welfare reform work.

Republicans will call a parade of witnesses to say that welfare reform has worked wonderfully in weaning the poor from federal dependency on welfare and does what is needed, but more must be done to get the poor to take personal responsibility for their lives. Health and Human Service Secretary Tommy Thompson says that he and Mr. Bush want to "help those families that have left welfare to climb the job ladder and become more secure in the workforce." And Mr. Bush has said he want $100 million for experimental programs to encourage welfare mothers to get married.

Very few of the witnesses will be beneficiaries of aid programs. It's unfortunate that we won't see many poor people telling of the advantages and disadvantages of one form of government as opposed to another or how they have to stretch the rules of any system of aid in order to pay the rent and feed the kids. We will not hear - at least not from a low-wage earning mother - of how the cost of childcare must be computed into necessary monthly expenses and of how an aid program to help her pay for childcare might help her keep her job. Nor will we hear from the poor themselves about how low-wage jobs provide no medical benefits and how sometimes a child who might benefit from medical attention gets none.

Instead, most of the witnesses will be "experts." We will see well-educated men and women in business suits boldly raise their right hands, promise to tell the truth and sit down, look directly at the committee chairman with the sharp, confident glaze of one who has made his way in the world.

However, instead of telling the truth, these experts will present a display of spin worthy of Ari Fleischer, buttressed by charts, graphs and fancy terms like "regression analysis" and "bivariate correlations."

These experts present themselves as social scientists. They draw their conclusions from data set to a variety of statistical analysis. Are these experts really scientists? Or are they, like attorneys in a trial court, in the business of spinning the facts on behalf of a particular client in order to sell a particular point of view that might or might not resemble the truth?

One of the Republican-friendly witnesses will no doubt be Robert Rector, an author and a senior research fellow at the Heritage Foundation (see http://www.heritage.org/staff/rector.htm). Rector recently said that, in light of the success of welfare reform and the reduced caseload, the federal government should reduce the states' block grants by 10 percent and spend $300 million, three times the amount suggested by the administration, on programs to promote marriage among the poor.

Mr. Rector is no stranger to giving testimony before Congressional committees about welfare. In September 1999, he testified before a subcommittee of the House Education and Workforce Committee on the success of TANF (see http://www.heritage.org/library/testimony/test090999.html). In his testimony, Rector argued for harsh sanctions against those beneficiaries who violate welfare rules as a way of ending welfare dependency. Rector argues that under TANF, child poverty has been reduced "when the Earned Income Tax Credit, Food Stamps and other means-tested benefits are counted as income."

Rector explicitly denies that the economic boom had anything to do with the considerable decline in welfare since the passage of the 1996 welfare reform. Rather, he goes through statistical acrobatics to show that harsh sanctions are the cause of the reduction in caseload, that in states that have harsher sanctions, the caseload is reduced in spite of the fact that in those states there were slightly higher unemployment rates.

Rector divides the various states into categories according to the severity of sanctions dealt to welfare violators vis a vis the beneficiary's compliance with work requirements. He put the states in four categories: (1) initial full-check sanction, in which the beneficiary has his or her TANF aid cut off at the first violation; (2) delayed full-check sanction, in which the beneficiary is subjected to increasingly more severe sanctions upon successive violations; (3) moderate sanction, in which the state reduces the aid by more than a third, but not all aid; and (4) weak sanction. Rector also analyzed the data by the timing of the state's work requirement, in order to draw a correlation between those states that require beneficiaries to engage in work sooner than the 24-months required by federal law.

One should not be surprised that Rector shows that states with early work requirements and harsher sanctions have a greater reduction in caseload. Of course, since increased dependency on welfare is harmful to the poor, because he can show that having a job is better than being on welfare, that's a good thing. After all, the beneficiaries loose their benefits and find real jobs. That just makes sense - at least it does to Rector.

However, Rector does not cite any data to show that those beneficiaries who lose their benefits in the states with earlier work requirements and harsher sanctions move to jobs, or that the jobs to which they move are of greater benefit to them than government aid. In fact, something particularly noteworthy in Rector's testimony is his use of the word "caseload." He uses it, not counting the bold-type subject division added by the editors, 50 times. Nowhere in this testimony does Rector speak of the beneficiary of aid in the active voice. The argument is always in terms of what the state must do to reduce caseload. Nowhere does Rector even use a term like "beneficiary" or "welfare mother" or anything else that might imply that the reduced caseload is made up of individual human beings, who are often poor women with children. Whether it was Rector's intention or not, the word "caseload" dehumanizes the subject of his testimony - America's poor people - and sweeps them under a verbal rug.

