By William J. Andrews, Silver Spring, Maryland Vol. 32, Mankind Quarterly, 04-01-1992, pp 311.
While there are other significant problems related to the future of this country that need serious attention, all of which must be resolved if our civilization is to survive, this paper discusses the need to react to the possibility of a dysgenic threat due to an intergenerational decline in human intelligence. This is a problem that is seldom publicly discussed or debated, and therefore most people do not have an adequate background upon which to base intelligent conclusions. It has been especially clouded by the hairsplitting academic debate concerning the exact degree to which intelligence is dependent on heredity.
The subject of this paper was suggested by a reading of a debate between Eysenck and Kamin[1] relating to the nature/nurturecontroversy. In this controversy, Kamin holds that environmental factors are dominant in determining a person's IQ, and that any hereditary contribution to IQ is negligible and that genetics therefore is irrelevant to any explanation of acknowledged group disparities in academic or professional success. Eysenck, speaking for the hereditarians, by contrast, argues that although environment plays a role in determining IQ, genetic factors are also important. Hereditarians believe that the threat to the level of IQ in the West is sufficiently real that solutions should be sought in both areas, at least until it has been conclusively determined that remedial intervention is not required.
In reading the referenced work it became apparent that a lay reader could make no serious assessment of the subject without studying the original references cited in that work as well as the many studies which have been carried out since then - a task of several months. Even then a lay reader might conclude that the task was beyond his capabilities as the level of statistical competence required is challenging. The debate focuses on the relative apportionment of IQ to heredity and environment. In what follows, an argument will be made that the need to adopt anti-dysgenic measures is not directly related to the outcome of that debate, but that pragmatically and ethically it depends on the extent to which we are prepared to ignore the threat and simply pass the problem on to our progeny, instead of accepting responsibility for our share of the required effort.
For a lay person interested in this subject, two sources of information are recommended. One is Arthur Jensen's Bias in Mental Testing[2] and the other is a review of that book published in Behavior and Brain Sciences.[3] Jensen's work is complete and thorough. For those seeking less challenging reading, however, the review article cited above will be appreciated. It includes not only a precis of Jensen's book, but also eight pages of peer review and commentary, followed by the author's response. This review thus presents a balanced overview of the entire issue.
As previously mentioned, the core of the Eysenck-Kamin debate is the extent to which a person's IQ depends on heredity and how much upon environment. There are two basic problems posed by the distribution of IQ. One of these is a drift in the average IQ in the downward direction, which makes a nation less competitive in world markets, lowers the standard of living, exposes more workers to obsolescence (due in part to increasing intellectual requirements caused by advances in technology), and makes a democratic government less viable. A second and equally obdurate problem is that significant differences in intellectual competence have been found to exist between groups, and this causes inevitable friction and hard feelings. The crucial parameter about which the debate rages is the degree to which intelligence is genetically determined.
It is the purpose of this article to demonstrate that, from the point of view of dysgenics, the importance of the debate as to the exact heritability of IQ is overblown and concerns largely technical issues, and that the ethical side of the question issue is not dependent upon the outcome of the environment-heredity debate - even though this is now widely accepted as being in the ratio of 30 to 70 (i.e. the heritability of IQ is believed to be in the region of 70%). The issue is an ethical one, and if people in general had a better understanding of the underlying ethical issues, it might be easier for us to develop the consensus badly needed for finding and implementing solutions.
Figure 1 illustrates in simplified format how both evolution and eugenics operate to make changes in our genetic heritage. This figure shows two curves, one labeled "original," the other "derived." The curve marked "original" illustrates the distribution of a parent population, assuming no crossbreeding, as a function of IQ or intelligence quotient. The number of persons with IQs falling between any two adjacent IQ points along the horizontal scale would be represented by the height of the curve at the center of that interval. The total number of persons in the population is proportional to the area under the curve, and can be found given the scale factors for the horizontal and vertical axes. For convenience of discussion this is centered about an assumed mean IQ of 100, and has a distribution characterized by a typical breadth as defined by a standard deviation of 15 IQ points, or more universally understandable, by half of the distance between the half- amplitude points of plus/minus 17.66 IQ points as shown. For the purpose at hand, the vertical scale is normalized to one for the peak of the distribution as a matter of convenience, as the amplitude is not significant in the development.
