Research relating educational attainment to earnings has consistently found dramatic benefits for employees with increased schooling. Over their lifetimes, high school graduates will earn $212,000 more than nongraduates, and each additional year of school attainment beyond high school is associated with increasing income. For example, college graduates will, over their lifetimes, earn $812,000 more than high school dropouts, and graduate students with professional degrees will earn nearly $1,600,000 more than college graduates.
The correlation between school attainment and earnings is likely due to a combination of an indirect effect and a direct effect. First, schooling has an indirect effect because schooling affects intelligence, which, in turn, affects economic outcomes. That is, individuals who attain more schooling are more intelligent, and intelligent workers are rewarded for the skills they display in training and in doing their jobs. Second, schooling has a direct effect because of minimum entry-level educational standards required for acquiring certain higher-paying jobs. For example, many civil service jobs require minimum educational credentials, and higher earnings are associated with jobs requiring more schooling.
The reasons for the indirect effects of schooling on intelligence are less clear. It could be that staying in school is related to later earnings because individuals who were more intelligent before they entered school stay in school longer. Thus, when employers hire based on high educational attainment, they may be selecting individuals who were simply smarter to begin with.
Alternatively, it could be that the very act of attending school makes one smarter regardless of one’s initial level of intelligence and this acquired intelligence produces workers that are more effective. Even the least intelligent student is bound to pick up some of the instruction, and this instruction is helpful when it comes to taking intelligence tests as well as performing jobs. Consider that IQ tests ask about information and events that may have been covered in school, such as “Who wrote Hamlet?” or ” What country are the Great Pyramids in?” While one can come across such information outside of school, most individuals acquire such knowledge through explicit schooling and school-related activities (e.g., participating in plays).
To support the notion that schooling increases intelligence, one has only to compare two types of individuals who were comparable in intelligence in eighth grade. One type continued to attend school until graduation from high school, while the other type dropped out of school sometime between 9th and 12th grade. When these two groups are retested at age 18, they no longer have comparable intelligence. The dropouts lose, on average, nearly two IQ points for every year short of high school completion. Because the two groups had equivalent intelligence before the latter group began dropping out, we cannot attribute the IQ differences at age 18 to anything other than differences in schooling. Thus, when employers choose individuals based on their educational attainment, they are likely to be choosing employees who have acquired higher levels of intellectual ability.
Over two decades ago, researchers demonstrated that so-called general intelligence was far more predictive of occupational success than were specific ability factors such as arithmetical reasoning, vocabulary, spatial skills, and memory. These findings clearly explain why employers want to hire individuals with increased educational attainment. Given that these individuals would likely have higher general intelligence because of increased schooling, they are likely to be more productive and successful employees. Furthermore, these findings suggest to some that training individuals with low general intelligence in specific skills is not likely to be as effective as hiring someone with higher general intelligence.
It is important to note that at each level of schooling completed—finishing high school, finishing junior college, finishing a four-year college, and so on—there are pronounced differences in intellectual ability among students at the same level of schooling.
No school-induced increase in IQ or general intelligence can fully explain the substantial differences in intelligence that exist among students at the same level of education; there will always be individual differences in intelligence. Even though dropping out of school can be shown to cause IQ to drop, it is nevertheless true that among those who stay in school, there are large differences in IQ.
The available evidence thus suggests that schooling and intelligence make independent contributions to job success and lifetime earnings. Consider that even among those with comparable levels of schooling, the higher a person’s level of intellectual ability, the higher that person’s weekly earnings will be. As shown in other studies, there is a linear trend for a rise in earnings to be associated with increasing intellectual ability. Workers with the lowest levels of intellectual ability earn only two-thirds of what workers at the highest level earn. Because differences in schooling are statistically controlled in such studies, the source of the rise in earnings must be due to other factors, such as variations in intellectual ability. Thus, intelligence conveys significant advantages in lifetime income.
Data provided by the U.S. Department of Labor show a significant statistical relationship between measures of general intelligence and job-related aptitudes such as manual dexterity, spatial ability, and so on. In addition, a composite of nine different aptitudes, including a measure of general intelligence, significantly predicts performance in over 400 different jobs. Although the prediction of job performance is far from perfect, it is strong enough to suggest employers can save some money by hiring persons with more intelligence. Furthermore, the IQ measure by itself also significantly predicts job performance.
