Shadow Education: Theory, Analysis and Future Directions: A Rejoinder

Commentaries and replies on a given article, especially those appearing in the pages of an academic journal like Social Forces, cut to the very heart of intellectual exchange and area-specific development in the field. Such conversations and any debates embedded within them also expose analytic decision-making and potentially promising conceptual and empirical developments to current and future scholars. For these reasons, we welcome and appreciate the comments of Eric Grodsky and Sigal Alon on our article "Shadow Education, American Style: Test Preparation, the SAT and College Enrollment." Both provide insightful arguments and extensions that, in conjunction with our reply, should be of broad interest to stratification and education scholars.

Our article builds on "shadow education," a concept most often utilized in the comparative education literature. We extend its theoretical utility to the U.S. case, with a substantive and empirical focus on college test preparation. Our analyses show that background inequalities in family income and parental education shape the likelihood that students engage in various forms of SAT preparation, and that these "shadow education" activities have important implications for both test performance and selective college enrollment. For example, students from the most advantaged families are significantly more likely to enroll in private courses, such as those offered by Princeton Review and Kaplan — a strategy that corresponds to significant SAT score gains. Higher SAT scores, in turn, increase the chances of getting into the nation's most selective colleges and universities.

In their comments, Eric Grodsky takes issue with several important theoretical and methodological aspects of our article and Sigal Alon highlights key processes pertaining to race/ethnicity. Grodsky's and Alon's emphases on theoretical conception, measurement and process, when taken together, offer fertile ground for contemplation and future research. We respond accordingly here. Beginning with theoretical issues, we restate and stand by our definition of shadow education relative to the alternative, narrower definition that Grodsky prefers. Our measurement and modeling strategy follows directly from our theoretical discussion and captures — more effectively than Grodsky's dichotomous approach — the interconnected and overlapping nature of student test preparation. We then turn to Alon's response and analyses, which elaborate on our results in important ways with respect to racial/ethnic minorities and suggest differential processes in the course of educational mobility. We then highlight some avenues for future research on shadow education.

Theorizing Shadow Education

We define shadow education as "educational activities, such as tutoring and extra classes, occurring outside of the formal channels of an educational system that are 'designed to improve a student's chance of successfully moving through the allocation process.'"(436) This includes behaviors occurring outside of the formal school day with the intent of "mastering curriculum, examinations, and earning grades for learning and skills used by schools to grant students further educational opportunities" (Baker and LeTendre 2005:56). Like a shadow, such activities go relatively unnoticed and take the shape of formal schooling in both purpose and curricula (Southgate 2009).

Grodsky takes issue with our definition, overstating that it equates to "any non-compulsory practice that occurs outside of regularly scheduled classes" (Grodsky:478) or, even more simply, to "other activities undertaken by schools with public funds [such as] sports, debate, [and] school yearbook."(Grodsky:476) This is a misrepresentation of our definition, and an extension that undermines the theoretical and operational value of the shadow education concept. Moreover, Grodsky suggests that it "fails to conform to the key elements of the definition offered by Stevenson and Baker (1992)" in so much as "Shadow education... exists in the private as opposed to the public sphere [and] has economic costs that bar most disadvantaged families from participation."(476) He then attempts to show that the effects of private tutoring and courses on SAT scores are smaller than our estimates and not much different from the returns to more accessible forms of test preparation. There are two theoretically problematic aspects of his definition that are better accounted for within our framing. These pertain to the application of general theory to a specific context and the configurational...