Death of a Paradigm

Why flawed social science persists.

Written by Gary Marks.

“Socioeconomic status” or “SES” dominates research, public debate, policy and politics on social inequality. From early childhood to graduate school, inequalities between children and students are largely attributed to parents’ education, occupational class, income and wealth. Since education is important for subsequent social and economic outcomes, SES is understood as the ultimate source of educational and labor market inequalities in society — and for their reproduction across generations. These inequalities are falsely assumed to be profound, pervasive and persistent.

This is the essence of the SES paradigm. It is assumed that SES is the key influence for almost every conceivable educational outcome: test scores, grades, examination results, school differences, school and classroom climate, truancy and student misbehavior. Of course, SES adherents also believe that other factors influence student outcomes, such as schools, teachers, parents, and peers. However, according to the paradigm, SES is the dominant influence, and these factors mainly mediate SES effects.

What’s more, the differences in student outcomes by race, ethnicity, family size and family structure are assumed to be attributable to SES (at least in part). SES is claimed to be the primary influence on cognitive development among very young and preschool children — exerting its influence through material resources, parental stress and parenting behavior. In addition, it is frequently asserted that SES impacts (either directly or through education) the entire range of labor market outcomes: participation in the labor force, unemployment, occupation, earnings and career trajectories.

The SES paradigm extends further into health, crime, socio-emotional development, well-being, mental disorders and delinquency. Occupational class and SES remain part of the rhetoric of political parties and political activists of the left — although the relationship between them and political preference has reversed in many Western countries (Kitschelt & Rehm, 2022).

An alternative to the SES paradigm is the “IQ + genes” paradigm. This paradigm contends that the relationships of SES with educational and labor market outcomes are largely spurious. The relationships are due to the associations of parents’ cognitive abilities with their education, occupation and income, combined with genetic transmission from parents to their biological children. There is considerable evidence for these contentions (Marks & O’Connell, 2023). Thus, the intergenerational correlations of education, occupation and income are primarily epiphenomena and, therefore, cannot be substantially reduced by policy.

Paradigms

Many readers will be familiar with Thomas Kuhn’s The Structure of Scientific Revolutions. Normal science is based on a paradigm — a set of practices, beliefs and knowledge shared by its practitioners and adherents. Scientific paradigms define the legitimate problems and methods of a research field. Its accomplishments attract an enduring group of adherents, but unlike religion or ideology, allow unresolved issues to be investigated through research. Within a paradigm, knowledge slowly accumulates through revision, refinement, and the resolution of alternative theories and hypotheses.

Scientific paradigms work well until empirical observations do not conform with them. Anomalies are first accommodated within the paradigm by minor modifications to the relevant theories. For a while, more serious anomalies can be ignored or dismissed because of uncertainties about data, measures or methods. A crisis occurs when inconsistencies between theory and empirical observations accumulate so much that it is no longer possible to explain or dismiss incongruent findings within the paradigm.

A paradigm shift occurs when an alternative paradigm provides far more compelling explanations for relevant empirical observations. The original paradigm remains consistent with many empirical observations, but it is further eroded with each new set of findings that cannot be accommodated. Eventually, the original paradigm collapses, although the timing is influenced by the political context.

George Ritzer noted that a more meritorious paradigm cannot gain ascendency without winning the political battle. Less worthy paradigms can maintain hegemony through political channels despite their inadequacies. This is the case for the SES versus IQ + genes paradigms.

Paradigm shifts have occurred despite strong political opposition. For example, the heliocentric solar system replaced the geocentric one; Darwin’s theory of evolution replaced creationism; and in the USSR, Mendelian inheritance replaced Lysenkoism. Kuhn also discusses paradigm shifts — Maxwell’s equations, relativity theories, and continental drift — that were largely unnoticed because they posed no threat politically to anyone outside the small group of scientists. Today’s SES paradigm is unique in the depth and breadth of its political support. In contrast, the alternative IQ + genes paradigm is unpopular since it contradicts cherished political beliefs and ideologies.

John Ioannidis concluded that most research findings in medical science fields are false. The social sciences are likely to fare much worse. Ioannidis identified several characteristics that reduce the probability of research findings being true: small effect sizes; lack of replicability; many variations in definitions, analytical designs and findings; and a high degree of prejudice (i.e., political ideology).

