# Studies of Welfare Populations: Data Collection and Research Issues. Predictions of Outcomes for High-Barrier Cases

This section uses the coefficient estimates from the models predicting the probability of leaving welfare, the probability of employment after leaving, and earnings after leaving to predict each of these outcomes for different definitions of "high-barrier" cases. 17 Seven definitions of high-barrier cases are examined. The first is the same definition used earlier--cases that had no high school diploma, received welfare for at least 48 months in the preexit period, fewer than four quarters of earnings in the preexit period, and had at least one child under the age of 5. The rest of the definitions build this basic definition. They are:

Definition 2 = Definition 1 + the case head is on SSI.
Definition 3 = Definition 1 + the case includes a child on SSI.
Definition 4 = Definition 1 + the case lives in Milwaukee County.
Definition 5 = Definition 1 + the case head is black.
Definition 6 = Definition 1 + the case head is black and lives in Milwaukee County.
Definition 7 = Definition 1 + the case head was sanctioned from AFDC.

For each outcome, the coefficients from the model that uses the long-termer, short-termer and cycler distinction are used. Table 13-21 shows the mean predicted probability of the three outcomes computed for cases that qualify as high-barrier cases under these definitions. The first column shows the mean predicted outcomes for all cases in the sample as a reference.

 All Cases High- Barrier Definition1(a) High- Barrier Definition2 (b) High- Barrier Definition3c High- Barrier Definition4d High- Barrier Definition5e High- Barrier Definition6f High- Barrier Definition7g Number of cases 34,726 1,410 173 304 1,051 877 792 138 Percent of total sample 100.0 4.1 0.5 0.9 3.0 2.5 2.3 0.4 Number of leavers 17,294 344 26 65 192 150 125 45 Mean predicted probability of leaving welfare 48.7 (19.6) 23.9 (11.9) 7.5 (4.9) 21.5 (11.6) 19.0 (7.0) 18.9 (7.8) 17.5 (6.2) 28.7 (11.4) Mean predicted probability of employment after leaving welfare 88.3 (11.2) 75.7 (15.0) 29.5 (9.5) 72.5 (16.5) 75.4 (15.8) 74.7 (16.2) 74.0 (17.0) 72.3 (7.1) Mean predicted quarterly earnings after leaving welfare 1,929.9 (645.3) 1,224.1 (407.9) 293.5 (144.7) 1,046.0 (381.6) 1,329.0 (407.6) 1,246.2 (408.0) 1,276.7 (408.7) 953.5 (230.0) Note: Cases with missing data for explanatory variables were eliminated. Predictions are based on the actual values of explanatory variables for each case. Standard deviations reported in parentheses.a Definition 1 = No high school diploma; received AFDC for at least 4 years between 7/89 and 7/95; had fewer than four quarters with earnings between 1/89 and 7/95; had at least one child under the age of 5. b Definition 2 = Definition 1 + case head is on SSI. c Definition 3 = Definition 1 + case includes a child on SSI. d Definition 4 = Definition 1 + case lives in Milwaukee County. e Definition 5 = Definition 1 + case head is black. f Definition 6 = Definition 1 + case head is black and lives in Milwaukee County. g Definition 7 = Definition 1 + case head was sanctioned from AFDC.

Probability of Leaving Welfare for High-Barrier Cases

For the entire sample, the mean predicted probability of leaving welfare is nearly 49 percent. This is close to the 48 percent of the caseload that actually left welfare during the time period. Under different definitions of high-barrier cases, the probability of leaving welfare varies substantially. Under the basic high-barrier definition (Definition 1), the probability of leaving is 24 percent or about half the probability of leaving for the entire sample. Across different definitions of high-barrier cases, by far, cases that receive SSI have the lowest probability of leaving welfare. The mean predicted probability of leaving welfare for this group (Definition 2) is only 7.5 percent. For those high-barrier cases that include a child who receives SSI (Definition 3), the probability of leaving welfare is not as low as cases where the mother receives SSI. The mean predicted probability of leaving welfare for this group is 21.5 percent. For those who are high-barrier cases and who live in Milwaukee County the mean predicted probability of leaving welfare is 19 percent. This is nearly identical to the mean predicted probability of leaving for high-barrier cases that are also black (Definition 5). High-barrier cases that are black and live in Milwaukee County (Definition 6) have a slightly lower mean probability of leaving welfare, 17.5 percent.

These results suggest that high-barrier cases are much less likely to leave AFDC than those who do not face these barriers. This is especially true for those who receive SSI payments. High-barrier cases who are black and live in Milwaukee County also have a lower probability of leaving welfare than other high-barrier cases. Those high-barrier cases with a child who receives SSI payments are only slightly less likely to leave welfare than all high-barrier cases.

