Department of Human Services

2008 Update of the Maryland
Child Support Guidelines Schedule

Submitted by:

1570 Emerson Street
Denver, CO 80218

303 / 837-1555
FAX: 303 / 837-1557
Jane C. Venohr, Ph.D.
jvenohr@centerforpolicyresearch.org

December 3, 2008

Points of view expressed in this document are those of the author and do not necessarily
represent the official position of the Child Support Guidelines Committee or the State. CPR
Center for
POLIC Y
RESEA RCH

Table of Contents

SECTION I: INTRODUCTION ......................................................................................................................1
Purpose of Report .............................................................................................................. ........................1
Organization of Report ........................................................................................................ .................... 2
Federal Requi rements........................................................................................................... .................... 2
Guidelines Models .............................................................................................................. ...................... 3
Basis of Maryland Guidelines................................................................................................... ..................................... 5

SECTION II: ESTIMATES OF CHILD REARING EXPENDITURES ............................................................... 7
Estimates of Child-Re aring Expenditures........................................................................................ ........ 7
Overview of Me thodologies ...................................................................................................... .................................... 8
van der Gaag (1981) estimates.................................................................................................. ..................................... 8
Espenshade (1984) estimates.................................................................................................... ..................................... 8
Betson (1990) estimates........................................................................................................ .......................................... 8
Betson (2001) estimates........................................................................................................ .......................................... 9
Betson (2006) estimates........................................................................................................ .......................................... 9
USDA estimates................................................................................................................. ............................................ 10
Comparisons .................................................................................................................... .............................................. 11
Data Sources of the Estimates.................................................................................................. ............... 11
Specific Consumption Items ..................................................................................................... .................................. 12
Usage of Estimates in State Guidelines......................................................................................... ..........15
Economic Estimates Considered by Maryland ...................................................................................... .................. 18

SECTION III: UPDATED OBLIGATION SCHEDULE ..................................................................................19
Overview of Steps Used to Develo p Updated Obligati on Schedule........................................................19
Proposed Updated Schedule of Basic Support Ob ligations ....................................................................21
Detailed Technical Step s and Additi onal Data ................................................................................... ....36
Technical Steps Used to De velop Obligation Schedule............................................................................ .............. 39

Section IV: Schedule Compar isons and Conc lusions.............................................................................4 5

APPENDIX A: SIDE-BY-SIDE COMPARISONS OF EXISTING TO UPDATED SCHEDULES

REFERENCES

Table of Contents - ii
.

EXHIBITS
Exhibit 1: Principles of St ate Child Support Guidelines ....................................................................... .......................... 3
Exhibit 2: State Usage of Child Support Guidelines Models...................................................................... ..................... 5
Exhibit 3: Comparisons of Estima tes of Child-Rearing Expenditures............................................................. .......... 10
Exhibit 4: Variations in Child-Rearing Expenditures by Income ................................................................. ............... 11
Exhibit 5: Partial List of Expenditures Items Considered ...................................................................... ...................... 12
Exhibit 6: Composition of Av erage Spending by Families........................................................................ ................... 13
Exhibit 7: Family Consumption and Income...................................................................................... ............................ 20
Exhibit 8: Updated 2008 Schedule.............................................................................................. ...................................... 21
Exhibit 9: Parental Expenditures on Children.................................................................................. .............................. 38
Exhibit 10: Table of Support Proportions...................................................................................... ................................. 41
Exhibit 11: Illustration of the Hidden Net Income Column ...................................................................... .................. 42
Exhibit 12: Comparison of the Economic Ba sis o the Existing and Updated Schedules ....................................... 45
Exhibit 13: Comparison of Existing to Proposed Schedule: One Child........................................................... ........ 46
Exhibit 14: Comparison of Existing to Proposed Schedule: Two Children........................................................ ..... 46

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 1

Section I: Introduction
PURPOSE OF REPORT

Child support contributes to the financial well-being of many Maryland children. In 2007,
the Census reported that there are 1,358,576 children living in Maryland and 442,210 of
those children live with only one parent.1 This amounts to one out of three children living
with only one parent. Most of these children are eligible for child support. An unknown
number of Maryland children living in other si tuations are also eligible for child support. One
of their parents may have remarried and they now live with a step-parent or they live with
foster parents or in other situations without both of their parents. The Maryland Depart-ment of Human Resources, Child Support Enforcement Program (CSE) collects about $475 million in child support annually for many of these children. An unknown amount of addi-
tional support is paid to non-CSE cases.

In Maryland, child support orders are set usin g the child support guidelines provided in
statute (Maryland Code, Family Law, Sections 12-201 through 12-204). The core of the guidelines calculation is a lookup schedule of basic obligations for a range of incomes and
number of children. The basic obligations re flect economic data on the costs of raising
children. The obligated parent’s pro rata share of the basic obligation forms the basis of the
child support order. Additional adjustments are made for actual child care expenses, the
actual cost of health insurance for the children , shared physical custody, and other factors.

With the exception of changes at incomes below minimum wage, most of the Maryland
schedule has not been updated since 1989. As a consequence, it does not reflect the current costs of raising children. Moreover, the Maryland guidelines ranks low when com-
pared to other states. A 2005 study (Venohr an d Griffith) found that the Maryland guide-
lines amount for a mother and father with median income ranked 40
th among all states and
the District of Columbia. In contrast, Maryla nd ranks first among all states and the District
of Columbia in median family income.2

The purpose of this report is to provide te chnical documentation of an updated Maryland
child support schedule. The updated schedule re flects the cost of raising children in Mary-
land in 2008. The report also fulfills Federa l and State requirements that the child support
guidelines be reviewed at least once every four years.3 As part of the federal requirement,
states must consider economic data on the cost of raising children.
The Child Support Enforcement Administration of the Department of Human Resources is
responsible for the review and reporting its findings to the Legislature.
4 To this end, they

1 U.S. Census American Community Survey (2007). Downloaded from http://factfinder.census.gov on Novem-
ber 25, 2008.
2 Supra note 1.
3 Title 45, Public Welfare, CFR 302.56 and MD Family Law §12-202 (c).
4 MD Family Law §12-202 (c).

