Significant variation exists between the non-IVD and the IV-D cases, with a difference of over $1400 of net income a month. The Administrative IVD cases display the lowest median income at $1389 while the Direct Pay non-IVD shows the highest at $2846 per month.

· The median value of the order amount is $327 for all noncustodial fathers.

This amount represents the total amount the noncustodial father is ordered to pay in child support for all the children associated with this child support order; that is, it is not the median amount per child. Additionally, this value includes any deviations – upward or downward - from the presumptive amounts established by the Schedule.

Given the differences in income across categories, order amounts vary significantly among them. Administrative IVD fathers are ordered to pay $287 while Direct Pay non-IVD fathers are ordered to pay $549.

· For all noncustodial fathers, the order amount represents 19.0 percent of their monthly income. While some variation exists among the four strata, it is fairly small.

The Administrative IVD cases are ordered to pay the highest percentage at 19.6 percent, while Court IVD orders are the smallest at 18.4 percent.

· The data indicate that deviations are common: 29.0 percent of the orders differ from the presumptive amounts. The vast majority of these (85.1%) are downward deviations with a median decrease of $113 from the presumptive amount.

Substantial variation in deviation rates exists among the four strata: the most striking is seen among the Administrative IV-D cases. Only 14.0 percent of these entail a deviation. This differs markedly from the other categories, even within the other IV-D category, where 30.3 percent have a deviation in their order. Among the Direct Pay and Payment Service non-IVD cases, 42.5 percent and 38.0 percent deviated, respectively. The significant variation in deviation rates suggests that different award-establishment processes are occurring for the Administrative IV-D cases than the other strata.

While no one reason predominated for noncustodial fathers, this is not the case for noncustodial mothers: almost half of the deviations (45.3 percent) arise from a singular cause: their incomes are below the poverty level (the AFDC Needs Standard).

The number of children in the non-IVD cases is slightly greater at an average of 1.59 children, compared to the IV-D cases with an average of 1.34 children.

These findings are consistent with the Schedule, which takes into account the fact that additional children entail additional costs, while at the same time recognizing that two children are not twice as costly as one.

The data show that both the likelihood of a deviation and the amount of the deviation increases with the number of children in the order.

Most of these (88.7 percent) are in non-IVD categories.

· The second most common order type is "Administrative Notice Default," representing 21.4 percent of all orders in the database

These account for 63.1 percent of the Administrative IV-D caseload.

· The four categories differ significantly in the types of orders they establish.

Divorces account for 77.8 percent of the Direct Pay caseload and 62.7 percent of the PSO category, but less than 10 percent of IVD cases. Paternity orders are established almost entirely within the IV-D category, with just over 15 percent handled within the non-IVD strata.

· Orders in the Urban West are somewhat higher in terms of their proportion of income (19.7 percent) compared to the orders in the East at 18.1 percent.

· Income imputation for the purpose of establishing child support orders is common: 37.8 percent of the orders for noncustodial fathers are based on imputed income values.

The proportion of income ordered is very similar: 18.4 percent for imputed income orders compared to 19.3 percent for those derived from actual income.

Among the four strata, median imputed income ranges between 75 and 80 percent of median actual income.

· Deviations are more likely among imputed income orders than actual orders in every category but PSO.

· Income imputation is as likely among noncustodial mothers as fathers: 38.4 percent.

Median imputed income ranges from 56 percent of actual income in the Direct Pay category to 92 percent in the Administrative category, a much larger spread than observed for noncustodial fathers.

Such orders are rare, comprising only 1.4 percent of noncustodial fathers and even fewer mothers. They have lower median income than other orders and much higher upward deviation amounts.

Slightly less than 14 percent of the orders have combined incomes over $5000, most of these involve noncustodial fathers. While most of these orders comply with the Schedule’s instructions, between 23 percent and 29 percent do not.

Similarly, for the 5.3 percent of orders with income in excess $7000, most adhere to the Schedule’s instructions. However, between 16 percent and 19 percent (depending on the number of children) are not ordered to pay the expected minimum amount.

For these 171 orders (4.1 percent of all orders), the median order amount is $25 for one child and $50 for two children. No child support is ordered for 20 percent of the noncustodial fathers in these cases and 38 percent of the noncustodial mothers.

There are 153 orders (4.7 percent) where the noncustodial father is ordered to pay nothing. Compared to those with nonzero order amounts, these are associated with lower median incomes and lower deviation rates. They are more likely to be an Administrative order and equally likely to be based on imputed income. The findings for noncustodial mothers with order of zero are similar.