The SES paradigm has most of these characteristics. SES effect sizes are only moderate and are much smaller when considering stronger predictors, such as cognitive ability. There is very little replicability, but an overabundance of similar analyses purporting to be novel. There is a great variety of theoretical explanations, operational definitions of SES, and statistical approaches. However, there are very few generally agreed-upon conclusions. Almost every point is contested with little or no resolution. Many research fields relating to social inequality in education, sociology, psychology and economics are highly politicized.

It could be argued the SES and IQ + genes paradigms do not constitute distinct paradigms at all. Cognitive ability can be simply accommodated within the SES paradigm. Many studies include measures of SES and cognitive ability in the same analysis. However, the two approaches are fundamentally different.

The SES paradigm assumes that the effects of SES represent purely sociological processes, such as home, school, economic and cultural resources. In contrast, the IQ + genes paradigm assumes that SES effects include non-sociological processes involving genetic transmission. Furthermore, the SES paradigm assumes that cognitive ability is a product of class background and other factors, with genetics playing little or no role (Fischer et al., 1996; Richardson, 2002).

A collapsing paradigm

The SES paradigm is collapsing because the very notion of SES is nebulous. Prominent theoretical concepts, such as economic and human capital, are construed to provide post-hoc justifications. There is no consensus on what SES is, and how to measure it. The variables that measure SES are often unreliable and are too weakly intercorrelated to support the claim that SES is a meaningful concept (Marks & O’Connell, 2021).

The collapse of the SES paradigm is clearly evident in the sprawling explanations for the effects of SES on education and other outcomes. There is no consensus on why SES matters, just innumerable overlapping theories. The sheer number and variety of theoretical explanations are a testament to their inadequacy.

SES-focused theories become prominent, not because they provide satisfactory explanations for a range of empirical phenomena, but because they become fashionable for a while and then fade away. Some have greater longevity than others. For a short time, prominent theorists strut the academic stage as if they have discovered something important, like a latter-day Newton or Darwin.

The ultimate reason that theories of SES inequalities go in and out of fashion is that they cannot account for the empirical phenomena they purport to explain. Plus, they ignore the fundamental empirical realities that SES effects are only moderate at best. The strong possibility that these effects substantially reflect parental abilities and their genetic transmission is almost universally ignored.

The IQ + genes paradigm is superior. It has much greater explanatory power and explains the observed relationships involving SES. It can account for empirical phenomena that the SES paradigm cannot.

It is well-established that intelligence is a valid and measurable concept. The numerous myths surrounding the concept, its measurement, and its importance have been debunked by extensive literature that has accumulated over the last century (Warne, 2020). For student achievement, educational and occupational attainment, and income and wealth, cognitive ability has greater explanatory power than comprehensive measures of SES. It accounts for a large proportion of the effects of SES, whereas SES only minimally accounts for the effects of cognitive ability (Marks, 2022). The effects of SES on cognitive ability and student achievement are primarily accounted for by the mother’s cognitive ability (Marks & O’Connell, 2023).

Furthermore, the most important variables involved in social stratification — cognitive ability, student achievement, educational and occupational attainment and income — all have sizeable heritabilities, that is, the proportion of the variation in a trait attributable to genetic differences (de Zeeuw, de Geus, & Boomsma, 2015; Plomin & Deary, 2015; Pokropek & Sikora, 2015; Hyytinen, Ilmakunnas, Johansson, & Toivanen, 2019; Silventoinen et al., 2020). The causal implications of high heritabilities cannot be dismissed (Egeland, 2023).

The strong explanatory power of cognitive ability and the sizable heritabilities are incompatible with the SES paradigm. The paradigm collapses because it is unable to account for these findings. Therefore, the vast SES-centred theoretical and empirical literature is rendered highly questionable. In addition, cognitive ability and genetics account for the lack of success for policies emanating from the SES paradigm aiming to eliminate, or at least substantially reduce, socioeconomic inequalities. It also explains why parents’ socioeconomic characteristics still matter under socialism and in the Kibbutz (Firkowska et al., 1978; Justman & Gilboa, 2012).

Sources of support

Given that the vast majority of SES-centered theories have poor explanatory power and are misleading because they ignore cognitive ability and genetics, why does the SES paradigm remain dominant? It survives because it is central to so much academic research, policy formulation, and political discourse.

Its fundamental contention that societal inequalities can be attributed to socioeconomic origins is largely unquestioned across academia, research institutes, government bureaucracies, international agencies, the commentariat, social media, teacher unions, the political left and even the political right. The adage that paradigms change “funeral-by-funeral” does not apply to the SES paradigm given its high level of political support.

