In 1995, Colorado passed legislation allowing the State Division of Child Support Enforcement (CSE) to suspend driver's licenses through administrative process as a tool for enforcing child support payments among cases meeting certain legal criteria. In 1997, the law was amended to allow for the suspension of Commercial Driver's Licenses. In 1998, the Center for Policy Research (CPR) conducted an initial evaluation of the impact of driver's license suspension on child support payments. In early 2000, CPR examined child support payment patterns up to 24 months following the notice of suspension. This report summarizes the findings reached in the 1998 and 2000 evaluations and explores in depth one aspect of the longer-term effects of driver's license suspension. Specifically, this report considers the following research questions:

Answering these questions will help to document the true long-term effects of driver's license suspension as a tool of child support enforcement. In addition, answering these questions is critical to understanding whether the substantial effort that goes into monitoring cases and generating subsequent notifications and suspensions is cost-effective.

Child support enforcement has a complex system of tools to encourage voluntary compliance, eliminate opportunities for non-compliance (through measures such as routine wage withholding), and, as a last resort, create negative consequences for non-compliance. Suspension of a driver's license falls into the latter category. Child support agency administrators recognize that suspending an obligor's driving license may make it difficult for the individual to work, thus impacting earnings and ability to pay support. To counteract this possibility, obligors are given numerous opportunities to prevent the suspension. An obligor will not be selected for license suspension to occur unless:

Colorado, like the rest of the nation, uses driver's license suspension to push the most resistant obligors into accepting financial responsibility for a child. Evaluation results suggest that it can be an effective tool. For example, a press release from the Department of Health and Human Services credits the threat of license suspension with raising nearly $35 million in nine states.^[34] CPR's evaluation of driver's license suspension in Colorado also finds that the tool produces increased child support payments, although its impact is limited by the large number of obligors who qualify for suspension but are already without a valid license. Highlights from the initial Colorado evaluations are outlined below.

CPR's original evaluation of the effects of driver's license suspension compared three randomly generated groups of noncustodial parents. The experimental group consisted of individuals who were reported to the Department of Motor Vehicles for license suspension and were also reported to credit bureaus. The comparison groups included cases that qualified for driver's license suspension, but were not reported to the DMV. One comparison group was reported to credit bureaus, the other was not. In summary, the three groups were:

The key results of CPR's evaluations are the following:

Obligors who were reported to the DMV made more child support payments than did obligors who were not reported for license suspension. In the first year following their license suspension, the experimental group paid an average of $958 dollars toward child support. During the same time period, the comparison group that received neither DMV nor credit bureau notification paid significantly less: $548.

By returning to their earlier non-payment patterns, obligors left themselves open for subsequent notifications of intent to suspend, and actual suspensions of, driver's licenses. The following analysis considers what happens when these obligors receive subsequent notices. Do they come into compliance again? If they do comply again, is this compliance also temporary? Does each subsequent notice of non-compliance produce effects, or do effects diminish with each subsequent notification?

To address these questions, a sample was generated from the Automated Child Support Enforcement System (ACSES) of 2,596 obligors who first entered a subsequent failure classification, or Status C, in 1998. For all of these cases, the suspension followed a temporary return to compliance following notification of the pending suspension. Figure 1 summarizes the actions that qualified the cases for inclusion in the sample.

The data extracted for this sample includes information about the movement of each case from non-compliance back to compliance -- and sometimes back to non-compliance again -- as well as information about actual payments of child support.

Of the 2,596 obligors in this evaluation, 2,461 (95%) have only a single order, while 136 obligors have two or more. Because compliance patterns may be more complex for multi-order obligors, complying on some but not all orders, they are considered in a separate analysis later in this report.

Table 1 provides a few key characteristics of the 2,461 obligors with single orders who received an initial notice of failure to comply and suspension, began complying, and were subsequently noticed for suspension due to non-compliance.

On May 28, 2000, the date when the extract was generated, 64 percent of the obligors were listed in Status E (complying), while 36 percent were in Status C (subsequent failure status). An average of 732.2 days had elapsed from the time the obligor received a notice of subsequent failure (entered Status C for the first time); the range was 575 to 910 days. On average, the obligors had spent 60 percent of this time in suspended or pending suspension status.