Since Rector doesn't show that the "reduced caseload" in states with earlier work requirements and harsher sanctions found jobs, what is his data really measuring? Only that if a state demands more of a beneficiary for aid and cuts off aid to those who do not meet the state's demands to the satisfaction of the social worker, caseload will be reduced. A real surprise, that. Indeed, if the caseloads were reduced most in states with harsher sanctions and higher unemployment, it shouldn't take an Aristotle to tell Rector that what is happening is that people are being thrown off welfare and not necessarily finding work.

Perhaps we need to remind Rector that a great problem with the old AFDC system was its inflexibility. Getting off welfare and going to work was often an all-or-nothing proposition for the recipient. If the beneficiary was working, then the beneficiary apparently did not need to draw AFDC. Thus, getting low-wage job with no fringe benefits like health insurance probably leaves the AFDC recipient worse off than before. After all, a full-time worker earning minimum wage with no benefits is earning only 60 percent of the poverty line.

Which brings up something else missing from Rector's study: how many of these former aid beneficiaries are better off in low paying jobs than they were receiving aid? Rector doesn't tell us how childcare expenses associated with employment, and not covered by a federal program, affect the income of those moving from welfare to work. Nor does Rector take into account that many welfare recipients bent the rules and found ways of receiving outside income off the books. The fact is that government aid was insufficient to cover living expenses and welfare cheating was often a necessity. In short, Rector takes no account of the increased expenses a former aid recipient incurs moving from welfare to work.

Of course, Rector didn't say in his 1999 testimony that Food Stamps are another form of welfare dependency. However, in June 2001, he did go before a sub-committee of the House Agriculture Committee to testify: "The current Food Stamp program is a virtual twin of the failed Aid to Families with Dependent Children (AFDC) program. Like AFDC, Food Stamps are anti-work, anti-family and anti-marriage" (see http://www.heritage.org/library/testimony/test062701.html). The Food Stamp program was more broadly based than AFDC. Although long associated in the minds of many with welfare, the Food Stamp program has been a traditional benefit to the working poor as well. It is a well-known fact that even military families receive Food Stamps (see http://www.cnn.com/ALLPOLITICS/time/2000/05/01/food.stamp.html).

While the old AFDC program was often derided as a handout to people who refuse to work (ignoring the fact that AFDC imposed penalties on benefits for income earned), the Food Stamp program is a supplement to the income of low-wage families.

However, this doesn't stop Mr. Rector from characterizing Food Stamps in terms of AFDC. In his 2001 testimony, Rector called the 1996 welfare reform legislation "a tremendous success leading to the dramatic drops in dependence in child poverty, increases in employment and slow down in the growth of out-of-wedlock child bearing." Rector further says:

"The current Food Stamp program is a fossil embodying all the errors of the old War on Poverty. It provides one way handouts, rewards non-work and idleness, fosters long-term dependence, rewards and promotes out of wedlock childbearing. As such, the Food Stamp program actively harms children and increases poverty in the nation."

Rector recommends for fixing the Food Stamp program essentially the same principles that distinguish TANF from AFDC: work requirements. Rector would deny aid to able-bodied, non-elderly adults who either do not have a job or refuse to perform community service work. Rector cites statistics to show that dependency of Food Stamps is often long term. Nearly 41% of all Food Stamp recipients have been receiving Food Stamps for over ten years, and another 29% receive aid for between five and ten years.

Rector also argues that Food Stamps, like AFDC, "erodes the work ethic and the growth in dependence, in turn, has profound negative effects on the welfare of children." Rector cites studies to show that poor boys are less productive adults. He cites statistics to show that the determining factor in the boy's eventual earning capacity is welfare received during childhood, "holding constant race, parental education, family structure and a range of other social variables." Rector even tries to get around the obvious objection, that it is not welfare but poverty that is the determining factor.

In citing the study supporting his contention that growing up in an environment of welfare is the cause of reduced earning capability later in life, Rector omits to explain a great deal about the study. In the footnotes, we find that the study was published in 1992; the welfare of which Rector speaks, then, is AFDC. However, no one is arguing with Rector that AFDC was a bad program. The question is: has TANF made significant improvements in the lives of the poor?

After only six years, it really isn't possible to determine the long-term implications of being raised on TANF as opposed to AFDC or with parents who try to make ends meet on low-wage income unsupported by public aid. Besides, Rector is arguing against Food Stamps in his testimony, not AFDC. Rector simply began his testimony asserting that Food Stamps were like AFDC. He then follows through with this assertion, without proving it in any convincing way, to equate the evils of AFDC with those of Food Stamps. Moreover, 1992 was in a recession. Could this have had an effect on the earnings of the subjects in the study? Rector doesn't suggest that it does?

Finally, Rector argues that Food Stamps and other means-tested aid programs penalize marriage. This encourages illegitimacy, which he claims causes poverty. "Children born out-of-wedlock to never married women are poor fifty percent of the time," Rector points out. "By contrast, children born within a marriage that remains in tact are poor 7 percent of the time."

