The Rug Rat RaceComments and DiscussionThe Rug Rat RaceComments and DiscussionGarey RameyValerie A. RameyBrookings Papers on Economic Activity, Spring 2010, pp. 129-176 (Article)Published by Brookings Institution PressDOI: 10.1353/eca.2010.0003For additional information about this article Access Provided by University of California, San Diego at 04/14/11 2:27PM GMThttp://muse.jhu.edu/journals/eca/summary/v2010/2010.1.ramey.html129GAREY RAMEYUniversity of California, San DiegoVALERIE A. RAMEYUniversity of California, San DiegoThe Rug Rat RaceABSTRACT After three decades of decline, the amount of time spent byparents on childcare in the United States began to rise dramatically in the mid-1990s. This increase was particularly pronounced among college-educatedparents. Less educated mothers increased their childcare time by over 4 hoursper week, and college-educated mothers increased theirs by over 9 hours perweek. Fathers showed the same patterns, but with smaller magnitudes. Whywould highly educated parents increase the time they allocate to childcare atthe same time that their returns from paid employment have skyrocketed?Finding no empirical support for standard explanations, such as selection orincome effects, we argue instead that increased competition for college admis-sions may be an important factor. We provide empirical support for our expla-nation with a comparison of trends between the United States and Canada,across ethnic groups in the United States, and across U.S. states.As time in paid work has increased over the last four decades, timespent on most home production activities has trended downward (see,for example, Robinson and Godbey 1999, Bianchi, Robinson, and Milkie2006, Aguiar and Hurst 2007). One notable exception, however, is timespent on childcare. Suzanne Bianchi (2000) and Liana Sayer, Bianchi, andJohn Robinson (2004) show that despite shrinking families, parents in thelate 1990s reported spending as much or more time on childcare than par-ents in earlier decades.In this paper we show that there has in fact been a dramatic increase intime spent on childcare. Linking 13 time-use surveys between 1965 and2008, we show that after declining for several decades, time spent perweek on childcare started increasing in the mid-1990s. The trends follow apronounced S-shaped pattern, rising markedly from the mid-1990s to theearly 2000s and then flattening out. Moreover, the increase in childcare12178-03a_Ramey-rev2.qxd 8/11/10 12:13 PM Page 129time has been twice as great for college-educated as for less educatedparents. This differential trend is particularly puzzling in view of thesharp increase in the average wages of college-educated individuals overmuch the same period. We also show that an important component of theincrease in childcare time was time spent on older children, and in partic-ular on coordinating and transporting them to their activities.Our estimates imply increases in average weekly hours of childcaretime ranging from 3 hours per week for less educated fathers to more than9 hours per week for mothers with a 4-year college degree. The implica-tions for the allocation of time are large by any metric. According to ourestimates, the time spent on childcare by the entire adult population in2008 is equal to almost 20 percent of the time spent on paid work. Theincrease in average weekly time spent on childcare during a 10-year periodfrom the early 1990s to the early 2000s was equal to 70 percent of theabsolute decline in work hours during the “Great Recession” that began inlate 2007. If those hours are valued at the market wage, the cost of theincrease in childcare time amounts to over $300 billion per year.The literature has offered several explanations both for why childcaretime has increased and for why it is greater among more educated parents(see, for example, Bianchi and others 2006, p. 87; Aguiar and Hurst 2007;Guryan, Hurst, and Kearney 2008). These explanations include selectioneffects, income effects, safety concerns, greater enjoyment of childcare, andmore flexible work schedules. We test each of these and find that they arenot consistent with the data.The inability of existing explanations to account for the evidence leadsus to offer a new explanation for the upward trends in childcare time. Weargue that much of the increase, particularly among college-educated par-ents, may be a response to an increase in the perceived return to attendinga good college, coupled with an increase in competition for college admis-sions. The size of college-bound cohorts rose dramatically beginning inthe mid-1990s, coincident with the increase in childcare time. John Boundand Sarah Turner (2007) provide evidence that these larger cohorts areassociated with increasingly severe cohort crowding at quality schools.The increased scarcity of college slots appears to have heightened rivalryamong parents, which takes the form of more hours spent on collegepreparatory activities. In other words, the rise in childcare time resultedfrom a “rug rat race” for admission to good colleges.To clarify the mechanics of this explanation, we develop a simple theo-retical model in which college admission depends on parents’ choice oftime spent preparing their children for college. College-educated parents130 Brookings Papers on Economic Activity, Spring 201012178-03a_Ramey-rev2.qxd 8/11/10 12:13 PM Page 130are assumed to have a comparative advantage in preparation time. Whenslots at good colleges are relatively plentiful, the marginal slots are filledby children of less educated parents. Competition among these parentsthen determines the preparation required for admission. When good slotsbecome relatively scarce, rivalry for the marginal slots shifts to the college-educated parents, who are better able to compete. A rug rat race emergesamong these parents, driving up both admissions requirements and thetime spent on childcare.We provide support for this explanation using three comparisons. First,we compare childcare trends in the United States with those in Canada.The two countries are similar along many dimensions but differ in onerespect that is key to our explanation: the Canadian higher education sys-tem lacks a steep prestige hierarchy, so that Canadians do not experiencethe same intense rivalry to gain admission into higher-rated colleges. Thusour theory predicts that time spent in childcare by more educated Canadiansshould