by Andrew Lynn
American schooling attempts to be meritocratic. It’s not like Britain where a small number of private schools effectively train up an elite and feed them into the great universities. Nor does it have much in common with the European system, which channels a small percentage of students into academic secondary schools and then university while sending the rest to vocational colleges. What America attempts to do is to provide its young people with a general preparatory education, allowing students from all tracks to continue into higher education. The American school system has been designed to open doors rather than to close them.
But what actually happens to students as they go through the system? To find out the answer to that, you need data such as that provided by the National Center for Education Statistics (NCES), which has carried out a number of ‘longitudinal studies’, each following a defined cohort of students over a sustained period of time. One of them – the ‘High School and Beyond’ study – followed a national sample of 1980 sophomores, checking up on them every two years through 1986 and then one more time in 1992.3
The egalitarian bent of the U.S. system makes it harder to detect any initial ‘head start’. You can’t just look and see whether an individual has been to a private school, for example, or an academic as opposed to vocational college. All the same, there is one feature that stands out: the nature of the curriculum. That’s because American students are still tracked: they are streamed on the basis of ability or curricular tracking, which continues, effectively, from high school into higher education. In twelfth grade, for example, some students get the benefit of courses in advanced algebra, geometry, trigonometry, chemistry, physics, and two or more years of foreign language taken from the ninth to twelfth grade. At the other extreme, some students get none of the maths courses, none of the science courses, and no foreign language courses. The rest of the students get the benefit of some of these courses but not all.
For convenience, let’s call these three broad categories of students Group 1 (the more privileged), Group 2 (the moderately privileged), and Group 3 (the less privileged). Given that we know where our students have started out, it’s now just a matter of checking in on them at intervals to see how they have progressed. In fact, we can do more than that: we can draw career lines that describe the students’ passage through the educational system to college and beyond.
What we might expect would be a more or less random smattering of career lines. For example, we might expect some students to perform uniformly well and some students to perform uniformly badly. We would also expect some students to ace high school and flunk college, and some students to underperform at high school but then ace college. We would even expect some students to ace high school, flunk college, and then make a surprise recovery in later years. And we would of course expect to see some students – the ‘B’ students – performing at consistent if unexceptional levels throughout. In short, we would expect to see the full range of human variability.
That’s not the way it turned out. Most of the cohort followed, in the words of one of the authors of the study, ‘a few well-traveled career lines’. Out of 192 theoretically possible career lines (and 183 career lines that were actually possible), fully 52.5 percent of the students followed just thirteen of them. What’s more, fully 29.2 percent followed just four career lines. Although nearly every career line was, in fact, followed by somebody or other, only a very small number of career lines were followed by a substantial proportion of students.
There is clearly a pattern here. So what is it? We need to look more closely at those four career lines that seem to draw such a disproportionate number of America’s youth. They are these:
Student Trajectory Tenth Grade Twelfth Grade Four Years Later Ten Years Later
‘Persistent High Road’ High curriculum High curriculum Four-year college Further education in for-profit institution
‘Short High Road’ High curriculum High curriculum Four-year college [No further higher education]
‘Consistent Low Road’ Low curriculum Low curriculum No postsecondary No postsecondary
‘Dropout’ Low curriculum Dropout No postsecondary No postsecondary
What’s striking about these four most commonly travelled career paths is that there’s an obvious bifurcation into two ‘roads’: there’s the ‘high road’ of early demanding curricula followed by further demanding curricula and then entry into a four-year degree programme – and there’s the ‘low road’ of undemanding curricula followed by further undemanding curricula or a straightforward dropout from education altogether. If you expand these categories to include those whose paths were primarily ‘high road’ or primarily ‘low road’ (rather than exclusively so), then you find that 42 percent of the entire cohort have been channeled in one direction or another. There are far fewer middling performers than we might expect – and also far fewer erratic performers.
Alan Kerckhoff and Elizabeth Glennie, the authors of the study, see this as being a classic example of what they call ‘cumulative advantage’. Children in the twelfth grade who get the benefit of demanding curricula are not only likely to do better in the twelfth-grade tests of maths and English. They are also more likely to get into a better position (a four-year or private college, say) four years and even ten years out. Advantage heaps upon advantage: a good start gives the children a capacity to do well which carries them through to the next opportunity, which gives them a further capacity to do well – and so on. All this occurs even when factors like social background and prior academic achievements are taken out of the equation. It’s not a result of family privilege or even innate ability – it’s a purely positional effect.
What’s more, the effect can be quantified. Imagine each child in the cohort is in a ‘hierarchy of achievement’, which can be represented by percentiles: the middling child is at the fiftieth percentile, for example, the truly exceptional child is at the ninety-ninth percentile, and the underperforming child is down at around the twentieth percentile. Once you’ve got these starting percentiles, you’re able to calculate the ‘deflection’ caused to each child’s percentile rank in the hierarchy by these positional effects. The scale of the deflections will reflect the scale of any cumulative advantage.