There is no mention of children of divorce. Children can be raised in a single parent family without being illegitimate. Of course, Rector tells of the evil consequences of being born to an unwed mother, citing yet another study that shows that young black men raised in single-parent families were twice as likely to engage in criminal activities as those growing up in intact families.

It is noteworthy that in his 2001 testimony about Food Stamps, Mr. Rector has a great deal to say about the effects of aid programs on children and how aid programs discourage marriage and promote out-of-wedlock childbirth. One might expect to hear much of this from Rector and others such experts as they testify in the coming weeks. It does, after all, support Mr. Bush's plan to promote marriage as a poverty-fighting program.

It might be noted that in all of Mr. Rector's citing of statistics, data and graphs to illustrate his point, he seldom reveals anything else about the studies which he cites. Were these randomly chosen cases? What was the sample size in these studies? Anyone who has studied statistics knows that a larger sample will better reflect the universe that the data purports to represent. Finally, who performed that study? Was it an independent group of social scientists? Or was it a group who made an assertion about welfare being evil and went looking for facts to fit the pre-ordained conclusion? One would have to spend a great deal of time researching that; that's time most us don't have.

These questions might not arise if the Heritage Foundation were not so predictable. The Heritage Foundation is a conservative research and educational institute - in other words, a think tank with a partisan mission. Their mission statement makes clear that they are formulating and promoting "conservative public policies based on the principles of free enterprise, limited government, individual freedom, traditional American values, and a strong national defense" (see http://www.heritage.org/whoweare/).

Given this, should we even expect intellectual objectivity from the Heritage Foundation? Or should we just expect them to formulate talking points for right-wing politicians?

Overall, Mr. Rector makes many specious arguments. In his 1999 testimony, he argued that Food Stamps counted as income in offsetting child poverty in an attempt to invalidate arguments that welfare reform had hurt the poor; however, in 2001, without stating that this was no longer the case, Rector attacked Food Stamps as part of a network of welfare that he blames for continued poverty. Is Rector saying that if aid offsets poverty, then it contributes to poverty? That seems to be the right-wing argument. In fact, Rector seems to be setting up the idea that all means-tested aid to the poor saps their ambition. His argument about welfare causing illegitimacy and in turn causing crime is almost laughable. While he has some reason to suggest that the old welfare rules discouraged beneficiaries from getting married, to turn around to say that illegitimacy causes crime is dreadful. How many of those illegitimate criminals were born to comfortably middle-class women with a future? Would Rector seriously argue that the child of an unmarried professional who chose to remain single is just as likely to turn to a life of crime as the child of a poor, inner-city teenager?

Mr. Rector avoids stating the obvious truth: that AFDC was at worst a bad solution to poverty and not a cause of poverty itself. Poverty causes crime, not welfare or illegitimacy. His little tricks of correlating welfare and illegitimacy with crime only work insofar as both can be correlated with poverty. The poor have been stealing bread and trafficking on the black market in order to survive since the beginning of time.

However, if he said that, that would fly in the face of the Heritage Foundation's mission of formulating and promoting conservative policies. Therefore, Rector must state that his charts and graphs and data analysis "prove" that reducing government aid to the poor actually makes their lives better. He cannot say that simply adjusting rules and requirements to make aid programs more flexible is a better idea. That is not the Republican line.

Rector's 1999 testimony shows that welfare reform as enacted in 1996 is about "reducing caseload" and not fighting poverty. Using the term "caseload" covers the fact that behind each individual case is a human being, often a single woman with children. The thesis he promotes is that poor are dependent on walfare and removing welfare will make them productive and happy people; that welfare makes the poor lazy and stupid. Yet the very struggle in which each individual poor person engages, even when this includes criminal acts ranging from armed robbery to violating welfare rules, shows that poor people are not lazy or stupid. They are very intelligent people. They can manipulate their environment as well as most congressmen and corporate executives manipulate theirs. Of course, such acts may be deemed dishonest. Yet they are no worse than the acts of the corporate fathers at Enron.

It is Robert Rector and many people like him who will testify before Congress about welfare reform. Some will support Rector's position and some will oppose it. They will be heard, and the poor people who receive these benefits will have to settle for people such as they - educated, self-sufficient and well paid - to speak on their behalf. Will any of the poor testify themselves? All of those experts who do speak will talk of their concern for the poor and use fancy language and attractive charts to mask the fact that they are there because they are supporting the nominally opposing positions of upper- and middle-class politicians.

In the end, the side with the most votes will win. Since the Republicans control the House and can count on a few blue-dog Democrats to join them on this issue, Mr. Bush and Mr. Thompson can expect to see most of their proposals emerge from the congressional floor fights as law. Those who must deal with poverty as a direct experience don't need Robert Rector's fancy charts, regression analysis or his fallacious arguments to tell them of its harmful effects on poor adults and even more on poor children. Bush will win, and they will lose; but perhaps they don't stand a chance in the first place.