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 2
convened a committee of diverse stakeholders (e.g., judges, family law attorneys, a legisla-
tor from the Maryland House of Representatives, a representative of the Office of the Attor-ney General, and representatives of CSE) to review the guidelines and make recommenda-
tions. CSE also contracted wi th the Center for Policy Resear ch (CPR) to provide technical
assistance; namely, to review current economic data and develop an updated schedule with
the Committee’s input.
This report focuses on the updated child supp ort schedule. It does not document the nu-
merous other considerations and recommendations of the Committee.
ORGANIZATION OF REPORT
The report is organized into four sections.
¾ Section I provides an introduct ion. This includes the purpose of the report, a summary of
federal regulations pertaining to state child support guidelines; and, an overview of the Maryland child support guidelines relative to other state guidelines.
¾ Section II reviews estimates of child-rearing expenditures. It reviews those underlying
state child support guidelines.
¾ Section III contains the updated schedule. The 2008 updated schedule is in Exhibit 8.
This section also outlines the data and steps used to develop the updated schedule.
¾ Section IV compares the amou nts under the new and updated schedule. It also summa-
rizes the changes in the economic factors underlying the schedule.
Side-by-side comparisons of the new and update d schedule are provided in Appendix A.

FEDERAL REQUIREMENTS
Federal law has required state adviso ry child support guidelines since 1987.5 The Family
Support Act of 1988 expanded the requirement. As of 1989, each state must have one
guideline that is to be applied presumptively rather than on an advisory basis.6 It also re-
quires each state to establish deviation criter ia that allow for the rebuttal of the state’s
presumptive guideline. The state-determined cr iteria must take into consideration the best
interests of the child. The Maryland guidelin es provide deviation crit eria in State statute
[MD Family Law §12-202 (a)].
Federal regulation requires states to review th eir child support guidelines at least once every
four years [45 CFR 302.56]. As part of that re view, states must consider economic data on
the costs of child-rearing expenditures and exam ine case file data to analyze the application
and deviation from the guideline.7

5Advisory statewide guidelines were required as part of Child Support Enforcement Amendments of 1984 [P.L.
No. 98-378].
6Presumptive guidelines were required as part of The Family Support Act of 1988 [P.L. No. 100-485].
7 The Family Welfare Research and Train ing Group, School of Social Work, University of Maryland conducted
the 2008 case file review. A report of their fi ndings is available from their website, URL:
http://www.familywelfare.umaryla nd.edu/reports/guidelines08.pdf

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 3
GUIDELINES MODELS
S t a t e s h a v e d i s c r e t i o n i n t h e g u i d e l i n e s m o d e l s t h a t t h e y u s e . Y e t , a c c o r d i n g t o f e d e r a l
requirements, they must: ¾ Be based on specific descriptive and numeric criteria;
¾ Take all earnings and income of the noncustodial parent into consideration; and
¾ Provide for the child(ren)’s health care needs.
Most states, including Maryland, base their gu idelines on the Income Shares Model, which
was developed through the 1984-87 National Child Support Guidelines Project.
8 Convened
by the federal Office of Child Support Enforc ement at a request of Congress, the Project
made recommendations for the development of state guidelines. Prior to the 1987 re-
quirement, few states had statewide guidelines.

Income Shares Model
The Income Shares model was developed to em body the principles of state child support
guidelines identified by the Guidelines Project’s Advisory Panel. (Those principles are shown in Exhibit 1.) It also incorporates actual evid ence of child-rearing expenditures. The Income
Shares guidelines model is based on the premis e that the child should be entitled to the
same level of expenditures that the child woul d have received had the parents lived together
and combined financial resources. As a co nsequence, the core of the Income Shares
model is a measurement of how much families sp end on child rearing. In turn, that amount
is often adjusted in a guidelines worksheet fo r different situations such as shared physical
custody situations, children from other relationships, and other factors.
Exhibit 1:
Recommendations of the 1984-87 Child Suppor t Guidelines Project Advisory Panel:
Annotated Summary of Basic Principles for State Child Support Guidelines
1. Both parents should share in the financial support of their ch ildren. The responsibility should be divided in proportion to
their available income.
2. The subsistence needs of each parent should be considered, but in virtually no case, should the obligation be set at zero.
3. Child support must cover a child’s basic needs as a first pr iority; but, to the extent either parent enjoys a higher standar d of
living, the child is also entitled to share in that higher standard of living.
4. Each child of a given parent has a right to a share of that parent’s income. (In other words, when a parent has other
children besides the children for whom support is being determined, an adjustment may be appropriate.)
5. The guidelines should not treat children of separ ated, divorced, and never-married parents differently.
6. The guidelines should not assume whether the mother or father is the custodial parent.
7. The guidelines should not create economic disincentives to re marry or work. (An economic disincentive to remarry could
exist if the guidelines considered a new spouse’s income. An economic disincentive to not work is avoided by imputing
income to a parent who is voluntarily unemployed or underemployed.)
8. The guidelines should consider the involv ement of both parents in the child’s upbringing. It should take into consideration
the financial support provided by parents in shared physical custody or extended visitation arrangements. Yet, this does
not necessarily obviate the child support obligation in 50/50% timesharing arrangements.