There are 114 orders (3.5 percent) where the noncustodial father has zero income. While the median transfer amount for these fathers is $25, 24 percent are ordered to pay nothing. These orders are much less likely to involve a deviation and more likely to be Court IVD orders. The most common order type for these cases is "Judgement/Paternity" and they are much less likely to rely on imputed income. Noncustodial mothers with zero income are more likely to be "Administrative Notice Default" and found in the Administrative IVD category. Like noncustodial fathers, they are much less likely to involve a deviation or to be based on imputed income.

This report provides an analysis of the findings from the database compiled for the purposes of investigating the outcomes that flow from the point of order origin. The purpose is to further the goals of the Strategic Plan of the Office of Child Support Enforcement of increasing collection of child support. The objective of the study is to understand the processes and components of how child support orders are established.

Tables I-A and II-A provide a general overview of some of the key variables in the analysis, for those orders where the noncustodial parent is the father. Tables I-B and II-B provide the same data for noncustodial mothers. Because one of the primary interests in our analysis is to compare the outcomes associated with non-IVD cases to those of Washington’s Division of Child Support IVD cases, the variables in these tables are categorized by four strata:

"Direct Pay," (Dirpay) where one party pays child support to the other directly;

"Payment Service Only" (PSO), where payment is made through the registry;

"Court Ordered (CourtIVD);

"Administrative" (AdminIVD).

The first two - "Direct Pay" and "Payment Service Only" - represent non-IVD cases, while the latter two - "Court Ordered" and "Administrative" - are IVD orders.

The variables presented in these first two tables are: net monthly income (net), actual order amount (trxpymnt), order amount as a percent of income (pctinc), whether the order deviated from the presumptive amount in the Schedule (whdev), the amount of the deviation if it increased the order (updev) and the amount of the deviation if it decreased the order (downdev).

The bottom row, first column of Table I-A indicates that the median net monthly income of noncustodial fathers is $1757. Significant variation exists between the non-IVD and the IV-D cases, with a difference of over $1400 of net income a month. The Administrative IVD cases display the lowest median income at $1389 while the Direct Pay non-IVD shows the highest at $2846 per month.

The differences in income are reflected in the variation in noncustodial fathers’ order amounts, as would be expected from the Child Support Schedule. The median value of the order amount is $327 for all noncustodial fathers. This amount represents the total amount the noncustodial father is ordered to pay in child support for all the children associated with this child support order; that is, it is not the median amount per child. Additionally, this value includes any deviations – upward or downward - from the presumptive amounts established by the Schedule.

Given the significant variation in income across the four strata, we would expect variation in the order amounts across the categories. Table I-A confirms this: Administrative IVD fathers are ordered to pay $287 while Direct Pay non-IVD fathers are ordered to pay $549.

For all noncustodial fathers, the order amount represents 19.0 percent (.1896 in Table I-A) of their monthly income. While some variation exists among the four strata, it is fairly small. The Administrative IVD cases are ordered to pay the highest percentage at 19.6 percent, while Court IVD orders are the smallest at 18.4 percent.

Thus, although the order amounts vary significantly, reflecting the differences in income among the strata, the percent of income ordered in child support does not. Thus, the proportion of income ordered in child support is fairly uniform across the subgroups, displaying neither income progressivity or regressivity.

Table II-A provides information on the prevalence and amount of deviations for noncustodial fathers. The data indicate that deviations are common, occurring in almost one-third of the cases for noncustodial fathers. Specifically, 29.0 percent of the orders differ from the presumptive amounts (as seen in Table II-A, in the mean value of .2896 for "Fwhdev" column, "Total" row). The vast majority (85.1 percent) of the deviations are downward (that is, the deviation reduces the order from the presumptive amount). The median value of the downward deviations is $113. This is a relatively large magnitude: it represents a 32.2 percent reduction in the order amount. Those 15 percent of the orders with upward deviations have a median value of $110, resulting in a 20.4 percent increase in their order amounts.

Of particular interest is the substantial variation among the four strata: the most striking variation is seen among the Administrative IV-D cases. Only 14.0 percent of these entail a deviation. This differs markedly from the other categories, even within the other IV-D category, where 30.3 percent have a deviation in their order. Among the Direct Pay and Payment Service non-IVD cases, 42.5 percent and 38.0 percent deviated, respectively. The significant variation in deviation rates suggests that different award-establishment processes are occurring for the Administrative IV-D cases than the other strata.