The amount of deflection between the tenth and the twelfth grade is noticeable but modest: 10 percent of the cohort were deflected downwards by at least six percentile points and 10 percent were deflected upwards by at least eight percentile points. Look at the deflection at ten years out and it’s a different matter. More than 10 percent had been deflected downwards by twenty percentile points or so – and more than 10 percent had been deflected upwards by more than twenty-five percentile points. At this ten-year point, the child who had benefited the most was deflected upwards 27.9 percentile points, whereas the child who lost the most was deflected downwards by 24.2 percentile points – a difference of more than fifty percentile points. That means that the effects resulting simply from a difference in curriculum at tenth grade are enough to propel an average child into the top quartile or, likewise, send the same average child spinning down into the bottom quartile.
These figures are just what we would expect if cumulative advantage were at work. Relatively small initial advantages – studying algebra and geometry at twelfth grade, say – snowball into frankly astonishing differences ten years down the line. And the longer the period we look at, the greater the dispersal appears to be.
Cumulative Advantage
The theory of cumulative advantage attempts to answer a big and important question: why is it that people who start out with more or less the same skills and qualifications end up with such disparate outcomes? Of two young science PhDs, why should one end up a Nobel laureate and the other a teacher in a third-tier school? Of two young lawyers, why should one end up on the Supreme Court bench and the other chasing ambulances? Or of two young actors, why should one end up on the big screen and the other waiting tables? In each case: was any initial difference in ability
or potential really so great as to justify the inequality in outcome?
This is the question that was addressed by a man called Robert Merton of Columbia University. Merton’s own career had been little short of stellar: born as Meyer R. Schkolnick, the child of East European Jewish immigrants in working-class Philadelphia, by the end of his life he had been granted honorary degrees by no fewer than twenty universities, including Harvard, Yale, Columbia, Chicago, and Oxford. Merton’s immediate interest was the divergence of outcomes in the scientific world, and he looked to interviews with Nobel laureates – as well as diaries, letters, notebooks, scientific papers, and biographies – to find the answer.
What interested Merton in the first instance was something potentially very trivial: he noticed that when scientists collaborated – as they often do – it would be the better-known scientists who would get the credit: ‘You usually notice the name that you’re familiar with,’ said one laureate. ‘Even if it’s last, it will be the one that sticks.’ That doesn’t seem surprising until you follow through on the consequences, as Merton did. Say Scientist A and Scientist B collaborate on a project. Scientist A is more famous than Scientist B. The project is a great success. Scientist A gets the credit – after all, Scientist A must have done the bulk of the work, we think, being the more eminent – and becomes even more famous. Merton’s contribution was to show how this sets in motion a positive feedback effect: Scientist A just keeps getting more famous – and keeps getting more credit. And so on.
Cumulative advantage describes a similar positive feedback process but goes beyond distribution of credit to distribution of resources and opportunities. Run through the process again – only substitute ‘resources’ for ‘credit’ – and see where it gets to this time. Scientist A is identified as having star potential as a mere undergraduate; he is picked up by an Ivy League graduate school; with the support of eminent faculty, his graduate thesis garners attention from the research community; he gets a fellowship, fully funded, and a well-equipped lab; more good research; and onwards and upwards. Scientist B, on the other hand, may have had a touch of star potential but it was never well recognized. He gets to a middle-ranking graduate school where he can do solid if unexceptional work supported by solid if unexceptional colleagues. There’s funding but not a great deal of it; he has to teach a lot of hours, which diminishes the amount of time he has for research. His work attracts little attention; he gets little in the way of new funds and little in the way of promotion; which gives him little opportunity to produce the kind of work that would attract attention. And so on.
Accumulation of advantage has been identified in a broad range of activities. Careers within corporations follow a path-dependent cumulative advantage process: early success in the competition for promotion greatly increases the prospects of further promotion; early failure more or less rules it out. Many criminals have, in their early years, received ‘negative pushes’, which have triggered cumulative disadvantage processes through labeling and harsh punishment. The same mechanisms have been shown to structure the careers of film stars, Protestant ministers, and even Communist Party members.4
The disheartening thing about all this is that it tends to extract the living, breathing human being from the picture. We’re at the mercy of a cold and impersonal process that will propel us forward towards a future of continually accruing advantage – or it won’t. Even more disheartening, on the face of it, is that what determines whether we get ‘locked in’ to such a mechanism will, at best, be some early demonstration of promise; at worst it will be nothing more than the luck of the draw.
Or is that an oversimplification?