8 National Center for State Courts (1987). Development of Guidelines for Ch ild Support Orders, Final Report.
Report to U.S. Department of Health and Human Services, Office of Child Support Enforcement, Williamsburg,
Virginia.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 4

The premise of the Incomes Shares model applie s to children of previously married parents
as well as never married parents. Children should not be forced to live in poverty because of
their parents’ decisions to never marry or divorc e. Children of disrupted families, regardless
of the reason for the disruption, should be afforded the same financial opportunities as
children of intact families with similar incomes.
Another major premise of the Income Shares model is that both parents are financially
responsible for their children. To this end, the average amount expended on children is
prorated between the parents. The obligated parent’s share becomes the basis of the child
support award. There may be other adjustme nts for physical custody or other factors.

Other Guidelines Models
Judge Melson of Delaware developed the Melson formula. It first considers the basic needs
of the children and each parent. If the obliga ted parent’s income is more than sufficient to
cover his share of the basic needs of the children and his basic needs, an additional per-
centage of his remaining income is assigned to child support. The percentage-of-obligor
income guidelines model is the simplest and ol dest guidelines model. It assigns a flat or
sliding-scale percentage of obligor income to support.
In recent years, various groups have introduced a few new guidelines models (i.e., The Chil-
dren’s Right Council first introduced a version of the Cost Shares model, the American Law
Institute introduced its guidelines conc ept); yet, no state has adopted them.
9

State Usage of Guidelines Models
As shown in Exhibit 2, as of 2008, there ar e 37 states that currently rely on the Income
Shares model; 10 states that use a percentage-of-obligor income guidelines model; three states that rely on the Melson formula; and on e state that uses a hybrid approach between
the Income Shares model and percentage-of-obligor guidelines.
Until recently, few states have changed guidelines models. However, beginning in 2005,
several states adopted Income Shares. Tenne ssee, Georgia and Minnesota moved from the
percentage-of-obligor model to Income Shares guidelines. The most recent guidelines change occurred in the District of Columbia, wh ich moved to an Income Shares approach as
of April 2007 after previously using a hybrid guidelines model. Massachusetts still uses a
hybrid guidelines model. The Massachusetts gu ideline starts with a percentage-of-obligor
income model when the custodia l parent had low income but sw itches to an Income Shares-
like approach when the custodial parent’s in come exceeds $20,000 per year with additional
adjustments for more than one child and work-r elated child care expenses. The premise is
that the custodial parent needs a little more at very low incomes to boost the family out of
poverty.

9 For more information on these gu idelines models see the 1999 Chil d Support Symposium published by
Family Law Quarterly (Spring 1999) and Beld and Biernat (2003).

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 5

Basis of Maryland’s Guidelines

Maryland bases its guidelines on the Income Shares model. The schedule was developed in
the late 1980’s based on measurements of child-rearing expenditures estimated from fami-lies surveyed in 1972-73.

The Committee discussed other guidelines models and agreed to retain the Income Shares
model.

Income Shares (37 States)
Percentage of Obligor Income (10 States)

Other [3 States (HI, DE, MT) are based on
Melson Formula; MA use a hybrid approach] Exhibit 2
State Usage of Child Support Guidelines Models

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 6
This page is intentionally blank.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 7
Section II:
Estimates of Child-Rearing Expenditures &
Expenditures Data

The purpose of this section is to describe th e estimates of child-rearing expenditures. The
estimates are first summarized. This is followed by a discussion of the data source used to
produce the various estimates. Finally, this se ction concludes with a discussion of the us-
age of these estimates in state guidelines.
ESTIMATES OF CHILD-REARING EXPENDITURES
Most state child support guidelin es rely on one of the following studies on the costs of rais-
ing children:

¾ Van der Gaag, Jacques (1981). On Measuring the Cost of Children . Discussion Paper
663-81. University of Wisconsin Institute for Research on Poverty, Madison, Wisconsin.

¾ Thomas J. Espenshade, Investing in Children: New Estimates of Parental Expenditures ,
Urban Institute Press: Wa shington, D.C. (1984).
¾ David M. Betson, Alternative Estimates of the Cost of Children from the 1980-86 Con-
sumer Expenditure Survey , Report to U.S. Department of Health and Human Services,
Office of the Assistant Secretary for Planning and Evaluation, University of Wisconsin In-
stitute for Research on Poverty, Madison, Wisconsin (1990).

¾ David M. Betson, “Chapter 5: Parental Expenditures on Children,” in
Judicial Council of
California, Review of Statewide Uniform Child Support Guidelines , San Francisco, Cali-
fornia (2001). Available at URL: http://www.courtinfo.ca.gov/programs/cfcc/1058files2001/CH5.PDF

¾ Betson, David M. (2006). “Appendix I: New Estimates of Child- Rearing Costs in State of
Oregon Child Support Guidelines Review: Up dated Obligation Scales and Other Consid-
erations, Report to State of Oregon, Prepared by Policy Studies Inc., Denver Colorado.
Available at URL:
http://www.dcs.state.or.us/ oregon_admin_rules/psi_guidelines_review_2007.pdf
¾ Mark Lino, Expenditures on Children by Families: 2007 Annual Report , U.S. Department
of Agriculture, Center for Nutrition and Poli cy Promotion. Miscellaneous Publication No.
1528-2007 (2008). Available at URL: http://www.cnpp.usda.gov/ExpendituresonChildrenbyFamilies.htm

