Miles Corak is Director of Family and Labour Studies at Statistics Canada.
Editor’s Note This article is reprinted as it appeared in the 25th anniversary issue of Policy Options (March-April 2005), and is also available at <www.irpp.org>.
In thinking about the Canadian welfare state, about past accomplishments and future challenges, I would like to begin by borrowing a few pages from the work of Amartya Sen, the 1998 Nobel laureate in economics. Sen’s thinking is neatly summarized in his popular book Development as Freedom, which in the first instance is directed to issues in development economics, but has broader implications and lessons for social policy in rich as well as less rich countries.
Sen argues that the major goal and the major means of development should be freedom. Society should offer its citizens the freedom to lead the lives they choose to value.
This idea has in fact been an important undercurrent in the development of the Canadian welfare state, from its earliest days in the post Second World War era. Then, the emphasis was on the importance of full employment and the need for economic security for those excluded from jobs because of business cycle recessions, sickness, or the physical limitations of old age. But from the start there has also been an emphasis on active measures involving investments in human capital. First through more and more years of compulsory schooling; then to the broadening of access to higher education; and more recently to early childhood development.
Sen’s thinking asks policy makers to focus on the capabilities of citizens and to remove barriers to full participation in society, thereby allowing them to make use of their talents and develop their full potential. He frequently cites the quest for freedom from poverty, adequate health care, and education as examples of specific social policies affording citizens these capabilities.
I choose this as my starting point to draw what I think is an obvious link, though one not directly made by Sen or others, to the circumstances of children. There is in the case of this particular group, a connection between on the one hand freedoms and capabilities, and on the other hand direct measures of these concepts as well as policy levers. Freedom means that children can become all that they can be. In some large sense, this means that they are free to realize their full potential, and that their outcomes in life are not tied to their family backgrounds. In other words, for children this means equality of opportunity.
As an economist, I am inclined to think about these issues in terms of labour market outcomes, though obviously the issues are broader than just earnings and incomes. Canadians often pride themselves in having a society that is not subject to the extreme inequalities that are sometimes seen as characterizing our neighbour to the south. But in this essay, I will ask you to think not just about inequalities in the here and now, but also about how these inequalities play out across generations.
There is a sense in which a high level of income inequality need not necessarily reflect, in Sen’s words, an “unfreedom” and would in fact be tolerated by even the poorest in society. This would be the case if there were flux and mobility across the generations; if, in spite of current circumstances, the children of the poor were as likely to grow up to be highincome adults as the children of the rich. Similarly, the same high level of inequality in the here and now may have very different implications for social cohesion and individual welfare when there is very little generational income mobility.
The percentage increase in sons’ adult earnings for every percentage point increase in fathers’ earnings varies by a factor of more than two across the rich countries
Source: Miles Corak, “Do Poor Children Become Poor Adults? Lessons for Public Policy from a Cross Country Comparison of Generational Earnings Mobility,” 2004.
Indeed, historically, the possibility that life will be better for their children is the promise that North American societies have held out to many immigrants and that makes the great difficulties of leaving their home country and settling in a new land tolerable.
At first glance, Canadians have a good deal to celebrate when it comes to intergenerational mobility, especially when it is compared with those of other countries. One direct measure of generational income mobility is the strength of the link between an adult individual’s earnings and his or her parents’ earnings when they were raising their family. This tie will determine the income advantage, relative to the average family, that higher income parents pass on to their children. Or for that matter the income disadvantage that lower income parents pass on.
The percentage increase in a child’s adult income for every percentage point increase in his or her parental income is presented in figure 1 for a number of OECD countries. The higher this statistic, the lower the degree of generational mobility. These findings suggest, that first, there is a good deal of variation across the rich countries – by at least a factor of two – in the degree to which an earnings advantage is passed to children. Second, in no country is the inherited parental advantage much lower than one-fifth.
The United States, the United Kingdom, and to a slightly lesser extent France stand out as being the least generationally mobile societies, with every 10 percent increase in parental income implying that the grown-up child income will be 4-5 percent higher. At the other extreme are Denmark, Norway, Finland, and Canada where the relationship is more than half as weak.
For example, in Canada during the late 1970s, the market income of families in the top fifth of the income distribution was about 23/4 times as great as those in the bottom fifth. The 0.19 statistic in Figure 1 implies that the adult income of someone born to a family at the top would be about 20% higher than someone born to a family at the bottom. With a value in the order of the US or the UK this income advantage would have been 66%.
