Credential inflation and educational strategies: A comparison of the United States and the Netherlands☆
Introduction
This paper asks the question to what extent the labor market value of educational qualifications affects educational decision making, and whether this influence varies according to the parents’ education. Based on theories on educational decision making developed in the past years, it is argued that children take their parents’ social class as a reference of own aspirations, and that their primary goal is to avoid downward mobility (Boudon, 1974, Breen and Goldthorpe, 1997, Goldthorpe, 2000, Chapter 11; Morgan, 2005). To the extent that this is the case, children's educational choices will depend on whether education is a ‘positional good’. According to an important group of mechanisms explaining the effect of education on labor market outcomes, education does not produce productive skills but rather signals attributes such as talent, trainability or motivation (Arrow, 1973, Hirsch, 1977, Spence, 1973, Thurow, 1976). These approaches share that education functions on the labor market as a positional good; not primarily the skills that are learned in schools, but rather the relative position of holders of a particular qualification level on the labor market determines the value of qualifications. If education functions as a positional good, children's educational decision making in order to avoid downward mobility assumes that they would need more education than their parents if qualifications have lost part of their value. Only by investing more in education if qualifications lose value, children are able to maintain their parents’ class position.
There are good grounds to expect that countries vary in the extent to which education functions as a positional good on the labor market, and thus in the extent to which children's educational decision making is affected by the relative value of educational qualifications in comparison to what they were worth for their parents. To understand these cross-national differences, we must first have a look at an important alternative theory of why education pays off on the labor market: human capital theory. Human capital theory assumes that people learn productive skills in school, and that employers are rewarding these skills because of the higher marginal productivity of skilled workers relative to unskilled workers.
One of the most important cross-national differences in educational systems is precisely this skills-producing function of schooling. In some countries, such as Germany and the Netherlands, educational systems are highly vocationally specific, aimed at the production of skills that are relevant for work and are rewarded by employers as such (Culpepper and Finegold, 1999, Shavit and Müller, 1998, Thelen, 2004). Given the large vocational orientation of the secondary and tertiary schooling system, it can be expected that education functions more in line with human capital theory in these countries relative to countries of which the educational system is far less vocationally oriented. One example of such an educational system that is far less vocationally oriented is the American system. Instead, the American educational system is largely hierarchical and linear, with less institutionalized tracking at secondary and tertiary levels, and fewer institutional linkages between schools and employers (Culpepper and Finegold, 1999, Lucas, 1999, Müller and Gangl, 2003, Thelen, 2004). Such a system would make employers far less likely to judge qualifications on the basis of the skills that have been acquired in school, and more likely to judge educational qualifications on the basis of their relative value. Such a linear schooling system makes families more likely to adjust their educational decision making strategies on this positional basis.
This paper studies the impact of credential inflation on educational decision making for two countries: the Netherlands and the United States. For both countries large combined datasets are available that allow the investigation of trends in the impact of education on occupational social class attainment for birth cohorts from the end of the 19th century up to the 1980s, as well as the impact of intergenerational credential inflation for birth cohorts who have parents born as far back as the 1880s.
Section snippets
Credential inflation and educational decision making
Recent theories of educational inequality state that children take the social position of the parents as a reference for own aspirations. The ‘primary goal’ of individuals is to avoid social demotion (class maintenance); only in secondary instance people strive for upward mobility (Goldthorpe, 1996, Goldthorpe, 2000). The decision to continue at a certain branching point in the educational career or to leave the educational system to enter the labor and/or marriage market is thus dependent on
Schooling in the American and Dutch labor markets
It is likely, but thus far untested, that countries vary in the usefulness of a mechanism to explain why schooling affects labor market outcomes. Thus far economists have paid much attention to comparing mechanisms for ‘the education effect’ largely by comparing human capital from positional good (or ‘screening’) explanations (e.g. Groot and Oosterbeek, 1994, Jaeger and Page, 1996, Layard and Psacharopoulos, 1974, Weiss, 1995). However, these studies have not aimed at explaining cross-national
Hypotheses
Based on aforementioned theories on the positional character of schooling on the labor market, where not primarily the skills that are learned are important, but rather the relative position in comparison to other job seekers, it is likely that people's educational decision making is responsive to changing values of qualifications across generations. More specifically, it is expected that a decreasing labor market value across generations of having made a particular educational transition makes
Data
Use is made of a great number of surveys held in the past decades, for both countries. For the United States, I use the cumulative General Social Surveys of 1972–2004. For the Netherlands I employ the collection of national surveys brought together by Ganzeboom and Luijkx (2004).