Harriet Zuckerman, who has studied the phenomenon in Nobel Prize–winning scientists, thinks that it may be. While cumulative advantage might be able to give people a head start irrespective of merit or hard work, that’s not the only way it works. It’s true that inequalities might unfairly open up solely on the basis of a whole range of ‘functionally irrelevant’ criteria such as race, sex, religion, or social class, says Zuckerman – but they won’t open up fast. In fact, it’s precisely when resources are allocated on strictly fair criteria – notably, competence for the job or task in question – that cumulative advantage really takes off. ‘Disparities in performance grow especially rapidly,’ says Zuckerman, when ‘the advantaged have been selected solely on criteria relevant to high-level role performance.’5
The Success Story of the Jewish People
How far does cumulative advantage take hold in the real world beyond the highly structured environments of schools and colleges? How, if at all, does it play out on the grand tapestry of human history? Can it explain the relative success not just of individuals over the course of decades but of peoples over the course of millennia?
Enter Zvi Eckstein. Eckstein has been the Deputy Governor of the Bank of Israel, has taught at Yale, Carnegie-Mellon, and Boston Universities among others, and has published in the fields of labour economics, monetary economics, macroeconomics, and development. What’s really interesting about Eckstein, though, is that he is one of the few world experts on Jewish economic history. If anyone can explain the mystery of the disproportionate performance of the Jews, it’s him.
A truly disproportionate performance it has indeed been. Take the number of Nobel Prizes awarded as a rough measure of achievement. In the first half of the twentieth century, Jews won 14 percent of the Nobel Prizes in literature, chemistry, physics, and medicine/physiology; in the next half-century the figure rose to 29 percent. The Jewish people, however, constitute just 0.2 percent of the world’s population. Two-tenths of one percent of the world’s population is responsible for about 30 percent of Nobel Prize–winning work. The maths is simple: the Jewish people are punching above their weight, so to speak, by a factor of more than one hundred.
Nobel Prize winners, however, are by nature unrepresentative. What about ordinary Jews? The statistics are every bit as curious. Even back in the early years of the twentieth century – when agriculture was still the mainstay of developed economies – as many as 99 percent of Jews from countries as diverse as those of Eastern Europe, Russia, the United States, and Canada had left tilling the land for more promising things. By the end of the century, approximately 53 percent of adult Jewish men were in the professions such as law, medicine, and academia, as opposed to only about 20 percent of white non-Jewish men. On the other hand, Jewish men are significantly underrepresented in the construction, transportation, and production industries: only 6 percent of Jewish men are out there working as builders, truckers, and factory workers, in contrast to 39 percent of adult males as a whole.
Follow the census and similar data and you find two other mysterious features of the Jewish story. The first mystery is the speed of Jewish ‘catch-up’. When Jewish immigrants reached America, they moved fast. In the first two decades of the twentieth century, it took less than six years on average for a Jewish immigrant to reach the occupational status levels of immigrants from Canada or Northwest Europe – and it took only about fourteen years to catch up with native-born whites. After just over five-and-a-half years in the United States, wages of Jewish immigrants were on a par with those of native-born Americans. What’s more, the Jews weren’t just making money and getting ahead in their careers; they were also getting themselves and their children educated. By the 1910s, Jewish immigrants were achieving higher levels of schooling than their native-born counterparts. By 1990, 71 percent of adult Jewish men were college graduates. (That compares with just 25 percent for adult white men.)
The second mystery is that there has been, effectively, a premium for being Jewish. By the mid-century, Jews had a 19.2 percent earnings advantage over non-Jews. What happens, though, if you control for other variables that might be responsible for this effect – living in urban areas, for example, or living primarily in the north? Jews still had an advantage (albeit slightly smaller) of 8.8 percent. The same phenomenon occurs into the latter half of the century. Jewi
sh men earned around 16 percent more than other men of similar age, schooling, and location.6
Outcomes so disproportionate cry out for an explanation.
Eckstein thought that the answer was to be found, if anywhere, in the unique history of the Jews.7 So he dug deep into the archives. On the one hand, there were the Cairo Geniza, a 200,000-strong collection of Jewish manuscript fragments found in the storeroom of the Ben Ezra Synagogue in Old Egypt. On the other hand, there was the Responsa literature, the thousands of written replies composed by the heads of the Jewish academy in reply to letters they had received from all over the world.
What Eckstein found was intriguing. There was no sign that the Jewish people had always been exceptional: in fact, prior to the middle of the first millennium, the Jews – from Eretz Israel and Mesopotamia to Egypt and throughout the Roman Empire – had (like their neighbours and hosts) been a primarily agricultural people. Nor did it appear that Jewish society had progressed in a slow and steady, linear fashion. Instead, what Eckstein found in the archives was evidence of a massive turning point in Jewish history squat in the middle of the first millennium. At that moment – and in spite of the fact that previously 80 to 90 percent of the labour force had been engaged in agriculture – the Jewish people simply stopped farming. By the end of the first millenium, in Eretz Israel itself only 20 to 30 percent of the population still farmed the land. In the other parts of the world to which the Jews had migrated (i.e., the Muslim Empire and Western Europe), there were as few as 10 to 20 percent and 5 to 10 percent, respectively, of Jewish workers involved in agriculture.