In Table III-A, the reasons for deviations (for noncustodial fathers) are shown. As this table indicates, no one reason predominates in explaining why deviations occur. The most frequent reason, accounting for 18.1 percent of the deviations, is because the noncustodial father’s income is less than the AFDC Needs Standard. The next most common reason, accounting for 17.5 percent of deviations, is a result of using the Whole Family Approach in establishing the order amount. Other common reasons arise because of residential credits (13.3 percent), mutual agreement (11.8 percent), and child support from other relationships (9.1 percent).

A case was determined to have deviated from the presumptive amount in the Child Support Schedule if both of the following conditions held:

the variable "amount of deviation" (deviati in the database) was nonzero and

the variable "reason for deviation" (devreas in the database) was nonmissing.

Imposing both of the above conditions resulted in a loss of some potential cases which might have involved a deviation. For noncustodial fathers, a total 56 cases were not assigned a deviation because both conditions did not hold; specifically, 51 cases indicated a deviation "reason" but no "amount of deviation " and the other 5 cases gave an "amount" but no "reason." For noncustodial mothers, 15 cases were not assigned a deviation because they indicated a "reason" but not an "amount."

Prior to the analysis reported here, the "amount of deviation" variable had been screened to determine whether the actual order amount differed from the presumptive amount in the Schedule simply from the process of rounding to a whole or even number. Such rounding would result in a nominal deviation; such a nominal deviation, however, would not constitute a conscious decision to deviate from the Schedule and therefore these nominal deviations are not included in this analysis.

Tables I-B, II-B, and III-B present the findings for the cases where the noncustodial parent is the mother. A comparison between mothers and fathers shows that substantial differences exist. Table I-B shows that for all noncustodial mothers median net monthly income is $1060, indicating that their income is only 60.3 percent of what fathers earn. Variation exists among the subgroups, as we observed with noncustodial fathers, but to a lesser degree: Direct Pay cases have incomes of $1523 compared to $994 for Administrative IVD mothers. As would be expected given their lower incomes, noncustodial mothers have lower order amounts: the median value for all mothers is $161. Again, although variation among the subgroups is observed, it is not as wide as for fathers: Payment Service Only cases are ordered to pay the highest amount at $209, while Administrative IVD order amounts are $143.

An important distinction between noncustodial mothers and fathers is observed in the "percent of income" variable. Recall that fathers’ median value is 19.0 percent; the comparable value for mothers is 15.2 percent. The explanation for this finding cannot be due to the lower income of mothers per se, because the variable measures the order amount as a percent of income. Similar to the fathers’ cases, the Administrative IV-D cases are ordered to pay the highest percentage of their income: 16.6 percent compared to 14.1 percent for the Direct Pay orders, which pay the lowest percent of income.

The differences between mothers’ and fathers’ awards deserves scrutiny in further analyses of the database. Table II-B, however, may provide some initial information. Mothers are not more likely to have a deviation in their order: 26.4 percent of the mothers’ orders (compared to almost 29.0% of fathers’) involved a deviation. However, the amount of the downward deviation may provide a partial explanation for the differences in orders of mothers and fathers: the median downward deviation for mothers is $134, which represents a 54.4 percent reduction in their order. (Recall that fathers’ deviations were considered significant at a 32.2 percent reduction in their orders.) The reasons for deviations (for noncustodial mothers) are given in Table III-B. The results are similar to those of noncustodial fathers, in that similar reasons are most likely for the deviation (specifically, the most common are: Income Less than AFDC, Residential Credit, and Whole Family Approach). However, while no one reason predominated for noncustodial fathers, this is not the case for noncustodial mothers: almost half of the deviations (45.3 percent) arise from a singular cause: their incomes are below the poverty level (the AFDC Needs Standard).

The average number of children in each order is 1.43 with 66.7 percent of the orders involving one child and 25.3 percent with two children. Thus, 92.0 percent of the orders have one or two children. The number of children in the non-IVD cases is slightly greater at an average of 1.59 children, compared to the IV-D cases with an average of 1.34 children.

Tables IV-A and B show the income, order amounts, and order as a percent of income for noncustodial fathers and mothers, respectively. The tables here, however, provide that data categorized by the number of children in the order. As Table IV-A shows as the number of children increases both the dollar amount of the order and the order as a percent of income increases. For one child, the median order amount is $285, which represents (as seen in the last column) 17.8 percent of net income. For two children, the amount rises to $531 or 25.0 percent of income. These findings are consistent with the Schedule, which takes into account the fact that additional children entail additional costs, while at the same time recognizing that two children are not twice as costly as one. That is, the Schedule has incorporated the economies of scale associated with larger families. While a similar pattern is shown for noncustodial mothers in Table IV-B, it is much less pronounced. The median order amount rises for the second child in an order, but only by $13; the order as a percent of income rises, but only from 15.2 percent to 17.8. The higher award amounts for fathers associated with the second child may reflect, not only the costs associated with an additional child, but also the much higher net income associated with those orders: father’s net income for the one child orders is seen in Table IV-A to be $1563 compared to $2289 for those with two children. The orders for noncustodial mothers show only a very small increase in income between the one child and two child cases: $1042 compared to $1104.

