Infine, per chi si fosse appassionato a Schelling, vediamo insieme come ragiona su un problema demografico, quello della scelta del sesso dei figli (ricordiamoci che quando scriveva non c’era la tecnologia per farlo, che adesso è invece disponibile):
Imagine that it were possible to choose in advance the sex of children. It is an easy idea to toy with; there is no difficulty in knowing what it means.
We can already choose whether to have children at all, at what age to have them, how many to have and how to space them over time; we can even somewhat control the sex composition by, for example, stopping when we already have a boy and a girl or trying again if we don’t yet have what we want. Exercising a choice of sex would not le ad to any new kinds of families: all the family combinations of boys and girls already exist.
Our interest is in the consequences, not the technology, and in how we deal with a choice that has never mattered before. But the technology itself can affect some of the questions we want to pursue. For example, is the technology under the control of the mother alone or does it require cooperation between the parents; will it be known whether or not the choice was exercised and to whom will it be known; and if a girl is decided on and a boy is born is it likely due only to the imperfect reliability of the method, or due to carelessness, or due to cheating? If a child ever wonders whether he or she was “wanted,” the advent of contraception can affect the child’s acceptance of a positive answer; will the technology of sex choice be such that the child will know what sex its parents tried for, and how likely it is they succeeded?
Leaving those questions behind, let’s speculate on how people would choose if they did choose. Speculate is all we can do. There is no real evidence. We cannot investigate what people do in fact choose, because they do not in fact choose. And even if we ask them, as several researchers have done from time to time, it is hard to take the responses very seriously.
It is a little like the question, what would you ask for if you caught an enchanted sturgeon and were offered three wishes to let it go free? Not expecting the opportunity, you are unlikely to spend much time making plans for it. The sex of children is a question about which most people – not everybody but most people – are unprepared, especially people who have not yet had their first child. Nobody would dream of making a decision within the short time the interviewer will wait for an answer; and no couple is going to attempt to reconcile any differences they have, or even delicately explore each other’s preferences, for the sake of providing a hypothetical statistic in a survey.
There have been attempts, now that contraception has been widespread and additions to American families can be interpreted as partly intentional, to look at any sex preferences that may be revealed in actual choices of whether or not to go on and have another child. The idea is simple: if families with two girls or two boys more frequently have a third child than families with one of each, this could mean that people want at least one of each and keep trying if they don’t get them in the first two. But the statistics don’t show much, and there are other interpretations. It is widely observed that families with girls and families with boys are different kinds of families. Most parents agree on that. It is possible that families with two girls, in deciding to have a third child, are not seeking a boy but find children a pleasure and, since two are not too many, they look forward to a third; however, a family with two boys may face a different noise level, or be slightly less satisfied with family life, or, equally satisfied, be more impressed that two is a large number. Or maybe boys and girls have a different effect on relations between the spouses or anything else (like the divorce rate) that has a statistical influence on the birth of a third child. In other words, we couldn’t be sure that we were observing preferences for either boys or girls if we did find some of these differences in the census figures.
Furthermore, there are at least two ways that preferences might change if the choice became an actuality. There are many cultures in which boy babies are a sign of virility or of God’s favor; there is a slightly coercive tradition that fathers want boys, and the congratulations sound more self-assured when the father has announced a boy. Even grandparents have been known to offer condolences when a third child is a third girl. All of that may evaporate once it’s known that the sex of the baby indicates nothing more than whether the mother took a blue pill or a pink pill. The father who insists he really is glad that his second child is a girl like the first won’t be thought merely keeping his chin up if it’s clear that, had he wanted a boy, he could have had one.
And a new set of social and demographic influences will come to bear on the choice if parents have to observe and anticipate sex ratios that depart from the approximate 50-50 in which boys and girls have traditionally been born. If the sex ratio within the ethnic groups or region or social class with which parents identify their children for school, marriage, and career, departs substantially from the historical ratios, and especially if there are government programs to tilt the incentives, people will have to think about the relative merits of being in the majority or minority sex. How they would make those calculations is at present just more speculation, but it is a fair guess that they would make them. If the little boy reports that two-thirds of his kindergarten class are little boys and only one-third girls, his parents will reflect on those figures before deciding whether their next should be a girl or a boy.
Will people be glad to have this choice available? Or will it just add one more decision to make, one more source of conflict, one more opportunity far remorse, when life is already full enough of decisions and married couples have enough to disagree about?