In the social sciences, each new cohort of students is exposed to the SES paradigm. Many academics have built successful careers around some aspect of SES — perhaps by becoming an expert in a particular theory or theorist, proposing a slightly different theoretical explanation or resurrecting an old one. Most academic journal editors and referees accept the major tenets of the SES paradigm. It provides fertile, although very much over-tilled, ground for journal articles, books and book chapters, research grants, and PhD theses. This has been happening with little progress for over 50 years and is unlikely to change anytime soon.

Critical theory and post-modernism are also concerned with social inequality. However, they are not scientific — often explicitly rejecting scientific assumptions, methods and practices. Importantly, they deny the possibility of objective knowledge. However, they are accepted as academic disciplines with all their paraphernalia of conferences, journals and research grants. They are not rejected from academia for being unscientific and highly politicized because their general political orientation is consistent with that of the SES paradigm. So highly tendentious “research” that does not adhere to the evidentiary rules of normal science has become quite acceptable. This is disastrous for the accumulation of scientific knowledge and effective policies.

Academic research into social inequality has been strongly influenced by Marxism. Much of the inspiration for stratification research comes from the belief that socioeconomic inequality is morally repugnant, and the need to understand the capitalist system in order to change it. Many of the theoretical accounts for socioeconomic inequalities have Marxist legacies. For example, cultural capital theory purports to explain class inequalities with an emphasis on social exclusion and cultural hegemony. During the 1970s and 1980s, there was an enduring, and ultimately fruitless debate on Marxist and Weberian approaches to class. Many of the most strident critics of cognitive ability and behavioral genetics were self-identified Marxists (e.g., Gould, Lewontin and Kamin).

The SES paradigm is central to the political agenda of left-wing political parties and activists.

For the centre-left, socioeconomic inequality is a problem that can be alleviated by policies. Once in power, left-wing political parties promote, debate and implement policies consistent with the SES paradigm. Notably, they ignore similar policies that have failed in the past. Academic research is cherry-picked to support the party’s political position and proposed policies. Opposition is depicted as insensitive — ignoring, if not supporting, socioeconomic inequality. Since research communities and education bureaucracies generally accept the SES paradigm, center-right governments tend to do the same.

The SES paradigm is a major plank of far-left politics too. The supposedly strong and enduring SES inequalities constitutes further evidence of the iniquity of Western capitalist societies. No amount of contrary empirical evidence will change this belief, despite enormous improvements in wealth, health and living standards, substantial reductions in real poverty, the expansion of social welfare and other government services, and the implementation of policies specifically designed to reduce socioeconomic inequalities.

What are the far-left policy prescriptions to combat the reproduction of socioeconomic inequalities? Generally, it is state control of all the major societal institutions. Entry to elite educational institutions would not be based on academic performance but by bureaucrats using political criteria. Similarly, the state would be involved in hiring, promotion, and dismissal in workplaces. They advocate intellectuals (like themselves) leading an agenda of radical social reform to obliterate disparities — a model that has historically proven disastrous wherever implemented.

A larger political agenda

No research in cognitive psychology or behavioral genetics argues that social outcomes are biologically determined. No one actually identifies as a “genetic determinist”. Behavioral genetics does not dismiss the role of the environment; the shared and unshared environment are fundamental concepts. Likewise, no one argues that heritability applies to individuals or is immutable across time and societies (see Sesardić, 2005).

There is no logical connection between cognitive IQ research and apartheid, Jim Crow laws, or state-sanctioned racism. Much of the support for eugenics in the late 19th and early 20th centuries was from progressives and socialists (Freeden, 1979; Paul, 1984).

Yet, the IQ + genes paradigm has been attacked as being offensive to low SES families and social minorities. This critique is not about sensibilities but, again, a political ploy to silence criticism of the SES paradigm. There is a vast industry and prominent political ideologies to protect.

Although the language surrounding SES-based explanations is usually couched in a sympathetic tone, such explanations could also be considered offensive. For example, they seem to imply that less educated parents are uninterested in their children’s education, have negative attitudes towards academic pursuits, use fewer words and less complex language, and do not read enough to their children.

So why are the attacks on cognitive psychology and behavioral genetics so vehement? It is because the IQ + genes paradigm poses a scientifically rigorous challenge to left-wing sensibilities. The SES paradigm, although failing as an explanatory model, survives because of ideology, not science.