Drawn through the peak of this distribution is a sloping (dashed) straight line, which represents the number of children the parents have, on an average, with respect to the replacement birth rate. For instance, from Figure 1 at an IQ of 85, one standard deviation from the average of 100, the number of children exceeds the replacement rate by a factor of 30%, and at an IQ of 115 the number falls short by 30%. These values are illustrative only, and the value of the slope, 0.03, is chosen to make the figure less cluttered. As implied above, the slope is expressed as a fractional value per standard deviation in IQ.
In our society, conditions are such that almost all will survive to maturity. Assuming for the moment that the children of the parents will have the same average IQ as the average of their parents, a curve similar to the "original" can be derived by the multiplication of the "original" distribution by the birth-rate slope, and the resulting distribution is labeled "derived." One can see that the peak and average of the "derived" curve has been shifted in the left direction to a new average value as shown. One effect not accounted for, which in a quantitative analysis would need to be included, is the effect of assortive mating. There is a relatively high level of correlation between parent's IQs, often estimated at 45%. This effect, as well as environment would need to be included explicitly before one could give quantitative significance to the figures.
Clearly this is a very simplified presentation, as in general the curve giving the ratio of the actual number of children to the replacement rate will not be a straight line function of IQ, and the equivalent slope for a real relationship may be steeper, shallower, or of the opposite sign. None of these considerations, however, are significant for the purposes at hand. It is not intended here to assume that this new average so found represents a new value for the average population IQ. In fact, it is almost certainly not that large. Most children' s IQs are located between the parent's IQs and the mean of the population. Taking this fact into consideration, the actual new "derived" average change will be some multiple (less than one) times that shown in our example. For the simplified model used here, the fraction that determines how much of the child average IQ change shown one actually gets is the same number which, when multiplied by the parent's IQ referenced to the group mean, gives the average of the children's IQs. This parameter is the "heritability" as defined for purposes of exposition in this very simple model. A more complete analysis would give a more complex but similar result.
Very crudely, this is the process by which evolution made changes to adapt man in his struggle for existence. Those unable to feed and care for their children, either collectively or individually, would have fewer surviving children. Those more capable of producing and raising children until they in turn can reproduce would have more descendants, tilting the curve to the right, indicating an increase in the successful traits. Intelligence is only one of the almost innumerable survival-oriented traits that have been similarly reinforced. In a Malthusian world, where the population is controlled by available resources and birth control measures are not available, life is competitive. In modern societies in which standards of living and social support are such that effectively all individuals survive to maturity, nature' s adaption mechanism now uses a new criterion - simply the number of births per mother irrespective of other parental characteristics. If there is a negative slope due the correlation of IQ with birth rate, the average IQ of a population will decline. The significant criteria here are two things, the equivalent slope of the birth-rate/IQ curve and the heritability. At least superficially, it would appear that the heritability is a very significant parameter.
Figure 2 shows the average IQ drop decline due to the above mechanism, as illustrated in Figure 1, for six values of heritability. Figure 2 uses a raw drift rate, before allowance for heritability, of 1.5 IQ points per generation. There is very little data on this subject, but the value of 1.5 is consistent with a value from Vining[4] of approximately 1 IQ point per generation, which is equivalent to the assumed 1.5 above with a heritability of 0.667. A generation is taken as 28 years, although this is different for different societies and is not constant in time in any case. The heritability assumed for the plot falls comfortably in the range of 0.4 to 0.8 which bounds most estimates. Once again, the exact value selected here is not crucial to the argument. The downward drift is proportional to the product of the amount of crude unbalance before taking into account heritability, the heritability, and time. If one assumes constancy of the values, the average IQ curves drop a constant amount each generation.
It is clear from the curves shown in Figure 2 that the serious problem is not so much in the near term as in the future, and for some assumptions, in the distant future. For one who has no concern for the future the question is moot. Current dysgenic affects are not important to those who are only concerned with the present. On the other hand, for someone who feels concern for those yet to be born, the question is clearly not moot. At almost any value of heritability, the problem is serious.