The studies rely on various methodologies to es timate child-rearing expenditures and differ-
ent data years. All of the studies except va n der Gaag rely on the Consumer Expenditures
Survey (CES), which we will di scuss in more detail later.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 8
Overview of Methodologies
One reason that we discuss so many different estimates of child-rearing expenditures is
because not all economists arrive at the sa me estimate. Moreover, economists do not
agree on which estimate best reflects actual child-rearing expenditures. Part of the problem
is that there is no perfect methodology for separating the children’s share of family expendi-
tures from the parents’ share. To illustrate th is issue, consider family expenditures for elec-
tricity used in the home. The children’s share of electricity is not obviously separable from
the parents’ share.
The most common methodology is a marginal co st approach, which compares expenditures
between two equally well-off families: (a) marrie d couples with children, and (b) married
couples of child-rearing age without children. The difference in expenditures between these
two families is deemed to be child-rearing expenditures. The Engel and Rothbarth method-
ologies, named by the economists who developed them, are both forms of the marginal cost
approach. The Engel methodology uses expendit ures on food, while the Rothbarth method-
ology relies on expenditures for adult goods (specifically, adult clothes in the Rothbarth
estimates that form the basis of state guidelines) to determine equally well-off families.
Most economists believe that the Engel estima tor overstates actual child-rearing expendi-
tures and the Rothbarth estimator understa tes actual child-rearing expenditures.

van der Gaag (1981) estimates
In his study, van der Gaag concluded that a co uple that adds one child to the household
needs 25 percent more gross income in order to maintain the standard of living they en-
joyed when they had no children. Wisconsin us ed van der Gaag’s estimates to develop its
child support table, although adjusted this percentage to account for taxes and other fac-tors. Since then, several states adopted Wisconsi n’s flat percentage of obligor gross income
as their guidelines formula.

Espenshade (1984) estimates
Most states, including Maryland , relied on Dr. Espenshade’s measurements when they first
developed child support guidelines in the 1980s because his was the most authoritative
study available at the time. It formed the ba sis of the prototype Income Shares model de-
veloped through the 1984-1987 National Child Support Guidelines Project.
10 Using the
Engel methodology, Espenshade found that families spend about $58,000 to $138,000 (in 1981 dollars, hence over twice as much in 2007 do llars) to raise a chil d from birth to age
18 years.

Betson’s Three Studies
In the past 18 years, Dr. Betson has conducte d three studies estimating child-rearing ex-
penditures. Each study us es more recent data.

Betson (1990) estimates . Dr. Betson applied five different methodologies to estimate child-
rearing expenditures using 1980-86 CES data.
11 He concluded that estimates using the

10 National Center for State Courts (1987).
11 The five approaches were (1) Engel, (2) Rothbarth, (3 ) ISO-PROP, (4) Barten-Gorman, and (5) per capita (i.e.,
average cost approach, similar to the USDA approach).

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 9
Rothbarth methodology were the most robust , and hence recommended their use. He re-
jected his estimates using the Engel methodol ogy, which was used by Espenshade, because
they approached implausibly high levels. Be tson’s application of the Rothbarth estimator
finds that the percentages of total household expenditures devoted to children are, on aver-
age: (a) 24 percent for one child, (b) 34 perc ent for two children, and (c) 39 percent for
three children in an intact family. Betson’s ap plication of the Engel estimator finds that the
percentages of total expenditures devoted to children are, on average: (a) 33 percent for
one child, (b) 39 percent for two children, and (c ) 49 percent for three children in an intact
family.

Betson (2001) estimates. In 2001, Dr. Betson updated his 1990 estimates based on the
Rothbarth and Engel methodologies using more recent data (1996-98, initially, but later
expanded it to include 1996-99). The only difference between the 2001 and earlier esti-mates was in the years the data were gathered . The source of data (CES), the estimation
methodologies, and the assumptions he used to develop the estimates did not change.
These estimates form the basis of many state child support guidelines including the current
Pennsylvania schedule. Using the more current data, Betson’s application of the Rothbarth
estimator found that the percentages of total household expenditures devoted to children are, on average: (a) 26 percent for one chil d, (b) 36 percent for two children; and (c) 42
percent for three children in an intact family . Betson’s application of the Engel estimator
found that the percentages of total expenditures devoted to children are, on average: (a) 32 percent for one child, (b) 46 percent for two chil dren, and (c) 58 percent for three children in
an intact family.

Betson (2006) estimates . In 2006, Dr. Betson upda ted his 2001 estimates using the
Rothbarth methodology with da ta from 1998 through the first quarter of 2004 for the state
of Oregon. The 2004 survey was the most recent data available from the CES at that time.
Dr. Betson did not update the estimates using the Engel methodology or other approaches.
The Oregon table at the time was based on th e Betson-Rothbarth estimates relying on sur-
vey data collected in 1996-99 and Oregon only wa nted to use updated su rvey data. (A more
complete discussion of Dr. Betson’s findings using the updated data is available in the 2006
Oregon guidelines review report.)
For this set of estimates, Betson relied on data from a wide range of years for two reasons.
First, it increased the sample size, hence the re liability of the estimate. Secondly, it spanned
several economic cycles: the high growth of the late 1990s; the short recession in 2002;
and the stable economic growth afterwards. Similar to the 2001 update, he applied the
same assumptions and method, but he used more recent data. His findings showed that
the child-rearing expenditures as a proportion of total household expenditures are, on aver-age: (a) 25 percent for one child, (b) 37 perc ent for two children, and (c) 44 percent for
three children. Over three data periods, the an alysis shows that the proportion of household
expenditures devoted to children has increased, albeit somewhat less for families with one child (from 24 percent using 1980-86 data to 25 percent using 1998-2004 data) than for
families with three children (39 percent to 44 percent).

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 10
USDA (2007 estimates)

The USDA estimates child-rearing expenditures individually for several expenditure catego-
ries (e.g., food, transportation, housing); then , adds them to develop a total. For some ex-
penditure categories (e.g., housing), the USDA uses a per capita approach to estimate the
child’s share of the costs. That is, they divide the expenditures for that particular good by the
number of family members. Most economists believe this approach overstates the child’s
actual share of expenditures. The USDA update s its estimates every year for changes in the
price level; however, the database for the current estimates is CES data from 1990-92. The
USDA estimates expenditures for one child in a two-child family to be between $7,830 and
$17,500 per year, depending on the age of th e child and the income of the parents.

