We show that events in previous generations can explain contemporaneous shifts in intergenerational mobility. We first study the dynamic response of income mobility to structural changes in a model of intergenerational transmission. Mobility today depends on past policies and institutions, such that major reforms may generate long-lasting mobility trends over multiple generations. These trends are often non-monotonic, as mobility tends to be highest when a structural change occurs. Times of change are thus times of high mobility, while declining mobility today may reflect past gains rather than a recent deterioration of “equality of opportunity”. We then exploit data over three generations and a compulsory school reform in Sweden to test the dynamic implications of our model. The reform had a large, long-lasting, and non-monotonic effect: it reduced the transmission of disparities in income and education from parents to their offspring in the directly affected generation, but increased intergenerational persistence in the next. ⇤Stockholm University, Swedish Institute for Social Research (
[email protected]) †University College London, Centre for Research and Analysis of Migration, and IZA (
[email protected]) ‡Financial support from the Swedish Council of Working Life (FAS), the German National Academic Foundation, and the Centre for Research and Analysis of Migration is gratefully acknowledged. We especially thank our thesis advisors Anders Bjorklund, Christian Dustmann, Markus Jantti and Uta Schonberg for comments and advice. We also received helpful input from Robert Erikson, David Green, Raquel Fernandez, Raffaella Giacomini, Helena Holmlund, Stephen Jenkins, Mikael Lindahl, Matthew Lindquist, Steve Machin, Magne Mogstad, Lars Nesheim, Marieke Schnabel, Gary Solon, and seminar participants at University College London, the 2013 RES Conference at Royal Holloway, the Workshop on Intergenerational Mobility at the University of Copenhagen in June 2013, the Swedish Institute for Social Research, and the 2013 EALE Conference at the University of Torino. Introduction The evolution of inequality in economic status over time is a fundamental topic in the social sciences and in public debate. Two central dimensions of interest are the extent of crosssectional inequality between individuals and its persistence across generations, as status differences are transmitted from parents to their children. Both have important implications for individual welfare and the functioning of political and economic systems.1 A significant rise in cross-sectional income inequality from the late 1970s in the US, UK and (more recently) other OECD countries is well documented, but much less is known about trends in intergenerational mobility.2 Yet, we do know that income mobility differs substantially across countries, and the observation that those differences appear negatively correlated with cross-sectional inequality has received much attention.3 A central theme in the recent literature is thus if income inequality has not only increased, but also become more persistent across generations. This question is debated particularly in countries that experienced rising cross-sectional inequality, such as the US, where commentators argue that low mobility threatens social cohesion and the notion of “American exceptionalism”.4 But how should evidence on declining mobility be interpreted – does it reflect a diminished effectiveness of current policies and institutions in the promotion of “equal opportunities”? In this paper we show theoretically and empirically that mobility trends may instead be caused by events in a more distant past, as structural changes affect mobility over multiple generations. We argue that such dynamic responses are of particular importance in the study of intergenerational persistence, since even a single transmission step – one generation – corresponds to a very long time period. Institutional reforms or other systemic changes generate therefore long-lasting mobility trends. The interpretation of such trends necessitates a dynamic perspective, but existing theoretical work focuses instead on the relationship between causal mechanisms and the implied long-run or steady-state level of intergenerational mobility. We thus contribute to the literature by examining the dynamic implications of a simultaneous equations model of inter1Intergenerational mobility is for example seen to contribute to the stability of liberal democracies, by legitimating income and status inequalities and by reducing the potential for class-based collective action (see Erikson and Goldthorpe, 1992). 2Autor and Katz (1999) discuss trends in wage inequality across countries. Atkinson, Piketty, and Saez (2011) find a substantial rise in top income shares in the US and various other countries. 3A large empirical literature (see Solon, 1999, and Black and Devereux, 2011) seeks to quantify how intergenerational mobility differs across countries, groups and time. Bjorklund and Jantti (2009), Blanden (2011), and Corak (2013) present evidence on the correlation between cross-sectional inequality and mobility. 4Exemplary articles are “Ever Higher Society, Ever Harder to Ascend” in The Economist (Dec. 2004), “Moving Up: Challenges to the American Dream” in the Wall Street Journal (May 2005), “The Mobility Myth” in The New Republic (Feb. 2012), or the recent “Great Divide” series on nytimes.com. The political importance of the topic is exemplified by a speech of Alan Krueger, Chairman of the Council of Economic Advisers, who warned that intergenerational mobility should be expected to decline further as of the recent rise in income inequality in the US (speech at the Center for American Progress, January 12th, 2012).