Mike Bonner, Ph.D. has a doctorate in Educational Psychology, with a specialization in organizational development. In the past, he has served as a practicing school psychologist, a university faculty, a full-time stay-at-home dad. He now balances his work with his family, which is comprised of a full-time working wife and two children, ages 9 and 7. When he isn’t providing “child care” for his kids, he teaches college courses in psychology, and provides community psychology services to organizations to improve organizational and human functioning.

In Part 1 Friday, I explained how the Census’s assumptions about parenting roles do not accurately capture how families are making child care arrangements. In Part 2, I breakdown the details of the “Who’s Minding the Kids?” report, show how the Census’s methodology, and its continued defense of that methodology is flawed and explain one simple fix.

Child Care Purpose of the Report Defense

Let’s look at the details of the report, “Who’s Minding the Kids?” In the introduction, it states: “Deciding which child care arrangement to use has become an increasingly important family issue as maternal employment has become the norm, rather than the exception”(emphasis added). If maternal employment is becoming the norm, and decisions about child care arrangements are increasingly important, then it would seem to be also increasingly important to accurately portray such arrangements for families with two jobs, or one job and a stay-at-home parent. To equate father-provided care with grandparent, or sibling provided care — as something functionally different from mother-provided care — does not accurately portray the arrangement, or the social, economic, and policy consequences of it.

Table 1 of the report displays data for “Preschoolers in Types of Child Care Arrangements” (p. 2).  Fathers are only counted as providing care when the designated parent (i.e., mother) is working or is in school. If the mother is off work, and the father is still responsible for the care, that isn’t counted, rather it is assumed the mother is providing the care. Many at-home fathers can tell you this isn’t always the case. The data provided tells us that 15% of fathers provide care while their wife is working or in school. But we don’t know how many mothers provide care while their husband is working or in school. And we don’t know how many fathers provide care, even if their wife is off work. Why not provide an accurate understanding of these variations?


Table 2 of the report displays data for “Preschoolers in Types of Child Care Arrangements by Employment Status and Selected Characteristics of the Mother” (p. 3-4). When estimating child care arrangements for children “living with father” this is defined as the mother not being present in the home. However, when estimating child care arrangements for children “living with mother” there is no differentiation about fathers’ presence (see footnote 6: “Mother present in the household, father may or may not be present” [emphasis added].) It is as if once a mother is present, the presence of a father and his role in the care of children is irrelevant – not worthy of discriminating in a meaningful way. Again, this is NOT a critique of how to define parent roles, but how we are not accurately measuring child care use patterns in this data.

When the report provides discussion of aspects of Table 2 (see p. 5) it really demonstrates how muddled, and un-meaningful this data analysis is: “Many (my note, many is 20%) mothers were involved as care providers…even though they did not live with them (their children). Later, it states, “Family members were important sources of child care for many employed mothers. Fathers and grandparents were regular care providers….  Grandparents cared for 30%…while a slightly smaller percentage (emphasis added, the number is 25%) of fathers cared for preschoolers. HOWEVER, we can’t tell if that 25% of fathers are fathers who live with the mother, or do not live with the mother, so we lose some important information here. The number of fathers raising children solo is TOO SMALL (2%) for them to provide employment breakdowns, but it wasn’t too small to report that 20% of non-resident mothers are “many” who provide child care! Further, in subsequent charts, father-provided care disappears entirely into a category of “family”! See Figure 1 (p. 6) and Figure 2 (p. 7) for example.

The report provides information about the average hours children spend in child care;  in the spring of 2005 it was 32%. However, when this figure includes only fathers as child care arrangements, and mothers as not a child care arrangement, you don’t fully understand the extent to which parents balance parent-care versus non-parent care (be it institutional, non-family, i.e., nanny, or family, such as sibling and grandparent). Further, they confuse this issue by discussing average child care time for employed vs non-employed mothers. “For children of employed mothers, this (time in child care) included time spent with their mother while she was working and time with their father while their mother was working. (Wouldn’t it seem important to distinguish between a mom taking a kid to work as a form of child care, and having a father supervise the child while she is at work?) But then, they go on to say, “If time in parental care is excluded…..” but does this mean parental care, as defined in the report, meaning mothers only, or does it refer to the previous sentence which lumps working mother-provided care with father-provided care? So what do these percentages really represent? They make this same confusing double meaning statement again on page 9.

