Miles Corak is a professor of economics at the University of Ottawa. He has published numerous articles on topics dealing with child poverty, access to university education, intergenerational earnings and education mobility, and unemployment. He has also contributed to two RSF volumes on economic mobilityFrom Parents to Children (2012) and Persistence, Privilege and Parenting (2011).
Q: In your chapter in Persistence, Privilege, and Parenting, you compare the effects of family background on mobility outcomes in Canada and the United States. First, why should people interested in inequality and mobility look in particular at the different results in these two countries? And second: is the ‘American Dream’ more achievable in Canada?
A: Perhaps the best way to answer this question is to consider what we mean by this phrase.
The "American Dream" is a very important metaphor in the United States, indeed some would say a defining metaphor. But it has a host of meanings, and our analysis speaks to the idea that economic outcomes should be the result of the energies, talents and motivations of individuals than of their family background. The American Dream is about becoming all that you can be regardless of your starting point in life.
In fact, we used similarly designed public opinion polls in the two countries to ask representative samples of Americans and Canadians what they understood the American Dream to mean, and surprisingly they answered in very much the same way, indeed in some scenarios we put to them, exactly in the same way. Both Americans and Canadians value freedom, being able to accomplish anything that you want, and to have the basic opportunities like health and education to do so.
Canadians differ from Americans in that they have a stronger tendency to see government policy as just another tool to help them achieve their individual goals. I don't want to overstate this tendency, but it is clear that Americans have a stronger tendency to see government as hindering rather than helping them, and that this could reflect a belief about the effectiveness of public policy as much as a principled stand against it.
There is value in comparing these two countries precisely because Americans and Canadians share some important values and goals, and because labour markets function in broadly similar ways on both sides of the border. But also because attitudes to public policy and the actual design of policy differ. These are differences that we potentially can all learn from, and that might be transferable across the countries.
To the extent that we measure the American Dream by the degree of relative earnings mobility across the generations --- by how closely tied a child's adult position in the earnings distribution is to the position occupied by his or her parents --- then on average it is pretty clear that there is more mobility, up to three times more mobility, in Canada than in the United States.
But the major point of our chapter was that this difference reflects differences in mobility at the two extremes of the earnings distribution. In fact, the broad majority of Canadians and Americans born to parents in the broad middle of the earnings distribution experience considerable mobility both upward and downward. But Americans born to parents at the top and the bottom are more likely than Canadians to become, respectively, the top and bottom earners of the next generation. It is in this sense that some might suggest that the American Dream is more of a reality north of the 49th parallel.
Q: In your second chapter in From Parents to Children, you and two Danish co-authors looked at how many children ended up with the same employer as their fathers. You found a surprisingly high proportion of young men in both countries fit that description – about 30 to 40 percent. But is this happening all along the income distribution? Are all young men – rich and poor – benefiting from the connections of their fathers or employer preferences?
A: The major contribution of this chapter is to underscore the importance of recognizing that child outcomes are related not just to the quality of the early years, but also to the structure of labor markets, and the resources parents have—though information, networks, or direct control of the hiring process—to influence the final transition children make in becoming self-sufficient and successful adults.
As you note our major finding is that up to 40% of young men have at some point worked for an employer that also employed their father.
This may appear to be a rather high level, but it may not be out of line with some of the basic facts of how young people find jobs. Families and friends are often cited as the most important source of information for new jobs. This is as true in Canadian and American labour markets as it is in European.
But our second major finding is that the transmission of employers between fathers and sons is greater, the greater the father’s earnings, and rising distinctly and sharply for top earners. The patterns are strikingly similar in both Denmark and Canada.
Most notable is a very sharp rise that sons whose fathers were in the top 10% had worked in the same firm. In fact, if the father’s earnings placed him in the top 1%, the majority of sons—indeed almost 7 out of 10 in Canada—had worked for an employer at which the father had also worked.
While the intergenerational transmission of employers is a common feature regardless of the father's earnings, it is distinctly much more common when fathers are top earners.
Q: It’s entirely possible that sons who work at their father’s firms do not climb the income ladder. How much does having the same employer as your father matter in terms of intergenerational mobility for those at the bottom and top of the income distributions?
A: The transmission of employers across the generations is a force for more stickiness in generational mobility, but this is almost exclusively because of what is happening at the top of the distribution.
In both countries sons born to fathers in the bottom 25% of the earnings distribution have about a 30% chance of ending up in the bottom 25% as adults, and about a 15% chance of rising to the top 25%. But the fraction working at the same employer as their father is the same whether they stay stuck in the bottom or move to the top, about 2 to 4%.
At the same time sons born to fathers in the top 25% also have similar rates of mobility in both countries, over a third staying in the top 25% as adults, and about a fifth falling to the bottom. However, the preservation of high income status across the generations is distinctly related to the tendency to be working in the same firm as the father: sons of high earning fathers are much more likely to be working at the same firm if they grew up to be high earners, than if they fell to the bottom.
Q: Are you aware of any research that has looked at this question in the United States or other OECD countries?
A: The findings are novel because our use of administrative data allows us to not only examine very large numbers of individuals, but also to link their employment outcomes to their father earnings and employers. There is a well-established literature in both economics and sociology stressing that family and friends are the major source of information about jobs and employers. Often used labor market surveys like those conducted by the Bureau of Labor Statistics show that asking family and friends about job contacts is one of the most common job search strategies, but the sample sizes and scope for intergenerational linkages are limited.
This said, Linda Dachter Loury recently published an interesting paper using US data showing the impact on wages of having found a job from a "prior generation male relative", and there is also a very interesting paper using Swedish administrative data documenting the extent to which young people get their first job in the very same plant as a parent. Though this research does not make the link to the parent's place in the earnings distribution, all of the findings are consistent with our research.
The kinds of questions social scientists can ask is often limited by the available data, and the scope for conducting more analysis of the kind we undertook would be greatly enhanced if a couple of questions were added to long standing cross-sectional and longitudinal surveys: "Did you ever work at a firm that also employed your father or mother?"; "Are you currently working at a firm that employed one of your father or mother?" This is something I would very much encourage the statistical agencies in other OECD countries to consider, particularly in the United States and the United Kingdom where inequality is much greater than Denmark and Canada.