Demographically, the main effects wilI be the aggregates – the overall sex ratio, or the ratio within particular age groups, ethnic groups, socio-economic groups, and other groups within which social life and marriage occur. But there could be effects on family itself, although it is hard to know how to appraise them. Far example, if the main direction of choice were toward balanced families – a boy and a girl in two-child families – fewer boys would have brothers and fewer girls would have sisters, more boys would have sisters and more girls would have brothers. With today’s technology, half the boys in two-child families have brothers; with a technology that leads to mixed families, none of them would.
For the overall ratio, we can do a little arithmetic to get an idea of the differences that various choices could make. A preference that appears to show up in interviews and questionnaires in America and Western Europe is a desire far at least one boy. This sounds like a modest male preference, but may not indicate a preference at all if people also wish to have at least one girl. There has also been occasionally observed in the surveys of hypothetical preferences some desire to have a boy first. These two choices could reflect the same preference: if you want at least one boy, a boy first relieves the suspense.
To get a feeling for the arithmetic, consider what would happen if every family elected a boy first. The result depends on whether the choice is to have a boy first and leave it to luck thereafter, or a boy first and balance out with a girl. The arithmetic also depends on how many families end up with a single child, how many with two or three or four. Suppose every family had first a boy and subsequent children at random. The one-child families would be all male, the two-child families would be three-quarters male, the three-child families would be two-thirds male, and so forth. Given the family sizes in this country, the children born in that fashion would be 70 percent male and 30 percent female, a ratio greater than two-to-one. If every family had first a boy and then alternated girls and boys so that an even-numbered family would have equal numbers and an odd-numbered family one more boy than the girls, births would be 60 percent boys and 40 percent girls. In this population, no girl would be without a brother; three-fifths of the girls would have no sisters and a third of the boys would have no brothers.
Suppose all families want at least one boy but will take what the lottery gives them until the last child and then, if they do not yet have a boy, choose a boy. Except for families that know in advance that they want only one child – and these are far fewer than the 21 percent in the United States that actually have only one living child – the effect will be small; only families that would have ended up all girls will be affected, and they will have a single boy in place of a girl – and even that will not happen in families that stop having children before they complete their plan.
What are the consequences of an imbalance in the sex ratio? Of all our institutions, monogamous marriage is the one most directly concerned. But in that regard there are already imbalances. First, there are geographical differences ranging from an excess of women in the Washington area to a large excess of men in some Western states and especially Hawaii and Alaska. Second, young women of an age to marry have recently outnumbered young men of an age to marry in this country because of the tendency for husbands to be older than wives at fìrst marriage; with new births increasing at 3 percent per year, as they did for the quarter century that ended in 1956, a three-year age difference means that the women are drawn from a more recent population that is almost 10 percent larger. Third, women live longer than men in this country, and there is a large excess of unmarried women over unmarried meno The ratio is nearly 4 to l in the age group beginning at 45. The difference in life expectancy for men and women in their early twenties is six or seven years; and the young woman who marries a man three years older can expect on average to outlive him by a full decade. Evidently the near equality of male and female births coexists with sizable imbalances for important age groups.
What does the government do, as a matter of policy, if boys and girls are born in very unequal numbers, or even if the ratio fluctuates in cycles, evening out in the long run but leaving large alternating imbalances in successive age groups? At the level of “technical policy,” the problem is probably no harder than coping with inflation or unemployment, energy, changes in the birth rate, or changes in the ratio of elderly retired to the working population. The government could attempt to “stabilize” the birth ratio by a variety of fiscal measures, like differential income tax deductions, differential eligibility of men and women for military service, arrangements for differential college tuition, and a variety of favoritisms and affirmative actions discriminating by sex. It wouldn’t be easy to devise successful policies, but it wouldn’t be analytically different from so many things that the government presently tries to stabilize.
But the social and even constitutional implications are awesome. Imagine the government’s having to have a policy on a “target” sex ratio for births. Imagine that Presidential candidates had to debate whether it’s better for men to exceed women by 5 percent or 10 percent or not at all, or for women to exceed men. Besides the need to incorporate a multitude of sexually discriminatory rewards and penalties throughout the government’s expenditure and revenue and regulatory programs, there would have to be a policy on the “correct” numbers of men and women to have.
There are already people who argue that federal programs to help the poor with family planning have racial implications, even racial motives. Imagine having explicit demographic targets: a President proposing measures that would hold inflation to 4 percent, unemployment to 5 percent, and excess little boys to 6 percent.
So it isn’t only parents who might like to be spared some of the choices that would have to be made if this particular technology became available. There are some things – the weather may be one, and the sex of a child at birth another – that it is a great relief to be unable to controI. The lottery dispenses arbitrary justice indiscriminately, but it may beat having to discriminate (pp. 197-203).