A different version of this article was previously published on Aporia.

Gary Marks is an honorary principal fellow in the Department of Sociology, Social and Political Sciences at the University of Melbourne.

---

Become a free or paid subscriber:

Like and comment below.

---

References

de Zeeuw, E. L., de Geus, E. J. C., & Boomsma, D. I. (2015). Meta-analysis of twin studies highlights the importance of genetic variation in primary school educational achievement. Trends in Neuroscience and Education, 4(2015), 69–76. https://doi.org/10.1016/j.tine.2015.06.001

Egeland, J. (2023). Heritability and etiology: Heritability estimates can provide causally relevant information. Personality and Individual Differences, 200, 111896. https://doi.org/10.1016/j.paid.2022.111896

Firkowska, A., Ostrowska, A., Sokolowska, M., Stein, Z., Susser, M., & Wald, I. (1978). Cognitive development and social policy. Science, 200(4348), 1357-1362. https://doi.org/10.1126/science.663616

Fischer, C. S., Hout, M., Jankowski, M. S., Lucas, S. R., Swidler, A., & Voss, K. (1996). Inequality by design: Cracking the Bell Curve myth. Princeton: Princeton University Press.

Freeden, M. (1979). Eugenics and progressive thought: A study in ideological affinity. The Historical Journal, 22(3), 645-671. https://doi.org/10.1017/S0018246X00017027

Hyytinen, A., Ilmakunnas, P., Johansson, E., & Toivanen, O. (2019). Heritability of lifetime earnings. The Journal of Economic Inequality, 17, 319–335. https://doi.org/10.1007/s10888-019-09413-x

Ioannidis, J. P. A. (2005). Why most published research findings are false. PLOS Medicine, 2(8), e124. https://doi.org/10.1371/journal.pmed.0020124

Justman, M., & Gilboa, Y. (2012). The scope for promoting equal opportunity in education: Evidence from the Kibbutz. Education Finance and Policy, 7(4), 489-515. https://doi.org10.1162/EDFP_a_00077

Kitschelt, H. P., & Rehm, P. (2022). Polarity reversal: The socioeconomic reconfiguration of partisan support in knowledge societies. Politics & Society, 0(0), 00323292221100220. https://doi.org/10.1177/00323292221100220

Kuhn, T. S. (1962/1996). The structure of scientific revolutions (Third ed.). Chicago: University of Chicago Press.

Marks, G. N. (2022). Cognitive ability has powerful, widespread and robust effects on social stratification: Evidence from the 1979 and 1997 US National Longitudinal Surveys of Youth. Intelligence, 94, 101686. https://doi.org/10.1016/j.intell.2022.101686

Marks, G. N., & O’Connell, M. (2021). Inadequacies in the SES–achievement model: Evidence from PISA and other studies. Review of Education, 9(3), e3293. https://doi.org/10.1002/rev3.3293

Marks, G. N., & O’Connell, M. (2023). The importance of parental ability for cognitive ability and student achievement: Implications for social stratification theory and practice. Research in Social Stratification and Mobility, 83(February), 100762. https://doi.org/10.1016/j.rssm.2023.100762

Paul, D. (1984). Eugenics and the left. Journal of the History of Ideas, 45(4), 567-590. https://doi.org/10.2307/2709374

Plomin, R., & Deary, I. J. (2015). Genetics and intelligence differences: Five special findings. Molecular Psychiatry, 20(1), 98–108. https://10.1038/mp.2014.105

Pokropek, A., & Sikora, J. (2015). Heritability, family, school and academic achievement in adolescence. Social Science Research, 53(September), 73-88. http://dx.doi.org/10.1016/j.ssresearch.2015.05.005

Richardson, K. (2002). What IQ tests test. Theory and Psychology, 12(3), 283-314. https://10.1177/0959354302012003012

Ritzer, G. (1975). Sociology: A multiple paradigm science. The American Sociologist, 10(3), 156-167. http://www.jstor.org/stable/27702185

Sesardić, N. (2005). Making sense of heritability. Cambridge: Cambridge University Press.

Silventoinen, K., et al. (2020). Genetic and environmental variation in educational attainment: An individual-based analysis of 28 twin cohorts. Scientific Reports, 10(1), 12681. https://doi.org/10.1038/s41598-020-69526-6.

Warne, R. T. (2020). In the know: Debunking 35 myths about human intelligence. New York: Cambridge University Press.