Another study found that the ability of a measure of general intelligence to predict job performance (that is, its predictive validity) was stronger as the job demands became more complex. In this study, IQ (and related measures) predicted occupational success almost as well as IQ has been shown to predict school grades. Although selection based solely on general intelligence is certainly not perfectly accurate, including a measure of general intelligence in a selection battery holds great promise because it can be combined with other measures (e.g., grades, letters of reference) to still further increase the predictive validity of the selection battery.
The value of using a measure of general intelligence as part of the hiring process depends on several factors: how many qualified applicants are in the pool, how intellectually demanding the job is, and how substantial the differences are between good and poor performance on the job. Earl Hunt commented in his 1995 book Will We Be Smart Enough? that it doesn’t pay to use tests to select someone for a job that almost anyone can do, such as cutting your lawn. However, a testing program can be useful when relatively few people can perform the job effectively, when not many people are needed to do the job, and when the costs of making a hiring mistake are substantial. In other words, testing for general intelligence can be useful to an employer, but only under certain conditions.
Because intelligence may be one of the best predictors of occupational success and because there is substantial variability in intelligence within groups of individuals with the same educational attainment, employers may also opt to use standard IQ tests in addition to level of schooling in choosing employees. To outsiders, the use of intelligence to select workers or trainees is fraught with problems. Regardless of any statistical relationship between measures of intelligence and job performance, there will inevitably be applicants who are mispredicted. This is also true of using Scholastic Aptitude Test (SAT) scores to predict who will succeed in college or Medical College Admission Tests (MCATs) to determine who will be successful in medical school; the future performance of some applicants will be underpredicted (i.e., they will achieve better grades than would be expected on the basis of their SAT scores), whereas the performance of others will be overpredicted (they will never live up to their high scores). No prediction based on intelligence is ever perfect or even close to it.
Given the fact that intelligence tests are imperfect predictors of employee performance, one may wonder why they should ever be used for selecting employees. Despite the fact that intelligence tests may be only moderately predictive of employee performance, mathematical models show that even a modest predictive value can lead to large economic benefits for employers. Increasing the probability that new hires may be more qualified for the job decreases the likelihood of attrition, lowers the amount of money that employers need to spend on training programs, and increases productivity. Thus, from the standpoint of employers, utilizing intelligence tests for the screening of applicants makes good economic sense under the conditions described above. Indeed, “utility analyses” that estimate the financial benefits of using tests in hiring decisions suggest that utilizing IQ tests to select employees translates into economic savings in the range of millions to billions of dollars.
In sum, both level of educational attainment and level of general intelligence (IQ) are related to employee earnings. The increases in employee earnings associated with increased educational attainment are likely the result of both the increased ability of employees with higher levels of education and the fact that higher-paying jobs have higher minimum-education requirements. Because intelligence predicts job performance and better performance is likely to be rewarded with increased salary, individuals with higher intelligence (and more schooling) are likely to earn more than those with lower intelligence (and less schooling). In addition, because performance on standard IQ tests also predicts employee performance, employers may opt to use IQ tests to select job applicants. Although these tests may not be perfect predictors of job performance, utilizing them can lead to significant economic benefits for employers, particularly in highly skilled fields where the applicant-to-openings ratio is large and the costs of poor hires are extreme.
- Card, D. 1999. “The Causal Effect of Education on Earnings, Volume 3A.” Pp. 1801-1863 in Handbook of Labor Economics, edited by O. Ashenelter and D. Card. New York: Elsevier.
- Ceci, S. J. and Williams, W. M. 1997. “Schooling, Intelligence, and Income.” American Psychologist 52:1051-1058.
- Heckman, J. J. and Kruger, A. 2003. Inequality in America: What Role for Human Capital Policy? Cambridge, MA: MIT Press.
- Hunt, E. 1995. Will We Be Smart Enough?: A Cognitive Analysis of the Coming Workforce. New York: Russell Sage Foundation.
- Morgan, S. L. 2005. On the Edge of Commitment: Educational Attainment and Race in the United States. Stanford, CA: Stanford University Press.
- Wagner, R. K. 1997. “Intelligence, Training, and Employment.” American Psychologist 52:1059-1069.