Consistent Data Defense

The U.S. Census Bureau defends the survey as needing to classify mothers as designated parents and fathers as child care providers in order to track changes in child care over time in response to the increase in working mothers. This mandate began in 1985. Society has continued to evolve since then, and as a result, it is time for the mandate of this survey to evolve with it.  To argue that the survey must maintain current definitions in order to keep comparisons with past surveys consistent is misleading and scientifically baseless.

Again, turning to “Who’s Minding the Kids” report, we see on page 9 a discussion of Historical Trends in the data: “Changes in the survey design over the period warrant caution when making comparisons between years.” (Yes that means there have been changes in how the survey was designed, administered or analyzed!) What kind of changes? In 1995 they expanded the number of options parents could select as child care categories; the data survey delivery was changed from paper to computer; the mode of data collection; the survey asked and presented categories and questions in a different manner; even the time of the year the survey is conducted has varied over the years. (This is important because the survey asks for child care arrangements of the previous month. It is reasonable to assume that for many families summer child care needs are different than fall, winter, or spring needs, and there may be important variations between fall, winter, and spring for some). The report goes on to correctly point out that “in addition to these survey design and seasonal issues, societal changes and the economic climate and stages of the business cycle during the time the surveys were conducted may also influence child care usage.” Hmmm. This sounds suspiciously like factors that might induce changes in father-provided care.  Let’s look at how other Census data collection and related instruments have dealt with methodological changes over time.

The decennial census (the 10 year population count) changes data collection and categorizing methods every time it is delivered. In fact, they design it each with intentional changes to test out the relative value of collecting data in different ways, or to prepare for desirable future comparisons. For example, there are numerous changes to how geographic data is categorized for each census (for an example, see: http://www.census.gov/geo/www/tiger/glossary.html). Racial designations, and how people are asked to select racial identity has changed over time (for examples, see: http://www.npr.org/2011/03/30/134974387/Race-Among-Hispanics, and http://www.census.gov/population/www/documentation/twps0076/twps0076.html) There are many other examples of this, acknowledged by the Census Burearu: “Because of the importance of this population count, procedural changes in the decennial census often reflect larger organizational shifts at the Census Bureau. …Political and technological changes, and the shifting public demand for information have all shaped the modern census and the mission of the Census Bureau.”

It is time for the “shifting public demand for information” to re-shape how the SIPP measure child care, and to distinguish parent care from non-parent care, and to include fathers as defined parent care.

The most recent SIPP presents a muddled picture of child care use, and then defends it based on statistical consistency, which doesn’t exist, and shouldn’t exist, if the Census is to fulfill their mission. The supplemental census surveys are meant to track characteristics of the population, and changes in those characteristics (such as child care use) over time. However, it is natural, and common, for such longitudinal surveys to make changes that allow the data collection to keep up with changes in society. A spokesman with the Census Bureau told Daddyshome, Inc., that:

“We collect data on a variety of child care arrangements to provide a full picture of the child care market and to help distinguish between the use of relative and nonrelative care.”

“The purpose of the SIPP child care topical module is to examine child care arrangements used primarily by working mothers – a valid subject of inquiry in what is fundamentally a financial and labor force activity survey.”

“It is not designed, nor intended, to make evaluative assessments of the level of ‘parenting’ activities engaged in by mothers and fathers.”

Each of these responses is problematic. Distinguishing between relative care and nonrelative care means paternal care is equal to grandparent and sibling care, but distinct from maternal care, based on current survey methodology. The need to examine child care arrangements used primarily by working mothers was a valid subject of inquiry for a fundamentally financial and labor force survey in 1985 and subsequent years, when working mothers were becoming more predominant. However, married couples are now making more joint decisions and arrangements about child care, and such reality no longer fits a 30 year old paradigm. Our criticism here is not seeking to make the SIPP child care module an evaluation of parenting activities. We are demanding a more contemporary and accurate accounting of the data on child care arrangements to better reflect a full picture of the contemporary child care market.

What Can Be Done?

A potential first step is to simply allow respondents to self-designate which parent assumes primary (i.e., 50.1% or greater) responsibility for the daily management and delivery of child care arrangements (to give the U.S. a more accurate picture of who is indeed “minding the kids”). Such data could then provide an additional benefit of distinguishing mother and father care from relative care and non-relative care (which gives a more accurate picture of the child care experience for the market place). Such a straight-forward change would also allow the data to be more in-line with the perspective of the child when assessing child care arrangements.

 Share and sign Daddyshome’s petition to the Census requesting recognition of dads as equal parents at www.dadsdontbabysit.com.