Tables V-A (for noncustodial fathers) and B (for noncustodial mothers) provide the findings for deviation rates and amounts, categorized by number of children in the order. In reviewing this table, note that, given most orders involve only one or two children, the cell sizes become quite small as we examine various subgroupings. Our purpose here is to determine if deviation rates and amounts vary by the number of children in the order. While the overall deviation rate for noncustodial fathers is 29.2 percent, we see here that deviation rate is generally higher for those orders with more than one child. While the deviation rate for orders involving one child is 27.6 percent, for orders with two or three children the deviation rate is 33.0 percent. Similarly, the amount of the downward deviation steadily increases as the number of children increases. Thus, both the likelihood of a deviation and the amount increases with the number of children in the order. This pattern is also observed for noncustodial mothers, as shown in Table V-B.

We begin here with an overview of order types found in the database and their representation in the four strata. The most common order type in the database overall is a child support order prompted by a divorce, accounting for 27.8 percent of all orders. Most of these (65.5 percent) are in the Direct Pay category, and divorces account for 77.8 percent of the Direct Pay caseload. (The bulk of the remaining Direct Pay orders were "Modifications, Court Order" accounting for 14.6 percent of the Direct Pay caseload.)

The second most common order type is "Administrative Notice Default," representing 21.4 percent of all orders in the database and they account for 63.1 percent of the Administrative IV-D caseload.

The next most likely order type is "Modification, Court Order," accounting for 14.9 percent of all orders in the database. The data show 61.1 percent of the Modifications are Court IV-D cases, and these represent 28.1 percent of the Court IV-D caseload. Another 23.0 percent (of the modifications) are Direct Pay, and (as noted above) modifications account for 14.6 percent of the Direct Pay orders. Finally, 15.9 percent of the modifications are Payment Service Only, and these account for 22.9 percent of the Payment Service caseload.

Another order type appearing with some frequency is "Judgement, Paternity" which comprises 11.0 percent of all orders. Virtually all (98.7 percent) of the "Judgement, Paternity" orders are Court IV-D cases, and these represent 33.5 percent of their caseload. Additionally, "Paternity Orders" (as distinct from "Judgement, Paternity" orders) comprise 7.1 percent of the orders, 84.0 percent of them are handled within Court-IVD, representing 18.3 percent of their caseload.

Thus, this review of the data indicates that the four categories differ significantly in the types of orders they establish.

Table VI-A presents the breakdown of noncustodial fathers’ income, order amount, and order as a percent of income by order type. As would be expected, substantial variation in income and order amount exists among the different order types. Those with the highest income values (greater than $2000 per month) are divorces, modifications by court order, and "other court orders." As mandated in the Schedule, these also have the largest order amounts. The lowest income values (less than $1300 per month) are the two order types involving paternities ("paternity order" and "judgement/paternity"). Variation in the order amount as a percent of income also exists, although the pattern is less discernable. Ordered to pay the largest proportion of income are divorce (20.7 percent) and temporary court orders (21.0 percent; however, this latter category has relatively few observations.). The smallest percent of income is ordered in agreed settlement orders (16.6 percent). The two types of paternity orders are required to pay close to 18 percent of net income. The Administrative Default orders (recall these comprise the second most common order type overall) are ordered to pay 19.9 percent of net income.

Table VI-B presents the same data for noncustodial mothers. Cell sizes for some of these order types are quite small and thus the findings may not be representative. One finding that stands out, particularly given the previous findings about Administrative IV-D cases, is that the Administrative Notice Default cases are ordered to pay the highest percent of income (17.6 percent compared to the median for all order types of 15.2 percent).

To further analyze the data, the orders were examined by "Region." Depending upon the county where the order originated, the case was assigned to one of three regions within the state:

Urban West, Non-urban West, or East. The Urban West accounts for the highest proportion of the orders (45.7 percent), while 22.6 percent are from the Non-urban West and 31.7 percent are from the East.