But what exactly does this statistic mean? The information in Figure 1 is purely descriptive and on its own tells us nothing about the reasons for the degree of generational mobility or for the differences among countries. Effective policy intervention requires not just an awareness of raw correlations, but also an understanding of causal processes. I would like to focus on two of several mechanisms that underpin these generational mobility figures and that open up distinct opportunities for the conduct of public policy. The first has to do with how labour markets work, and the second with the relative benefits of public policy.
More labour market inequality implies less generational mobility. One important determinant of the degree of earnings inequality is the return to higher education. Figure 2 shows that a higher return to university education is associated with tighter links between father and son outcomes. The figure presents a scatter plot between the estimates of generational income mobility from Figure 1, and the private rate of return to university education (relative to secondary education). For these countries there is a clear positive relationship. The three countries with rates of return higher than 10 percent – the United States at 18.9 percent, the United Kingdom at 18.1 percent, and France at 13.3 percent – are the countries with the least generational earnings mobility.
Generational Earnings Mobility is Lower the Higher the Return to University Education
Source: Miles Corak, “Do Poor Children Become Poor Adults? Lessons for Public Policy from a Cross Country Comparison of Generational Earnings Mobility,” 2004 and OECD, Education at a Glance, 2003.
There are a number of possible explanations for this pattern, but it may be that countries with higher rates of return also have very different structures and policies in place in terms of access to higher education, since one of the reasons for an elevated rate of return is the presence of restrictions in the supply of university graduates. In this sense it may also be that a higher rate of return reflects inequality of opportunity. And, further, higher rates of return do change the incentive for parents, leading those who value education most to invest even more in their children. This explanation focuses on the structure of rewards and private investments in children, but it also raises the question of differences in opportunity, i.e., the extent to which children from higher income families are more likely to capitalize on the rewards available to higher education. Thus, the second factor likely to explain cross country differences in generational mobility has to do with the opportunities children are afforded and the nature of public investment in them.
Increases in “progressive” investments – those of relatively more benefit to the less well off – will loosen the link between parent and child outcomes. Traditionally, this was seen as an important aspect of public schooling. Societies differ a good deal in their levels of public spending on education. This is illustrated in Figure 3, which plots the information from Figure 1 against the education expenditure per student.
More spending per student is generally associated with more generational mobility. However, the relationship between public spending on education and the degree of generational mobility is not as simple as that. For instance, the United States is a clear outlier, with the highest levels of public spending but also one of the strongest ties between fathers’ and children’s earnings. On the other hand, the United Kingdom and Finland both spend relatively low and similar amounts per student but are characterized by very different degrees of generational mobility. The important point we get from these results is that not only the overall level of public spending on education matters, but also how the money is spent. In other words, the way education systems are structured and how the cognitive capacities of children are developed to allow them to take advantage of whatever opportunities are made available is important.
Both the Level and Structure of Expenditure on Education Determine the Degree of Generational Mobility in Earnings
Source: Miles Corak, “Do Poor Children Become Poor Adults? Lessons for Public Policy from a Cross Country Comparison of Generational Earnings Mobility,” 2004 and OECD, Education at a Glance, 2003.
The best way to understand these patterns is to recognize the extent to which public investments are of relatively more benefit to the disadvantaged. One possible measure of this is how tightly related the abilities and skills of children are to the educational levels of their parents. For example, if literacy and numeracy skills in adulthood are strongly correlated with parental educational levels, this suggests that spending on education has done little to level out relative advantages and disadvantages that are based in the home and possibly that public investments have not been terribly “progressive.”
This information is available for seven of the countries under study, and the relationship with generational earnings mobility is illustrated in Figure 4.
In the countries under study there is a very strong positive relationship between generational earnings mobility and the gradient between the numeracy/literacy skills of adult children and the educational levels of their parents, with Germany being the only significant outlier.
This pattern reflects the inequality of private and public investment in children. If the degree of income inequality is higher in one country than another it might be reasonable to expect that not only will the rewards of a given level of investment be higher, but also that the level and distribution of investment in children will be different. A country with more income inequality might also have more inequality in the investment that rich and poor parents can make in their children, and hence a lower degree of generational income mobility.