Comparisons

Exhibit 3 compares the estimates of child-rearing expenditures by the number of children. It
shows that those estimated with the Engel meth odology result in higher amounts on average
than those estimated with the Rothbarth method ology. It also shows that the USDA esti-
mates generally fall between the two methodologies.12
Exhibit 3: Comparison of the Estimates of Child-Rearing Expenditures
(averaged across all income ranges)
24%41%51%
33%49%59%
30%44%52%
25%37%44%
25%35%41%
25%37%44%
26%42%48%
0%10%20%30%40%50%60%70%
1 Child 2 Children 3 Children% of total family expenditures devoted to
child-rearing costs
Espenshade-Engel (1972-73 data) Betson-Engel (1980-86 data)
Betson-Engel (1996-99 data)
Betson-Rothbarth (1980-86 data)
Betson-Rothbarth (1996-99 data) Betson-Rothbarth (1998-2004 data)
USDA (1990-92 data)

12The USDA estimates are based on gross income. The fa mily-expenditures equivalent of the USDA estimate,
which is cited in Lino (2007) comes from a much olde r study. We do not know what tax and expenditures
assumptions were made in the conversion. When we convert current USDA estimates from gross income to
family expenditures using prevailing tax rates and expe nditures data, we find that the USDA estimates are
generally higher than the Engel estimates.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 11
We note two caveats to Exhibit 3. First, it compares the average pe rcent of total family
expenditures devoted to child-rearing expendit ures, while the Maryland child support sched-
ule relates to “gross income” rather than “total family expenditures.” Gross income and
total family expenditures differ because of two factors: income taxes; and, some families
spend more or less than their after-tax income . Later in this report, we convert these
measurements back to gross income. Second, Exhibit 3 reflects “average” child-rearing expenditures across all income ranges, so it does not reflect how the estimates change
when there is more income. Most economists find that the percentage of total family ex-
penditures devoted to child-rearing expenditures declines as income increases. Exhibit 4
illustrates this by showing the most recent Betson-Rothbarth estimates for a range of in-
comes.

Exhibit 4: Variation in the Percent of After-Tax Income Devoted to Child-Rearing
Costs by Income Range
(Betson-Rothbarth estimates in 2005 dollars)
0%10%20%30%40%50%
0-$15,000 $15,001-
$25,000$25,001 -
$50,000$50,001 -
$75,000$75,001 -
$100,000$100,001 -
$125,000$125,001 or
more
After-tax incomePercent of income devoted to child-
rearing costs
One Child Two Children Three Children
DATA SOURCE OF THE ESTIMATES
With the exception of van der Gaag, all of th e economists estimated child-rearing expendi-
tures from the Consumers Expenditures Survey (C ES) that is administered by the Bureau of
Labor Statistics (BLS).13 Economists use the CES because it is the most comprehensive and
detailed survey conducted on household expenditures and consists of a large sample.14
The CES surveys about 6,000 households per quarter on expenditures, income and house-
hold characteristics (e.g., family size). Households remain in the survey for five consecutive
quarters with households rotating in and out each quarter. Most economists use at least
three quarters or a year of expenditures data for a surveyed family. This means that family
expenditures are averaged for about a year rather than over a quarter, which may not be as
reflective of typical family expenditures.

13 van der Gaag’s study is more of a literature review of the evidence of child-rearing expenditures that existed
in the early 1970’s.
14 Detailed information about the CE S can be found at the BLS website: http://www.bls.gov .

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 12

The BLS designed the CES to produce a nationally representative sample and samples representative of the four regions (Midwest, Northeast, South, and West). The sample sizes
for each state, however, are not large enough to estimate child-rearing costs for families
within a state. We know of no state that has seriously contemplated replicating the CES at
the state level. The costs and time requirements make it prohibitive.

Specific Consum ption Items

The CES asks households about expenditures on over a hundred detailed items. Exhibit 5
shows the major categories of expenditures captured by the CES. The CES focuses on cur-
rent consumption. It includes the purchase pr ice and sales tax on all goods purchased
within the survey period.

Exhibit 5
Partial List of Expenditures Items Considered in the BLS,
the Data Source Used to Esti mate Child-Rearing Expenditures
Housing Rent paid for dwellings, rent received as pay , parking fees, maintenance, and other expenses for
rented dwellings; and interest on mortgages, in terest on home equity loans and lines of credit,
property taxes and insurance, refinancing and prepayment charges, ground rent, expenses for
property management and security, homeowners' insurance, fire insurance and extended coverage, expenses for repairs and maintenance contract ed out, and expenses of materials for owner-
performed repairs and maintenance for dwellings used or maintained by the consumer unit. Also
includes utilities, cleaning supplies, household te xtiles, furniture, major and small appliances and
other miscellaneous household equipment (tools, plants, decorative items).
Food Food at home purchased at grocery or other food stores as well as meals, including tips, purchased
away from home (e.g., full service and fast food restaurant, vending machines).
Transportation Vehicle finance charges, gasoline and motor oil, ma intenance and repairs, vehicle insurance, public
transportation, leases, parking fees and other transportation expenditures.
Entertainment Admission to sporting events, movies, concer ts, health clubs, recreational lessons, televi-
sion/radio/sound equipment, pets, toys, hobbies and other entertainment equipment and services.
Apparel Apparel, footwear, uniforms, diapers, alterations and repairs, dry cleaning, sent-out laundry, watches
and jewelry.
Other Personal care products, reading materials, educat ion fees, banking fees, interest paid on lines of
credit, and other expenses.