Tables VII-A (for noncustodial fathers) and B (for noncustodial mothers) display income, order amount, and order as a percent of income categorized by region. According to the data for noncustodial fathers, as would be expected, the Urban West region orders enjoy the highest incomes, with a median value of $2073 per month, while the East region cases show the lowest incomes at $1428. The order amounts differ according to the income values, as established in the Schedule. The orders in the Urban West, however, are also higher in terms of their proportion of income (19.7 percent) compared to the orders in the East at 18.1 percent. A similar pattern (although lower values overall) is shown in Table VII-B for noncustodial mothers.

Tables VIII-A and VIII-B show the deviation rates and amounts categorized by region. Small variation is seen in deviation rates by region: orders emanating from the Non-urban West have a somewhat higher deviation rate at 29.6 percent than the Urban West at 29.3 percent, while those in the East have the lowest at 28.0 percent. As we noted before, the vast majority of deviations result in a reduction in the order amount, so we will focus our attention on those. While the dollar amount of the deviation does not vary much by region (for example, the median downward deviation in the Urban West is $117 compared to $123 in the East), the percentage the deviation represents of the order varies somewhat more substantially. That is, the $117 deviation is a 29.9 percent reduction in Urban West orders, while the $123 translates into a 36.0 percent reduction for orders in the East.

More substantial variation in deviation rates by region is shown in the data for noncustodial mothers. Table VIII-B shows that mothers in the Non-urban West are much more likely to receive a deviation than mothers in the Urban West: their deviation rates are 30.1 percent and 24.0 percent, respectively. The amount of the deviation is again substantial, representing a reduction in the order amount of 50.0 percent in the East and 63.1 percent in both regions of the West.

Income imputation for the purpose of establishing child support orders is common: 37.8 percent of the orders for noncustodial fathers are based on imputed income values. Table IX-A shows that the median imputed income for noncustodial fathers is $1363, compared to $2082 for those orders based on actual income. While the order amount is correspondingly lower for those using imputed income, the proportion of income ordered is similar: 18.4 percent for imputed income orders compared to 19.3 percent for those derived from actual income.

Table X-A shows the findings on how income imputation varies among the four strata for noncustodial fathers. Given their caseload, it is not surprising that imputation is much more common in the IV-D categories than within the non-IVD categories: 50.3 percent of the IV-D orders utilize imputed income compared to 28.0 percent of the non-IVD.

In the Direct Pay category, we see that imputed income is $2371 compared to $2954 for actual income. Thus, median imputed income in the Direct Pay category is 80.3 percent of median actual income, as compared to 74.7 percent in the PSO subgroup. Median imputed income is a similar proportion of median actual income in the IV-D categories: for the Court IV-D orders, imputed income is 74.9 percent of actual income, and in the Administrative IV-D orders, imputed is 80.8 percent of actual income.

In terms of the percent of income the order represents, within the non-IVD categories, orders based on imputed income are ordered to pay larger proportion of income: approximately 21 percent of income for imputed orders compared to 19.0 percent for orders based on actual income. Within the IV-D categories, the reverse holds: imputed income orders are a smaller fraction of income than actual income orders. For example, within the Administrative IVD category, imputed income orders are 18 percent of income compared to 21 percent for actual income orders.

The last issue to be addressed with respect to imputation is how it affects the deviation rate among the categories. Deviations are more likely among imputed income orders than actual orders in three of the four subgroups. As shown in Table XI-A, the deviation rate for Direct Pay orders which utilize imputed income is 45.0 percent (compared to 41.6 percent for those using actual income). The other non-IVD category, PSO, is the only one where the deviation rate is lower for the imputed income orders, but only slightly so: 36.9 percent for those based on imputed income compared to 38.4 percent for actual income orders. (However, in that category - and that category only - the amount of the deviation (downward) is greater for the imputed income orders ($190) than for the actual income ones ($143). Given the small cell sizes in that subgroup, these results may not be representative.)

Income imputation is as likely among noncustodial mothers as fathers: 38.4 percent of the noncustodial mothers’ orders are based on imputed income (compared to 37.8 for noncustodial fathers). The next set of tables (IXB, XB, and XIB) provides the findings on imputation for noncustodial mothers and display similar trends to those of noncustodial fathers. Table IX-B indicates that the values for net income, order amount, and order as a percent of income are less for orders with imputed income compared to those based on actual income. Specifically, for orders based on actual income, median actual income equals $1209 compared to imputed income of $977; order amount based on actual income is $222 compared to $171 for imputed income orders; the order amount as a percent of income is 15.3 percent for actual income orders in contrast to 15.2 percent for imputed ones.