Figure 2 provides a summary of what happened to the 2,461 obligors who were in subsequent failure status for the first time in 1998 following a period of compliance (i.e. , entered Status C for the first time after being in Status E). As this figure shows, approximately 18 percent have been in the subsequent failure category ever since. On average, they have been in subsequent failure status for 705 days, and half have been in this status for more than 697 days.

On the other hand, slightly more than a third of the obligors who entered the subsequent failure status for the first time in 1998 returned to compliant status and remained there. There was considerable variation in the amount of time these obligors stayed in subsequent failure status before returning to complying status. The range was one to 905 days, with an average of 209.5 days and a median of 93 days.

Returning to a complying status does not necessarily mean that payment is being made. Approximately a quarter (24%) of the 920 obligors who returned to and remained in the compliant status showed no record of child support payments from the time of their first notice of non-compliance to the time they permanently returned to compliant status. Over half of the cases that returned to and remained in compliant status have since been closed. Another 33 percent are cases where the only arrearage is for a TANF case (AF or BC).

With the exception of obligors who remain in the subsequent failure status continually, all groups of obligors pay significantly more child support as a result of both the initial suspension of their driver's license and the subsequent failure monitoring process. Table 2 compares child support payments at three intervals of time: in the 12- month period of time prior to the obligor's first notification of non-compliance (prior to Status A); in the period of time during which the obligor received his initial notice of failure to comply, began to comply, stopped complying, and entered subsequent failure (Status A through C); and in the 12-month period after the obligor entered subsequent failure status and his or her payment behavior was monitored with repeat notices for non-compliance (following Status C). The average amount of time elapsing between entry to Status A and C is 11.9 months and the median is 10.2.

Overall, across all the 2,461 cases in the study, child support payments rose 34 percent as a result of the initial suspension process and another 18 percent in the 12 months following the obligor's entry into Status C and his or her exposure to monitoring in subsequent failure. Thus, total payments went from $2,195,179 to $3,475,810, an increase of $1,280,631.

For some sub-groups, the increases due to monitoring in subsequent failure were even more substantial. For example, the group with the most dramatic improvement is the 37 percent of obligors who became and remained compliant after the first subsequent failure notice. In the pre-suspension period, 46 percent of these obligors paid no child support. The average paid was $1,028 and the median was $192. During their first exposure to driver's license suspension, payments for this group increased by 26 percent, with individuals paying an average of $1,299 and a median of $600. The percent paying nothing dropped to 27%. In the period after these individuals received a subsequent notice of failure and returned to compliance, 27 percent continued to pay nothing, but average payments rose another 32 percent to $1,115 and the median went up to $1,713. During this time period, these obligors paid 52 percent of their child support obligation, as compared with the 29 percent they had paid before being exposed to the DLS initiative.

Somewhat less than half (44%) of the obligors in the study moved in and out of compliance. Cases that move in and out of compliance status have spent most of their time since the initial suspension in subsequent failure status (on average 69% of their time), although at the time of the extract most (59%) were listed as complying. On average, these cases move in and out 3.3 times, and 71 percent become compliant on two or more occasions.

Moving cases into compliance, even if only temporarily, does result in increased collections of child support. For example, in the 12 months prior to their exposure to the DLS initiative, these 1,094 obligors paid an average of $932 and a median of $200, with 43 percent paying nothing at all. As a result of their first exposure to suspension, payments for this group rose by 30 percent, with the average and median moving to $1,214 and $676, respectively. The percentage paying nothing dropped to 18 percent. As a result of the notice of subsequent failure and monitoring and exposure to repeated episodes of threatened or actual suspension, payments for this group of obligors increased another 33 percent, with average and median payments in the 12 months following entry to Status C rising to $1,610 and $1,266, respectively. Only 12 percent paid nothing at all.