Mortgage Payments
The CES does not include mortgage principal pa yment as part of current expenditures.
However, the CES includes payment of the mortgage interest, rent among households that rent, utilities, property taxes, and other housin g expenses as indicated in the above table.
As shown in Exhibit 6, even with the exclus ion of the mortgage principal payments, these
housing items comprise the largest share of total family expenditures. Housing expenses contribute to over a third of family expenditures.
For purposes of developing child support schedule s, we have found it beneficial that the CES
excludes mortgage principal payments. In most situations, any equity in the home is con-
sidered as part of the property se ttlement during the divorce.

Finance Charges
The CES does not capture finance charges (wit h the exception of finance charges for homes
and vehicles). Specifically, it does not reflect any interest charges for items paid by credit
card. Nor, does it capture interest charges fo r installment plans often used to pay for furni-

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 13
ture, household appliances, televisions and other electronics. Nonetheless, data from the
Federal Reserve Board’s Consumer Finance Surv ey (CFS) suggest that the finance charges
on items not included in the CES are nominal. 15 Almost half (46%) of families make fixed
installment payments on automobiles loans; st udent loans; and, loans for furniture, appli-
ances and other durable goods. Automobile loan s, which are included in the CES, comprise
many of these loans.
For the purpose of developing child support schedules, it would be ideal to include finance
charges for furniture and other items consumed by children. Yet, the CFS data do not break down finance charges in that way. Further, ba sed on the limited amount of information that
is available from the CFS, it suggests that the finance charges are likely to be nominal when
averaged across families.

Exhibit 6
Composition of Average Spending by Families
(adopted from Betson 2006)
Expenditure Category Childless
Couple Family with
One Child Family with
Two Children Family with Three
or More Children
Total Annual Expenditures $44,728 $46,140 $49,834 $48,341
Budget Share (% of Total Expenditures)
Food 16.0% 16.7% 17.2% 19.5%
Housing 36.6% 38.2% 38.3% 37.6%
Apparel 3.6% 4.1% 4.3% 4.6%
Transportation 21.7% 21.8% 21.0% 19.7%
Entertainment 6.0% 5.6% 6.3% 6.1%
Health Care 6.0% 5.3% 5.0% 4.8%
Personnel Care 0.9% 0.8% 0.8% 0.8%
Reading 0.5% 0.4% 0.4% 0.3%
Education 1.2% 1. 5% 1.4% 1.4%
Personal Insurance 1.6% 1.2% 1.3% 1.1%
Miscellaneous 5.8% 4.3% 7.5% 3.5%

Transportation and Vehicle Payments
As shown in Exhibit 6, transportation expens es account for about one fifth of total family
expenditures. In the category of “transportation,” the CES includes net vehicle outlays;
vehicle finance charges; gasoline and motor oi l; maintenance and repairs; vehicle insurance;
public transportation expenses; and vehicle rent als, leases, licenses and other charges. The
net vehicle outlay is the purchase price of a ve hicle less the trade-in value. It accounts for
about 40 percent of all transportation expenses among families with children in the CES,
which is seven percent of total household expenditures.16
In developing estimates of child-rearing expenditures, Professor Betson excludes net vehicle outlays because it does not reflect that the vehicle can be sold again later after the survey
period. In contrast, the USDA starts its estima tes with all transportation expenses including
net vehicle outlays.

15 Brian Bucks, et al. “Recent Changes in U.S. Family Fi nances: Evidence from the 2002 and 2004 Survey of Con-
sumer Finances”, Federal Reserve Bulletin .
16 Bureau of Labor Statistics , Table 5: Composit ion of consumer unit: Average annual expenditures and char-
acteristics, Consumer Expenditures Survey 2007.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 14
There are some advantages and disadvantages to Betson’s approach. Like home equity,
vehicles are considered as part of the property settlement in a divorce. Nonetheless, the ideal would be to include a value that reflec ts depreciation of the vehicle over time.

Other Adjustments to the CES
Professor Betson also excludes other expendit ures items captured by the CES because they
are obviously not child-rearing expenses. Specifically, he excludes contributions by family members to social security and private pension plans; and, cash contributions made to
members outside the surveyed household. The USDA also excludes these expenses from its
estimates of child-rearing expenditures.

Net Income
Gross and net incomes are report ed by families participating in the CES. The difference
between gross and net income is taxes. In fact, the CES uses the terms “income before
taxes” and “income after taxes” instead of gr oss and net income. Income before taxes is
the total money earnings and selected money re ceipt. It includes wages and salary, self-
employment income, Social Security benefits, pensions income, rental income, unemploy-ment compensation, workers’ compensation, veteran’s benefits, public assistance, and
other sources of income. Income and taxes are based on self-reports and not checked
against actual records.
The BLS has concerns that income may be unde r-reported in the CES. Although underre-
porting of income is a problem inherent to surveys, the BLS is particularly concerned be-
cause expenditures exceed income among low-in come households participating in the CES.
The BLS is unclear about whether the cause is underreporting of income or that low-income
households are actually spending more than their incomes because of an unemployment
spell, being a student, or otherwise withdrawing fr om their savings. In an effort to improve
income information, the BLS added and revise d income questions in 2001. The BLS be-
lieves that this has improved income data but there are no notable differences in the com-piled data reports.
The Relationship of Expenditures to Income
Beside mortgage principal payments, the BLS also does not include changes in net assets or
liabilities as income or expenditures. In a ll, the BLS makes it clear that reconciling differ-
ences between income and expenditures or prec isely measuring income are not part of the
core mission of the CES. Rather, the core mission is to measure and track expenditures.
The BLS recognizes that at some low-income levels, the CES shows th at total expenditures
exceed after-tax incomes; and, at very high incomes, the CES shows total expenditures are
considerably less than after-tax incomes. However, the BLS does not try to explain these differences.