Most of us will accept that we have a minimum obligation to leave the world in no worse condition than we found it. Given that assumption, one might ask what the genetic burden will be, remembering that the degree of heritability is the multiplier for any eugenic program, just as it is for the dysgenic effects of the adverse birth-rate imbalance. In order to have no degradation in population average IQ, the curve corresponding to the "derived" curve of Figure 2 must average 100 or greater. The short-term seriousness of the slippage, if any, of course, would be ameliorated by low values of heritability. However, under those circumstances, even though the IQ drop is not large, the amount of eugenic effort to bring it back will correspond to the raw dysgenic trend rather than to the drop in IQ. If a dysgenic trend is allowed to proceed unchecked, it accumulates in time. Any postponement of necessary eugenic activities means that posterity will have the onerous chore of remedying our neglect- something it will by definition be less well-equipped to do.
Figure 3 has been prepared to illustrate this issue. For illustrative purposes the value of the downward drift rate is taken to be 1 IQ point per generation of an assumed 28 years. Starting at the left at a presumed average IQ of 100, three options are illustrated. The lowest path illustrates the "do nothing" option. The shadings below the solid lines indicate the presence of dysgenic pressures that cause a constant drop in average IQ. The shadings above, not present in this instance, indicate eugenic efforts. Sometimes the effort above the line is shown with twice the amplitude of the dysgenic pressures. This indicates that adequate eugenic effort must be expended to overcome the existing dysgenic trend before one can effect any positive compensatory changes. This lower curve, followed to its logical conclusion, results in social chaos.
The response of a group or society to such a drift depends upon the population for which that society has (not claims) ethical concern. If the reference population consists of adults over seventy who happen to have no ethical concern for their children or progeny, certainly the subject is moot. With an expected life of a fraction of a generation and no concern for others, even under the worse circumstances they would have no reason to act. Suppose, however, that the population of ethical concern included all those now living and to be born before 3000 AD. Now one has a vastly different situation and projected consequences for a thousand years into the future are significant.
To take a pragmatic case, predictions from known imbalances in the birth rate of different social classes suggest that the imbalance is now on the order of 1 to 2 IQ points per generation, before allowing for the heritability rh. Taking a generation at about 25 years or four generations in a century, this would give an approximate six IQ point decline in 100 years for an rh of one, which value is almost certain to be incorrect. If it were 0.5 then the six-IQ point would be reached in 200 years. In 1000 years, it would be five times that or 30 points. With today's requirements on intellectual tasks due to our complex technological society, and not the slightest hint that things will get simpler in the future, that should count as a disaster within the context of that person's value judgement.
The value of heritability, therefore, is important because it defines the level of short-sighted selfishness at which the problem of declining intelligence can be ignored. Anyone who has no concern for the future, then, can even ignore the postulated problem if the heritability were one (100%). On the other hand, one might make a basic assumption, not unreasonable, that at a minimum we should not bring children into the world and pass to them an inheritance inferior to that which our generation recieved - notably, one with an inferior gene pool. That is, we are in no mood to participate in eugenic or other highly "dubious" endeavors unless it is forced upon us by logic. Not letting things get worse we will accept, but progress we will leave to the future. Making such an assumption completely bypasses the question of the value of rh. If rh is high, then a eugenics program is very efficient, but on the other hand the dysgenic stress is also powerful due to the high heritability. Thus the appearance of the heritability in both the requirement side and the eugenics side means that it is irrelevant under the "hold our own" assumption.
If rh is very low, then society's average IQ is a valuable resource, valuable because it is durable, and also because the cost in obtrusive eugenics programs is very high if one wishes to raise the level of IQ for future generations. While a policy of maintaining the status quo sounds like an acceptable objective, there is another objective which couples into the problem in a very significant way. That is the case where one wishes to remove significant differences in intellectual competence between groups as a matter of equity and compassion. Graceful acceptance of such differences is scarcely possible. Even for a very obtrusive program, the period of time is long under almost any assumption, and is on the order of 100 to 150 years, depending upon what one means by "getting there." If the value of rh were very low, it might be very difficult to get anything started, given that the short-term payoff would pragmatically be small. On the other hand, any such a gain made would last for quite a while, even if the country allowed itself to slip into a situation where some dysgenic trends were tolerated. And one should be careful not to overlook intangible gains due to the self-respect one generates when one tackles a tough and difficult problem and passes a very significant benefit to one's children.