Table X-B shows the findings on how income imputation varies among the four strata. In orders for noncustodial mothers, as observed for noncustodial fathers, income imputation is more common within the IV-D orders than the nonIV-D orders. The difference between these categories, however, is not as large as that observed for noncustodial fathers. Just over sixty percent (61.4%) of the IV-D orders are based on imputed income compared to slightly less than sixty percent (57.7 %) of the non-IVD orders. We need to be aware, however, as we examine these cases that the cell sizes are becoming fairly small.

As with noncustodial father, in all categories, imputed income is less than actual income. However, the proportion of imputed income to actual income displays a different pattern for mothers than fathers. For example, in the Direct Pay category, median imputed income is just over fifty percent (56.2%) of actual income for noncustodial mothers, contrasted to 80.3 percent for fathers. In the Administrative IVD category, median imputed income is a very high 91.7 percent of median actual income (compared to 80.8 percent for fathers). Another finding with respect to the Administrative IV-D orders compared to the other categories is that for all other categories the order amount and the percent of income is less for the imputed orders than those based on actual income; in the Administrative orders, however, the order amount and percent of income is higher for the imputed income cases than the ones using actual income.

Deviation rates and income imputation for noncustodial mothers are shown in Table XI-B. As we have observed in earlier sections of this analysis, deviation rates vary significantly by category. Within the non-IVD categories, orders based on imputed income are more likely to include a deviation than those based on actual income and, as we have seen earlier, overall the non-IV strata are much more likely to deviate. Compared to the findings for noncustodial fathers, two reversals are observed: 1) for noncustodial mothers, the PSO orders based on imputed income are more likely to deviate than those based on actual income, and 2) the Court IV-D orders are much less likely to deviate when based on imputed income: 15.6 percent deviation rate for imputed income orders compared to 34.2 percent for actual income. (The caution regarding small cell sizes is particularly advised here.) The amount of the downward deviation (throughout our analysis, the vast majority of the cases) is smaller for imputed orders than actual orders within the non-IVD, while the deviation is larger for the imputed orders than actual orders in the IVD cases.

The proportion of income a noncustodial father is ordered to pay in child support falls slightly as income rises. Although the relationship is significant at the .00 level, it is very weak: the Pearson Correlation coefficient is -.072. This finding is visually reinforced by the scatter plot diagram of Figure One (included in the Tables), showing the dense clustering of orders with some outliers. It is to those outliers – those orders which a represent a very small or very large proportion of net income and those orders based on very high or very low incomes - that we now turn.

One of the limitations standards established by the Washington State Child Support Schedule is that "neither parent’s total child support obligation may exceed 45 percent of net income" (without clear justification). Among forty-one orders (1.2 percent of all noncustodial fathers), the percent of income ordered in child support exceeds 45 percent of income. As Table XII shows these orders are associated with lower median incomes compared to the rest ($1441 compared to $1800). The median order amounts for those with awards greater than 45 percent of income are more than double the amount of other orders: $758 compared to $338.

Table XII also that the deviation rate is similar both groups: 30.5 for those with orders less than 45 percent of income compared to 31.7 percent for those in excess of 45 percent of income. While caution must be exercised given the very small cell sizes, nine cases with an upward deviation display very large upward deviations: a median value of $539 and a mean value of $818. Such large upward deviations have not been observed elsewhere in this analysis.

Further analysis of these orders shows that those with an order greater than 45 percent of income involve a greater number of children (on average, 2.32) compared to 1.40 children for the remaining orders. The majority of these orders (62.8 percent) are established within the IV-D categories (41.5 percent by Court IVD and 19.5 percent by Admin. IVD); 29.3 percent are in the Direct Pay category. Almost half (48.8 percent) of these orders arose because of a divorce. These orders are less likely to be based on imputed income values (31.7 percent of them compared to 42.9 percent for the rest). The vast majority (80.5 percent) originate in the West (61.0 percent in the Urban West and 19.5 percent in the Non-urban West). It is difficult to discern from the information available whether clear justification was shown in establishing these orders in excess of 45 percent of income.

The limitation standards of the Child Support Schedule state that "the economic table is presumptive for combined monthly net incomes up to and including five thousand dollars. When combined monthly net income exceeds five thousand dollars, support shall not be set at an amount lower than the presumptive amount of support set for combined monthly net incomes of five thousand dollars unless the court finds reason to deviate below that amount. The economic table is advisory but not presumptive for combined monthly net income that exceeds five thousand dollars." The limitation standards continue regarding combined income greater than seven thousand dollars, "the court may set support at an advisory amount of support set for combined monthly net incomes between five thousand and seven thousand dollars or the court may exceed the advisory amount of support for combined monthly net income of seven thousand dollars upon written findings of fact."