These datasets are used for two different types of analysis. First, using data for men only, we detect the changing value of having made three educational transitions across cohorts by running ordered logit models on
Operationalizations
Social class is used to detect the trend in the effect of education (hence social class as ‘destination’ in the mobility literature), as well as to measure the impact of parents’ class on transition probabilities. We look at class destination to observe trends in the impact of education for two main reasons. First, the theory of relative risk aversion underlying our positional goods hypothesis is developed to explain rather stable levels of class mobility (Breen and Goldthorpe, 1997,
Results 1: trends in the impact of transitions on class attainment
We estimate ‘returns’ to having made a particular educational transition in terms of class attainment, using the following ordered logit regression model for each transition j separately, in a similar manner as in Van de Werfhorst and Andersen (2005) and Van de Werfhorst (2005):where TRANSj is a dummy variable indicating whether the individual has at least completed transition j, COH is a variable of K cohorts, of which K − 1 have
Results 2: credential inflation and educational decision making
Now we turn to the second part of our analysis, where educational transition models are estimated with the IVE ratio as independent variable. We analyze the three transitions separately. For each transition we analyze three models: one baseline model including gender, father's social class, father's years of education, and the dummy indicating whether the father has completed the transition under study. This is a model similar to the model by Mare and Chang (2006), which tests for
Summary and conclusions
In this paper we analyzed the trend in the effect of education on social class attainment, and used this information to test hypotheses on the impact of credential inflation on educational decision making in the Netherlands and the United States. It was shown that, with few exceptions, the completion of three educational transitions became less influential for class attainment across birth cohorts between the 1880s and 1980s. Across two generations each transition has lost value, except for
References (43)
Higher education as a filter
Journal of Public Economics
(1973)Education, opportunity, and social inequality
(1974)Social mechanisms without black boxes
- et al.
Explaining educational differentials: Towards a formal rational action theory
Rationality & Society
(1997) - et al.
Analyzing educational careers: A multinomial transition model
American Sociological Review
(2000) - et al.
Testing the Breen–Goldthorpe model of educational decision making
- et al.
The German skills machine: Sustaining comparative advantage in a global economy
(1999) - et al.
The relative risk aversion hypothesis of educational choice
Journal of Population Economics
(2002) Cultural capital, gender, and school success: The role of habitus
Sociology of Education
(2002)
The constant flux: A study of class mobility in industrial societies
Restructuring the employment relationship
Recent trends in intergenerational occupational class reproduction in the Netherlands, 1970–99
Class analysis and the reorientation of class theory: The case of persisting differentials in educational attainment
British Journal of Sociology
On sociology: Numbers, narratives, and the integration of research and theory
The economic basis of social class
Earnings effects of different components of schooling; human capital versus screening
The Review of Economics and Statistics
Social limits to growth
Degrees matter: New evidence on sheepskin effects in the returns to education
The Review of Economics and Statistics
The screening hypothesis and the returns to education
The Journal of Political Economy
Cited by (0)
- ☆
Earlier versions of this paper were presented to a conference on ‘The Causes and Consequences of Educational Expansion’, Monte Verita, Ascona, 7–13 July 2007, at the European Science Foundation conference ‘The Transfer of Resources across Generations: Family, Income, Human Capital and Children's Wellbeing’, Vadstena, Sweden, 9–13 June 2008, and at the Summer meeting 2008 of the Research Committee 28 on Social Stratification and Mobility of the International Sociological Association, Stanford University, 6–9 August 2008. Participants to those meetings are thanked for their comments. The Netherlands’ Social Mobility File has been made available by Harry Ganzeboom and Ruud Luijkx, for which I am very grateful.