The only group that did not respond to monitoring was the 18 percent of obligors who became noncompliant after their first cycle of suspension and compliance and remained noncompliant ever after. Unlike their counterparts in other compliance sub-categories, these individuals exhibited all the payment gains they were going to show during their first exposure to driver's license suspension. During their transition from Status A to C, payments for this group of obligors increased 86 percent, with averages going from $519 to $970 and medians going from $0 to $357. Correlatively, the percent paying nothing dropped from 66 percent to 30 percent. In the 12 months following their entry to Status C, however, these individuals gave up. Their payments dropped by 67 percent with annual averages going down to $317, medians returning to $0, and the percent paying nothing reverting to 60 percent.

Does each subsequent failure result in diminishing returns? This does not appear to be the case. Table 3 compares payment behavior for individuals exposed to one, two, and three additional notices of subsequent failure beyond the initial DLS intervention and the first suspension and compliance cycle that brought them into the study. As expected, the number of individuals exposed to increasing numbers of threatened or actual suspensions diminishes as individuals move into compliance. Nevertheless, the intervention remains effective and brings in more payments and more payers with each iteration. Thus, after one additional cycle, the proportion paying nothing drops to 18 percent. After two additional cycles, the proportion paying nothing drops to 10 percent. And after three or more additional cycles of threatened or actual suspension and compliance, the proportion paying nothing drops to 7 percent.

Figure 3 graphs the child support revenues generated for the sample of 2,461 cases with single orders. It shows actual monthly revenues generated in the 12 months prior to the original DLS intervention and during the monitoring process. Since we only had aggregate amounts and lacked monthly payments figures for cases during the DLS intervention, we present payments at a constant monthly level during 1997.

There were 136 obligors in this evaluation with two or more orders. The average number of orders for obligors with multiple orders was 2.14. One obligor had four orders and one had five. For the sake of simplicity, we analyzed the first three orders shown on ACSES for the two individuals with four and five orders. On average, obligors with multiple orders owed $6,323 per year for monthly child support and arrears payments. This was substantially higher than the $3,993 owed by obligors with single orders.

One key question about obligors with multiple orders is whether their payment behavior is consistent across their orders. Do they pay nothing on all their orders? Do they pay something on all? Do they have a mixed payment pattern, paying something on one order and nothing on another? Table 4 provides some answers to this question. In the 12 months before they received a notice of non-compliance and became eligible for driver's license suspension, only 26 percent of the obligors paid at least something towards all their child support orders. More than half (51%) paid nothing on any of their orders. The remaining 23 percent had a mixed pattern, paying something on one order and nothing on others.

Table 4 also shows how the monitoring process affects obligors with multiple orders. In the 12 months following a second episode of threatened or actual suspension (i.e. , entered Status C for the first time after having received an initial notice of failure [Status B] and having complied and been in Status E), the proportion of obligors with a mixed payment pattern remained the same at 23 percent, but there was a shift in the percent paying nothing and those paying something on all their orders. After having experienced suspension, compliance, and re-suspension, only 19 percent of the obligors with multiple orders were paying nothing on any of their orders, while over half (57%) were paying something on all. Clearly, the monitoring process had its intended effect, with the proportion of those paying at least something on all their orders more than doubling.

As of May 28, 2000, the date of the extract, 47 percent of obligors with multiple orders were suspended. This is higher than the 36 percent suspension rate for obligors with single orders. Individuals may be suspended on the basis of their behavior on one of their child support orders irrespective of their performance on others. Theoretically, an obligor may be totally compliant on one order and still be in subsequent failure status if his or her delinquency on another order meets the threshold for license suspension. If the orders are being handled in different counties, child support technicians are encouraged to coordinate their actions so that individuals who make an effort to be compliant are not unduly penalized.

Table 5 examines the 12-month payment behavior of obligors with multiple orders who are classified as having a subsequent failure (Status C). The analysis shows that subsequent failure classification is more likely to be invoked when obligors pay nothing on all their orders and when they exhibit a mixed payment pattern. In these instances, suspension occurs 69 and 47 percent of the time, respectively. However, 40 percent of obligors who pay something on all their child support orders also wind up in Status C, suggesting that they, too, have a suspended license. On average, these individuals paid about half (54%) of what they owed, with half paying more than 49 percent and half paying less.