But the extent to which this is so will also depend upon the degree to which children from less advantaged backgrounds disproportionately benefit from public programs. Under certain circumstances universal government programs can reduce the investment gap between rich and poor children. If the first dollar of investment creates the greatest increase in the well-being of the child, then when government invests equally in all children, poor children are likely to gain more than affluent children. The information in Figure 4, for example, suggests that the UK and the US get relatively little punch in terms of higher generational mobility for every dollar spent on education because the structure of their educational systems reinforces the relative advantages and disadvantages children get from their home environment rather than leveling them out. Figure 4 suggests that Canada stands at the other extreme with countries like Denmark and Norway. Indeed, the great promise of government investment in the expansion of universal education and increased access to higher education during the post-war period was that it would give children from relatively disadvantaged backgrounds an extra push and put them on a par with their fellow students Figure 4 suggests that there are still very large differences in the extent to which this has been done in the welfare states of the rich countries, and as such offers an important hint as to why countries differ in the degree to which economic advantage is passed on between parent and child.
In sum, the generational mobility of earnings in Canada is higher than other rich countries because first, labour market inequalities and the returns to education are relatively lower, and second, the mix of public and private investments in children has been relatively progressive. But this story reflects the situation of the last 30 years or so, a generation that was born in the 1960s, came of age and went to high school and university in the 1970s and 1980s, and found its place in the labour market of the 1990s. It cannot be uncritically transposed to today’s newborns and elementary school children, who will be attending the colleges and universities of the 2020s and working in the labour market of the 2030s.
Generational Earnings Mobility is Lower the Stronger the Gradient Between Children’s Cognitive Abilities and Their Parents’ Education Level
Source: Miles Corak, “Do Poor Children Become Poor Adults? Lessons for Public Policy from a Cross Country Comparison of Generational Earnings Mobility,” 2004 and Esping-Andersen, “Unequal Opportunites and the Mechanisms of Social Inheritance,” in Miles Corak (editor), Generational Income Mobility in North America and Europe, Cambridge: Cambridge University Press, 2004.
The future holds two challenges. The first has to do with access to education. A climate of higher tuition fees is on the horizon, and if the experience of the 1990s and the recently released Rae Review of post-secondary education in Ontario are any indication, we may also witness more decentralization in the way fees are set in the short run. Fees on average will likely continue to rise, but they will also vary a good deal more, perhaps between provinces, but also between institutions and fields of study. Thus, a much more differentiated postsecondary system is in the offing.
In this context there is a growing concern about access to university education. And while there may be a need for more public funding and for reconsideration of the structure of support to students, particularly to those from low-income backgrounds, the issues of access are more than just financial.
Access to higher education is often discussed in very broad terms, i.e., whether students are more or less likely to attend university according to their family income. In the coming years the significance of access to higher education may be more narrowly framed to refer to access to particular institutions or fields of study. Generational mobility will also be influenced by the nature of universities’ student selection criteria.
Recent Immigrants Living in Low Income Families, 1980, 1990, and 2000
If children from higher income families are more likely to have the skills to gain admittance to those areas most highly rewarded in the labour market, a rise in admission standards and the development of other selection criteria may lead to stronger links between family background and post-secondary participation in particular institutions or fields of study. In this sense it is important for policy-makers to understand the non-financial barriers to accessing higher education, particularly circumstances earlier in the lives of young people that help them continue their education after high school.
This is why concerns about generational mobility may lead policy-makers to focus increasingly on the family, particularly on its role in the development of children’s’ cognitive abilities. Cognitive and social skills are an important determinant of long-run earnings; they are only loosely correlated with educational attainment; and cognitive performance is more closely related to the “cultural capital,” or to put it more broadly, the “parenting style” of the family, than it is to its material wealth. In fact, the kinds of parental investments that are decisive are not the monetary kind. The inheritance of education, occupation and income is influenced in the first instance by the impact parents have on a child’s cognitive performance, and as Figure 4 illustrates, societies leveling the playing field with respect to these circumstances have had the most success in promoting generational mobility.
This is also why in Canada over one-third of children from low income backgrounds grow up to be low income adults, in spite of the loose overall tie between the incomes of children and parents. On average, Canadian society is very mobile across the generations, but a significant proportion of children fall through the cracks in this fabric.
Early childhood initiatives may play an important role in determining the degree of generational mobility. If this is the case, social policy may increasingly be called upon to equalize the impact families have on children’s skills, beliefs, and motivation. The impact of early childhood initiatives on generational mobility will depend upon the extent to which they prove to be effective and of relatively more benefit to children in disadvantaged families.
The second major challenge to generational mobility has to do with immigration.
The face of immigration has changed, and the experiences of the past are little guide for the future. In the 1960s, 75 percent of immigrants to Canada were from Europe and the US; during the 1970s this fell to 40 percent, and by the 1990s it was only 20 percent. More than 6 in 10 recent immigrants to Canada now arrive from countries in the Middle East and Asia, double the proportion during the 1970s.