There are 599 cases where the combined monthly income exceeds $5000; in 89.1 percent of these orders the noncustodial parent is the father. For these cases, according to the instructions from the Schedule cited above, we would expect these fathers to pay at least their proportion of the obligation associated with the amount set for combined incomes of five thousand dollars. To determine whether their orders were consistent with those amounts, the data need to be analyzed by number and age of children and then, based on that information, the noncustodial father’s share of the basic obligation from the Schedule must be compared to the actual order amount. For example, according to the Schedule, if combined net monthly income is $5000, the basic obligation for one child, less than twelve years old, is $738. If the father’s share of family income is .7, then his proportion of the obligation would be $517. Thus, if the instructions above are being applied correctly, we would expect that noncustodial fathers whose combined income exceeded $5000 with one child less than 12 years old, would have an award no less than $517.

The data show that for cases with one child less than 12 years old, the expected median order amount is a minimum of $447. This compares to the actual median order amount for such cases of $577. Similarly, for cases with one child between 12 and 18 years old, the expected median order amount is $534 which compares to the actual median order for these cases of $555. Thus, using this measure of central tendency, the instructions of the Schedule are being applied. The data were further analyzed to explore the variation around the median; this analysis showed that 22.6 percent of the noncustodial fathers with one child (with combined monthly income in excess of $5000) were not ordered to pay the expected minimum.

For the orders involving two children, again the median expected minimum amount is less than the median actual amount ($775 compared to $839). However, there is a sizeable percentage which are not ordered to pay the minimum: 29.4 percent of the orders with two children were not ordered to pay the expected minimum amount.

This same type of analysis applies to those cases where combined monthly income exceeds $7000. Of the 211 orders with such incomes, the father is the noncustodial parent in 91.6 percent of the cases. These cases represent 5.3 percent of all noncustodial fathers’ orders. For those orders involving one child less than 12 years old, the expected minimum amount is $488 compared to the actual median order amount of $779. For those with one child over 12 years old, the expected minimum is $569 compared to the actual median order amount of $762. In 15.7 percent of the cases the father is not ordered to pay the expected minimum. Given their high incomes, it is not surprising that over 70 percent of those not ordered to pay the expected minimum are found in the non-IVD categories. For the orders involving two children, the expected minimum ($791) is again less than actual order amount ($1044), while 18.5 percent of the orders do not pay the expected minimum.

A third limitation standard imposed by the Schedule pertains to those orders with combined monthly income of less than six hundred dollars. In these cases, "a support order of not less than twenty-five dollars per child per month shall be entered for each parent…" The database contains 171 orders (4.1 percent) where the combined monthly income is less than six hundred dollars. In 43 percent of these cases the noncustodial parent is the father. For those with one child, the median child support order is $25 and, for those with two children, the median order amount is $50. Thus, using this measure of central tendency, the Schedule’s instructions are being followed. In 15 of the 74 cases (20.3 %), however, no child support is ordered. For those cases where the mother is the noncustodial parent, the actual median order amounts adhere to the Schedule’s instructions. In 38 percent of the cases, however, no child support is ordered.

In the analysis for those orders with combined income of less than six hundred dollars, we observed that for some noncustodial parents the child support order equals zero. This prompts the question as to how many noncustodial parents in general are not ordered to pay any child support.

There are 153 orders (4.7 percent) in the database where the noncustodial father’s order amount is zero. The median income of these fathers is $1388 (compared to the $1795 for fathers with nonzero transfer payments). A curious finding is that these fathers have a lower deviation rate than noncustodial fathers with nonzero order amounts: 18.3 percent have a deviation compared to 29.8 percent. However, the amount of the downward deviation is substantial and much larger than those with nonzero orders: $203 compared to $112. Fathers with an order of zero are more likely to be Administrative IVD cases (49.0 percent of them compared to 26.9 percent of those with nonzero order amounts). Among order types, they are more likely to be "Judgement/Paternity" than their counterparts: 25.5 percent of them compared to 13.3 percent of the fathers with nonzero orders. They are also more likely to be an "Administrative" order: 15.0 percent of them compared to 2.5 percent of their counterparts. There is a small difference in the likelihood of relying on an imputed income figure rather than actual income values: 49.7 percent of them had imputed income compared to 41.2 percent of those with nonzero orders. Finally, these orders are more likely to originate in the East region of the state: 37.6 percent of the them compared to 30.4 percent of the orders with nonzero orders.