Probability of Employment in the First Year After Leaving

The next row shows the mean predicted probability of ever being employed in the first four quarters after leaving. These predictions are based on the coefficient estimates in Model 1 in Table 13-19, and are computed only for those who leave welfare. First, the overall mean predicted probability of employment after leaving is 88.3 percent. For the basic definition of high-barrier cases, the mean probability of employment is 75.7 percent, which is about a 14-percent difference from the overall mean probability. This is still a sizable difference, but not nearly as big as the difference in the mean predicted probabilities of leaving welfare for high-barrier and nonhigh-barrier cases. Furthermore, for nearly every additional definition of high-barrier cases, the mean probabilities of employment are approximately 75 percent. There are some exceptions. First, those high-barrier cases that receive SSI (Definition 2) have quite different mean predicted probabilities of employment than the overall sample and than the basic high-barrier definition. The mean predicted probability of employment after leaving for this group is only 29.5 percent. 18 Second, those with a child on SSI (Definition 3) and those who were sanctioned from AFDC (Definition 7) have slightly lower mean predicted probabilities of leaving (72.5 percent for Definition 3 and 72.3 percent for Definition 7). These results indicate that even sanctioned high-barrier cases and high-barrier cases with SSI-eligible children have fairly high employment rates after leaving welfare and do not appear to have trouble finding employment after leaving welfare.

Mean Predicted Quarterly Earnings After Leaving Welfare

The last row in Table 13-21 shows mean predicted quarterly earnings for leavers under the different definitions of high-barrier cases. These means are based on Tobit coefficient estimates from Model 1 of Table 13-20.

The mean predicted quarterly earnings of all leavers (column 1) in the first year after exit are \$1,930. The mean quarterly earnings of high-barrier cases (Definition 1) are \$1,224, which translates into a nearly 37-percent difference. Different high-barrier cases do better than this, however. The mean predicted earnings of those from Milwaukee County (Definition 4) are \$1,329, higher than mean predicted earnings of the basic high-barrier cases. This result is probably a result of wage differences between Milwaukee and other areas of the state. High-barrier cases who are black (Definition 5) also have higher earnings (\$1,246) than other high-barrier cases, although their means are not as high as the mean earnings for high-barrier cases from Milwaukee. Accordingly, those who are black and live in Milwaukee (Definition 6) have predicted earnings that fall between the predicted earnings of those from Milwaukee County (Definition 4) and those who are black (Definition 5). Their mean predicted earnings are \$1,277.

The predicted earnings of those with other barriers are not as high, however. Again, those high-barrier cases that receive SSI (Definition 2) are the worst off. Their mean predicted earnings are just \$293.5 per quarter. Again, only 26 observations fall into this category. High-barrier cases that have a child who is eligible for SSI also have low mean earnings, at \$1046. Finally, sanctioned high-barrier cases have low mean earnings, too, at \$953.5. Their mean is less than half of that for the entire sample of leavers and 22 percent lower than the basic high-barrier cases. So although the employment rates of sanctioned cases were not that different than other high-barrier cases, there are substantial earnings differences between sanctioned high-barrier cases and other high-barrier cases, and between sanctioned leavers and all other leavers.

Table 13-21 illustrates that it is likely that certain high-barrier cases will have a difficult time making it on their own. High-barrier cases in general are much less likely to leave welfare than other cases. Although their employment rates are not vastly different from all other leavers, their earnings are substantially different. High-barrier cases that are eligible for SSI are likely to have even greater problems making it on their own, according to these predictions. For other types of high-barrier cases, employment may not be a significant problem for them; however, earnings do seem to be a problem.

Results found here should supplement similar simulations conducted in Cancian et al. (2000b), where much wider differences in predicted outcomes between high-barrier cases and low-barrier cases were found. The Cancian et al. definitions of high-barrier cases are more restrictive than definitions used here.

#### View full report

"01.pdf" (pdf, 472.92Kb)

#### View full report

"02.pdf" (pdf, 395.41Kb)

#### View full report

"03.pdf" (pdf, 379.04Kb)

#### View full report

"04.pdf" (pdf, 381.73Kb)

#### View full report

"05.pdf" (pdf, 393.7Kb)

#### View full report

"06.pdf" (pdf, 415.3Kb)

#### View full report

"07.pdf" (pdf, 375.49Kb)

#### View full report

"08.pdf" (pdf, 475.21Kb)

#### View full report

"09.pdf" (pdf, 425.17Kb)

#### View full report

"10.pdf" (pdf, 424.33Kb)

#### View full report

"11.pdf" (pdf, 392.39Kb)

#### View full report

"12.pdf" (pdf, 386.39Kb)