In developing child support schedules, a long -standing assumption has been that the differ-
ence between after-tax income and expenditures at higher incomes is a form of “savings.”
This includes traditional savings (i.e. deposits into a bank account) and other contributions
to family wealth such as mortgage principal payments. For example, according to the most
recent CES, high-income households (i.e., households with incomes over $150,000 per

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 15
year), the ratio of expenditures to after-tax income is 55 percent.17 This suggests a consid-
erable amount of “savings.”
A high level of “savings” seems to contradict reports about the national savings rate being
low. However, economists calculate the national savings rate using a different methodol-ogy.
18 Some of the differences concern the tr eatment of housing and medical expenses.
When calculating the national savings rate, ec onomists define savings to be the difference
between disposable income and consumption. In defining consumption, economists impute
the rental value of housing to homeowners ev en though the rental value may exceed the
mortgage payment. Similarly, economists impute the value of all medical services received
even though there was insurance coverage and the family incurred no out-of-pocket ex-
pense. These imputed values increase consumption considerably; hence, reduce the na-
tional savings rate. In fact, the escalating cost of health services contributes significantly to
the declining national savings rate.19

Other exclusions made by Betson in his estima tes further exasperate the seemingly low ratio
of expenditures to after-tax income at high in comes. As discussed above, Betson also ex-
cludes net vehicle outlays, cash gifts, and co ntributions to private pension plans when esti-
mating child-rearing expenditures. These items, which are significant, could explain some of
the gaps between after-tax income and annual expenditures.20

USAGE OF ESTIMATES IN STATE GUIDELINES
States rely on various estimate s of child-rearing expenditures as the basis of their guide-
lines. Several states rely on whatever was the most current estimate available at the time
they developed or last revised their guidelines and have not updated as new estimates became available. Still other states made a deliberate choice to use one estimate over another. Often, these states chose the esti mator based on which one produced guidelines
amounts that differed the least from their current amounts.
Based on our current knowledge, we have coun ted the number of state guidelines by their
economic basis. We note that many states modified the estimates or combined them with
other information to arrive at their guidelin es amounts. Consequently, even though some
state guidelines share the same estimates, th eir guidelines amounts may differ. Another
c a v e a t t o o u r c o u n t s i s t h a t s o m e s t a t e s h a v e r e c e n t l y c h a n g e d t h e i r g u i d e l i n e s o r h a v e adopted new guidelines that are not yet promulgated.

17 Calculated from BLS, Table 5: Composition of consum er unit: Average annual expenditures and characteris-
tics, Consumer Expenditure Survey, 2007” downloaded on December 1, 2008 from:
http://www.bls.gov/cex/tables.htm
18 More information about this difference can be found in California’s guidelines review report (Judicial Council,
2006).
19 Supra note 17.
20 Published reports from the 2007 CES ( supra note 15) indicate that households with $150,000 or more in
annual income have annual expenditur es that average: $6,708 on net ve hicle outlays; $7,064 on cash contri-
butions; $18,604 on pension and social security; and $9,008 on mortgage principal.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 16
¾ The van der Gaag (1981) estimates form the ba sis of about five state guidelines (i.e.,
California, Idaho, Nevada, New York, Wisconsin.) Most states that rely or have relied on
the van der Gaag estimates use a flat percentage of obligor gross income to compute the
child support obligation; that is, there is no consideration of the custodial parent’s in-
come. (California and Idaho are exceptions.)
¾ The Espenshade (1984) estimates forms the ba sis of about seven state guidelines (i.e.,
Alabama, Florida, Indiana, Kentucky, Maryland, Michigan, Virginia). With the exception of
Alabama and Michigan, these states have not updated their schedules since new studies
have been released. (Alabama recently adop ted a schedule based on Betson’s most re-
cent study, but it has not been promulgated. Michigan updates Espenshade’s estimates
annually for changes in the price level.)
¾ The Betson-Rothbarth (1990) es timates form the basis of ab out nine state guidelines.
Many states that updated their guidelines beginning in the mid-1990’s relied on these
estimates. West Virginia, a state neighb oring Maryland, relies of these estimates.
¾ The Betson-Rothbarth (2001) estimates form th e basis of about twelve state guidelines.
Many states that updated their guidelines at least twice since the mid-1990’s rely on the
second set of Betson-Rothbarth estimates. Maryland’s neighbors, Pennsylvania and the
District of Columbia, rely on these estimates.
¾ The Betson-Rothbarth (2006) estimates form the basis of about four state guidelines
(i.e., Arkansas, North Carolina, Oregon, and Rhode Island). Schedules based on these es-
timates have also been recommended or adop ted in a few other states (e.g., Alabama,
Pennsylvania, Louisiana) but have not yet been promulgated.
¾ The average of the Betson-Rothbarth and th e Betson-Engel (2001) estimates form the
basis of Georgia guidelines.
¾ Lino’s USDA estimates forms the basis of the Minnesota guidelines.21 Minnesota is the
only state to rely on the USDA estimates. Adopted in 2005, Minnesota adjusted them
because some economists believe that the USDA estimates overstate the child’s share
of housing expenses.
¾ Kansas bases its guidelines on per capita estimates of child-rearing expenditures that
are adjusted for routine parenting time (also, called the “dissolution factor”).