To examine the relationship between payment behavior and the suspension sanction, we compared individuals in Status C who make at least some payments on all of their orders with their compliant counterparts. The results of that analysis appear in Table 6. While individuals in Status C pay 50 percent of what they owe, their compliant counterparts only pay 57 percent. And while their average annual payments of $2,672 were lower than the $2,892 paid by their compliant counterparts, the differences were not statistically significant. This suggests that there is a good deal of worker discretion in the imposition of the suspension sanction in cases with multiple orders where the individual is paying something on all his or her orders but not everything that is owed. Depending upon the worker, an identical payment scenario may well result in suspensions being imposed and lifted. Obligor behavior may be a factor, too, with some making a lump sum payment that results in the reinstatement of their driver's license even though they do not continue to pay on a monthly basis.

Do child support collections for obligors with multiple orders increase with exposure to the subsequent failure process? Do these obligors wind up paying more as a result of being repeatedly threatened for suspension? The answer appears to be "yes." In the 12 months before they were selected for driver's license suspension, obligors with multiple orders paid an average of 15 percent of what they owed, with the average annual payment being $856. Half paid nothing at all (see Table 7).

Twelve months after their entry to Status C, when they received a notice of subsequent failure at least one more time (and more often if they lagged behind in making payments), obligors with multiple orders paid nearly half (45%) of what they owed. On average, these same individuals paid $2,048, as compared with the $856 they had paid a year before entering Status A. This represents an increase of 140% and is more substantial than the average increase of 58 percent exhibited by obligors with single-order cases whose average annual payments rose from $893 to $1,413.

The original evaluation of the effects of driver's license suspension showed that DLS generates an extra $514 for every selected case, or $5.14 million per year when imposed on 10,000 obligors. This study looks at the effects of imposing repeated threatened or actual suspensions for those who comply but subsequently slip out of compliance. The analysis shows that the monitoring process generates an extra $212 for every case with a single order. Since approximately 3,000 orders enter Status C for the first time each year, this translates into an additional $600,000 per year.

The analysis shows that repeated threatened or actual suspension is an incentive for many obligors who have been suspended, complied, and become delinquent a second time. Although 18 percent remain noncompliant and are immune to this intervention, the remaining group of obligors either become consistently compliant (37%) or move in and out of a compliant status (44%). Both obligors who become compliant and those who move back and forth between delinquency and compliance, pay significantly more than they did before being exposed to driver's license suspension. The intervention does not lose its effect. The second, third, and fourth suspensions all generate revenue. Monitoring clearly generates child support revenues and enhances the effectiveness of the driver's license remedy.

Monitoring also enhances payment among obligors with multiple orders. Whereas only one quarter (26%) of obligors with multiple orders made partial payments on all their orders before they were selected for the DLS intervention, this was the case for 57 percent of the sample after they had been selected for subsequent license suspension. The 12-month average paid by this group of obligors rose from $1,087 to $2,805.

The uneven treatment of obligors with multiple orders who are making partial payments on all their orders warrants attention. Essentially, individuals with identical payment circumstances are experiencing suspensions or avoiding them, depending upon the proclivities of the technician. Although obligors in this sample do not appear to be deterred from making payments, this type of subjectivity can reduce public confidence in the child support agency and erode payment behavior over time. To avoid inconsistencies, technicians in different counties should work together to set reasonable monthly payment expectations that reflect an individual's ability to pay across all his or her obligations.

The history of the child support program has been marked by the evolution of a variety of enforcement remedies. Child support workers and policy makers have long concluded that there are no magic bullets; no single intervention works with different types of obligors. The driver's license initiative itself works best with people who have a valid license and presumably would be disadvantaged without it. Monitoring works for a sub-group of that population who respond to suspension by becoming compliant but subsequently slide back and revert to non-payment. While most of these individuals toe the line after they get suspended a second or a third time, others cycle back and forth. These results confirm that some individuals only respond to the continual threat of being caught and that monitoring is an effective way to produce behavioral change with these types of people.