There are 87 cases (12.1 percent) where noncustodial mothers have orders equal to zero. Many of them (n=38) have incomes of zero. Including those with net incomes equal to zero, the median income for those noncustodial mothers with a zero transfer amount is $744 compared to $1006 for those mothers with nonzero transfers. Like the noncustodial fathers ordered to pay nothing, these mothers have a lower deviation rate (17.2 percent) than noncustodial mothers with nonzero order amounts (21.5 percent) and the amount of their downward deviation is slightly greater, at $186 compared to $178 for those with nonzero orders. Most of these orders (83.9 percent) are in the Administrative IVD category. Regarding "Type of order," like the noncustodial fathers, these are more likely to be "Administrative Orders" (11.5 percent compared to 5.5 percent of mothers with nonzero orders). The most likely order type for these cases is "Administrative Notice Default," which accounts for 44.8 percent of them. (That is comparable to the figure of 43.0 percent for those mothers with nonzero orders.) Like the noncustodial fathers, these orders are less likely to be based on imputed income: 34.5 percent of them use imputed income compared to 56.2 percent of those mothers with nonzero orders. Regarding the region of the state, these orders are somewhat more likely to be from the Urban West (51.2 percent) than their counterparts (44.2 percent) and somewhat less likely to be from the East (32.1 percent compared to 36.3 percent).

The final subgroup to be analyzed is those orders where the noncustodial parent has zero income.

These orders have comprised a portion of each of the previous two sections. (That is, some of those with combined incomes less than six hundred have incomes of zero, and some of those with order amounts of zero have incomes of zero.) Here, however, we will examine, as a distinct subgroup, all those orders where the noncustodial parent’s income is zero.

There are 114 orders (3.5 percent) where the noncustodial father has zero income. The median transfer amount for these fathers is $25, suggesting adherence to the Schedule’s instructions that a minimum order of $25 per child will be established. However, as seen in the findings above regarding incomes less than $600, some cases have order amounts of zero: this is true for 24 percent of the orders where the noncustodial father has zero income. Their deviation rate is lower (12.3 percent) compared to orders with nonzero income (with a 29.9 percent deviation rate). A majority (60.5 percent) of these noncustodial fathers with zero income are Court IVD orders (compared to 38.3 percent of those with nonzero incomes) and 29.8 percent are Administrative IVD (compared to 27.9 percent of those with nonzero incomes). The most common order type for these cases is "Judgement/Paternity" which comprises 24.6 percent of them (compared to 13.5 percent of those cases where the noncustodial father has nonzero income). The other most common order types for these orders are: Administrative Notice Default (15.8 percent of these orders) and Modification/Court (19.3 percent). Perhaps the most striking difference is found in the likelihood of imputation: only 7.0 percent of orders based on zero income are imputed compared to 42. percent of those based on nonzero incomes. Finally, in terms of region, little variation exists: 42.7 percent of those with zero income are from the Urban West compared to 45.9 percent for those with nonzero income; 20.7 percent for zero incomes compared to 23.7 percent are from the Non-Urban West; and 36.7 percent of those with zero incomes originate in the East compared to 30.4 of those with nonzero incomes.

There are 112 noncustodial mothers with zero income (representing 15.6 percent of all orders involving noncustodial mothers). The median value of the order amount for these cases with zero income is $25; again, however, many of them have order amounts of zero: 38 of the 112 cases have order amounts of zero. Similar to noncustodial fathers, their deviation rate is much lower: 10.7 percent for them compared to 30.2 percent for noncustodial mothers with nonzero income. The data indicate 83.9 percent of these cases are processed in the Administrative IVD category (compared to 58.6 percent of those orders where the noncustodial mother has nonzero income). Regarding the type of order, 66.1 percent of these orders are "Administrative Notice Default" (compared to 39.1 percent of those orders with nonzero income). These orders are much less likely to rely on imputed income: only 16.1 percent of the orders where the noncustodial mothers has zero income were based on imputed income compared to 60.4 percent of those where the noncustodial mother has nonzero income. Finally, little variation is displayed regarding where the order originates: 46.8 percent are from the Urban West (compared to 44.6 for the orders with nonzero income); 24.8 percent of the zero income orders are from the Non-Urban West compared to 18.2 percent of the orders with nonzero income; 28.4 percent of the zero income orders come from the East compared to 37.2 percent of those orders for noncustodial mothers with nonzero income.