The above list accounts for the economic basis of about 39 state guidelines. In the remain-
ing states, the economic basis is unknown or th e basis is a combination of factors including
previous county guidelines amounts, guidelin es amounts in bordering states and other
factors.
State-Specific Data. We know of no state that uses stat e-specific data as the basis of its
guidelines formula.
22
Estimates for Single-Parent Families. We also know of no state th at relies on expenditures
in single-parent families as the basis of its guid elines formula. States that have considered
expenditures in single-parent families typica lly reject those estimates because they often
result in near-poverty amounts, are not availa ble for high incomes (because too few single-

21 We do not know the year of the USDA study that forms the basis of the Minnesota schedule.
22 Some states have attempted to estimate child-rearing costs for their state but have found the effort difficult
and expensive and have not used the stud y findings to develop their guidelines.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 17
parent families have high incomes), and are not consistent with the premise that the child
should share the standard of living that both or either parent can afford.

Switch from Espenshade-Engel to Betson-Rothbarth Estimates

About twenty states including Maryland origin ally based their guidelin es on the Espenshade
estimates. Many states that updated their guidelines after 1990 switched to the Rothbarth estimates for several reasons. First, after e xamining estimates using five different method-
ologies in his first study, Betson recommended its use for states guidelines. Other econo-
mists with expertise in child-rearing expend i t u r e s h a v e a l s o r e c o m m e n d e d i t ( B a r n o w
1994).
23 Finally, and arguably the most important reason, the use of the Rothbarth esti-
mates produced the least price-sticker shock to existing guidelines amounts. Although the
Rothbarth methodology produces lower estima tes than the Engel methodology, when ad-
justed for price increases and after-tax income increases resulting from federal tax reform, it
typically increases most guidelines amounts. (There are exceptions for some states at higher incomes.) In contrast, updating the schedule using the Betson-Engel estimates would result in inordinate increases in many states because the Betson-Engel estimates are much
more than the Espenshade-Engel estimates.

Use of Other Estimates
Several states have seriously considered th e Betson-Engel and USDA estimates when re-
viewing their guidelines formulas.

States that consider the Betson-Engel es timates, typically adapt the Lewin/ICF (1990)
framework for analyzing the appropriateness of their schedule amount by comparing their
guidelines amounts to the most recent Roth barth and Engel estimates. Since Lewin/ICF
found that the Rothbarth and Engel estimators formed the lower and upper bounds of credi-
ble estimates, they concluded that any amount in between these estimates is an appropri-ate guidelines amount.
24

Starting with this framework, one state, Ge orgia, eventually adopted a schedule based on
the average of the Betson-Engel and Betson-R othbarth estimates. The decision-makers
assumed that the average would be the closest to actual child-rearing expenditures. Fur-
ther, the average produced amounts more simi lar to the existing Georgia’s guidelines
amounts than the pure Betson-Rothb arth estimates did at the time.

23 Barnow (1994) wrote, “The Rothbarth estimator then emerges as the closest to a consensus estimator,
although, as noted above, it may well underestimate expend itures on children. States should bear in mind the
potential bias of the Rothbarth estimator, and they m ay wish to consider increasing the Rothbarth figures
slightly to account for the bias.
24 When using this approach in 1990, Lewin found that mo st state guidelines were within the range of the
estimates. Only eight state guidel ines contained amounts below the low estimate and no state guidelines
contained amounts above the high estimate. A subseque nt study (Venohr and Griffith 2005), in part, updated
the Lewin study. It found that that the number of st ate guidelines containing amou nts below the low estimate
increased to 22 states and included Maryland. They a ttributed this to states not updating their schedules.

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule

Page 18

Minnesota is the only state known to use the US DA estimates. Other state guidelines com-
mittees (e.g., Ohio) have recommended an upda ted schedule based on the USDA estimates
but the proposed legislation containing the USDA-based schedule was not passed.

Adjustments to the Bets on-Rothbarth Estimates

Several states have adjusted the Betson-Rothbarth estimates.
¾ Some states with above- or below-average income realign the Betson-Rothbarth esti-
mates, which are based on national data, to the income of their state. For example,
Connecticut, a high-income state, bases its sc hedule on the Betson-Rothbarth estimates
that were realigned to Connecticut by comparing national and Connecticut Census data on income distribution.
¾ The District of Columbia applies the Betson-R othbarth estimates to all after-tax income
including what an intact family would spen d on mortgage principal and “savings.”
¾ Rhode Island bases its schedule on the Bets on-Rothbarth estimates with a modest up-
ward adjustment to account for Rhode Is land’s relatively high housing costs.

Economic Estimates Considered by Maryland

CPR compared the existing Maryland schedule to several updated schedules that varied in
their underlying economic data. They included: ¾ A schedule based on the pure Betson-Rothbar th estimates similar to the Pennsylvania
schedule;
¾ A schedule based on the average of the Betson-Rothbarth and Betson-Engel estimates
similar to the Georgia schedule;
¾ A schedule based on realigned Betson-Rothb arth estimates similar to the Connecticut
schedule;
¾ A schedule based on the Betson-Rothbarth estimates applied to all after-tax income
similar to the District of Columbia schedule;
¾ A schedule based on the Betson-Rothbarth estimates with adjustment for Maryland’s
above average housing costs simila r to the Rhode Island schedule;
25 and,
¾ Other schedules that were modificat ions or combinations of the above.

In reviewing these schedules, the Committee fa vored using the Betson-Rothbarth estimates
with an adjustment to account for Maryland’s above average housing costs. Many states
rely on the Betson-Rothbarth estimates includ ing states bordering Maryland. Further,
economists have recommended the Rothbarth esti mator. An upward ad justment is justified
on two accounts: the Rothbarth estimator is kn own to understate actual child-rearing ex-
penditures; and, it is based on national data so it does not reflect Maryland’s above-average
housing costs.

25 Maryland housing costs are about 25 percent more than the national average. This is based on the differ-
ence in gross rent, which includes utilities, betwee n the nation and Maryland (2006 U.S. Census American
Community Survey).

Center for POLICY RESEARCH
2008 Update of the Maryland Child Support Schedule