Therefore, for the midrange of values of rh, the only rationale for the failure to respond to a downward drift in IQ applies if we wish to, as it were, sponge on future generations. While extremely low values of heritability provide an attractive rationalization for doing nothing and passing the problem on to our children, such rationalizations are scarcely consistent with an ethical system.
Which is to say that the degree of heritability of intelligence, as debated in the nature/nurture controversy - while significant from the standpoint that the more we know and understand, the more effective we can be - does not affect the ethical obligation to recognize the need for some kind of eugenic policy. The threat of dysgenic trends surely exists in our modern world, but what the debate about the precise degree of heritability is about is whether the impact is fast or slow, that is, whether we can excuse ourselves from action and pass the problem to our children by rationalizing excuses for inaction. Viewed in that context, there is no reason why we should delay consideration of eugenic measures just because some scientists still question the exact degree of heritability, or argue that it might vary slightly from race to race. Those who argue for a low value of rh are opting to pass the problem to our children. Those who perceive a high value of rh are trying to point out that dysgenic trends may not only be real but that the issue is an urgent one. To do our moral duty we should investigate the threat of dysgenic trends and attempt to take some kind of action -- if nothing more than alerting the public to the situation.
As long as rh falls within a reasonable range, and is not zero, he who claims the high moral ground should at the very least argue for a very significant effort to clear the air with a "we did this, and this is what happened" type of research program[5-6] the results of which the electorate can understand and accept, in contrast to arcane arguments which only specialists can follow.
In sum, the academic battles that rage over the selection or prediction of heritability are much overblown and of less significance that they are made out to be. Many are arguing without any real understanding of the implications of the debate, which essentially revolve around practical issues concerning the need (or otherwise) to implement some form of eugenic policy. If this generation is ethically concerned about its responsibility to future generations, it should perhaps design a pragmatic test to determine the efficacy of a voluntary eugenics program - one that would be acceptable to both sides, the results of which could be clearly demonstrated to the public at large. As I hope to have shown above, the exact extent to which genetic factors determine intelligence is not a criterion for the acceptance or denial of the need for a eugenic program. Since none today deny that genetics plays some role in determining intelligence, the need to consider the evidence for dysgenic trends cannot ethically be avoided. A close relationship between heredity and intelligence makes eugenic considerations all the more pressing, but dysgenic trends are quite as fatal whether they take place slowly over a long period of time or rapidly. If we are concerned for the future of the (hopefully) millions of generations still to be born, we must realize that their fate lies to a considerable extent in the breeding practices of those who are currently alive.
1 Eysenck H.J. and Kamin, Leon, The lntelligence Controversy, 1981, John Wiley and Sons, New York and Toronto.
2 Jensen, Arthur R., Bias in Mental Testing, 1980, New York Free Press.
3 Precis and review of Bias in Mental Testing, Behavior and Brain Sciences, Vol. 3, Sept. 1980, Precis pp 325-333, commentary pp 333- 359, author's response pp 359-371.
4 Vining, Daniel R., "Fertility Differentials and the status of nations: A speculative essay on Japan and the West", in Intelligence and National Achievement, Cattell, Raymond B., Ed., Cliveden Press.
5 Andrews, W.J., "Eugenics Revisited," Mankind Quarterly, Vol. XXX, No. 3, Cliveden Press, 1990, pp.235-302.
6 Andrews, W.J., "Addendum to 'Eugenics Revisited," "Mankind Quarterly, Vol. XXXI, No. 3, 1991, pp. 305-316.
GRAPH: Figure 1. Figure illustrating dysgenics and eugenics mechanism.
GRAPH: Figure 2. Time Impact of value of heritability, (Heritability shown in tenths.)
GRAPH: Figure 3. Schematic illustration of several scenarios,