While it should be noted that on average immigrants are more educated than the Canadian-born population, their command of English or French is at the same time not as high. This changing structure of immigration, along with the much changed economic climate that has put a higher premium on literacy than was the case in the past, has led to a marked decline in immigrants’ economic assimilation. In the early 1970s recent immigrants earned about 10 percent less than their Canadian-born counterparts, in the early 1980s it was about 30 percent, but in the early 1990s it was 60 percent, and it improved only slightly in the later part of the 1990s.
This situation has led to a marked increase in the percentage of immigrant families with low incomes, as illustrated in Figure 5. Between 1980 and 2000 this percentage fell slightly among Canadian-born families from 17 to 14 percent, but among recent immigrants it rose from 25 percent to just over 35 percent. Over one-third of recent immigrant families have low incomes.
In short, the ability of immigrant families to give their children the best possible start in life has worsened. This raises concerns for future generational mobility. However, the bright spot in this picture is the speed with which young children are able to adjust in the context of the Canadian elementary school system. The situation of children born to immigrant parents whose mother tongue is neither English nor French is the most remarkable in this respect. At the age of 4 and 5, when they start kindergarten, these children score 20 to 30 percent lower on a battery of reading, writing and math tests than their classmates, but by the age of 10 and 11, when they are ready to leave elementary school, their outcomes are no different and indeed in some cases exceed those of children whose parents are Canadian born and whose mother tongue is either English or French.
If we are to maintain the high degree of mobility experienced by previous generations, we must ensure that this performance is sustained over the coming decades and that it continues through high school.
Equality of opportunity implies that inequities of outcome are indefensible when they are due to differential circumstances, but societies and parents influence their children through a hierarchy of circumstances. If we are to understand what equality of opportunity means and how it can be influenced, we have to know what these circumstances are. We can imagine parents giving their children an advantage through social connections and financial support that facilitate access to education and jobs; family investments that influence skills; and finally aspects of family culture that influence beliefs and motivation.
These are the successively broader fields, each corresponding to a successively broader definition of equality of opportunity, that policy makers could seek to level. Equating equality of opportunity with complete generational mobility, with no statistical tie between parent and child earnings, implies that not only should the influence of social connections and investment be equalized, but so should the influence of family culture. In the words of the Yale political philosopher John Roemer, this is “a view that only a fraction of those who consider the issue would, upon reflection, endorse.”
This is a cautionary note: to eliminate entirely the income advantage that is passed from parents to children would require a degree of intervention into the lives of children and families that the majority in most societies would find untenable. The degree to which the parental income advantage passed on to children is consistent with equality of opportunity is not selfapparent. It requires a redefinition of the circumstances acceptable as sources of labour market success, an understanding of the effectiveness of policy interventions, and a recognition of the trade-offs between the gains in eliminating them and the losses in terms of other measures of welfare.
What is clear is that the postwar agenda of offering increased access to higher and higher levels of schooling seems to have reached its limit. If Canadian society wishes to continue to promote equality of opportunity, then we need to invest more in children earlier in their lives to ensure that they have the skills and the opportunities to succeed in the labour market. This shift in direction may involve a conception of equality of opportunity that has different degrees of support among Canadians, because it requires that public policy influences the impact that families have on their children’s skills, beliefs and motivation.
In this sense, whether the Canadian welfare state is able to offer effective programs of relatively more benefit to the least advantaged, in the context of a growing diversity, is the major challenge determining if the next generation will enjoy the same degree of opportunity, or to use Sen’s word – freedom – as the last generation.
The Emergence of Cross-Border Regions: Interim Report
The Canada-US Free Trade Agreement (FTA) and the North American Free Trade Agreement (NAFTA) opened a new era in Canada-US relations, not only in the area of economic relations but, more generally in the ways the two countries interact. The economic consequences have been extensive and largely positive, but another manifestation of their new relationship has been the growing extent of links of various kinds between the two countries – especially between adjacent and nearby areas along the border.
The PRI project on North American linkages has been studying the emergence of cross-border regions and the implications for the Government of Canada in terms of foreign, economic, and regional developmen policies.
The PRI is releasing project findings to-date in the Interim Report on the Emergence of Cross-Border Regions, as well as a working paper on a leadership survey on Canada-US Cross-Border Regions. A briefing note that provides an executive summary of these reports is also available.