Investing in the school system will not pay off for decades

In the UK, the average premium a college degree adds to weekly earnings is between 12 and 22 per cent for men and 38 to 42 per cent for women
Click to follow
What's so great about education? The Government has driven home the lesson that this is its top priority, but the lecture has the ring of motherhood and apple pie. It is fair to ask, in a spirit of intelligent enquiry, how exactly education boosts the economy and how it enhances fairness and opportunity.

Evidence from across the Atlantic about past improvements in the US school system sheds fascinating light on how much advances in individuals' skills contribute toeconomic growth and changes in earnings inequality.* The conclusions are that the big E has played more of a part than previously supposed in establishing America's economic dominance, and it can make an enormous difference to earnings differentials. The drawback is that the effects have to be assessed over decades - this is no quick fix.

The research, by Claudia Goldin and Lawrence Katz at the National Bureau of Economic Research, looks at the spread of high school education in the early part of this century. Only about 10 per cent of young Americans graduated from high school - or preparatory school, rather - in 1910. By 1940 this had risen to 50 per cent.

During this period the modern secondary school was virtually invented in America. "By the 1950s youths in the United States were three decades ahead of their British counterparts in terms of high school enrolment and graduation rates," write Goldin and Katz.

Like all economics, this is a tale of demand and supply, the demand for and supply of the skills needed in the workforce. Demand for skill depends on the kind of capital equipment businesses need their employees to work with and how they work with it, so changes in demand are determined by technical developments and investment. Supply of skill depends on the education system.

The technical changes introduced during the Industrial Revolution tended to deskill the workforce, requiring factory fodder rather than artisans and craftsmen. By contrast, during the 1910s and 1920s, the spread of electricity led to a profusion of processes that required new skills, such as knowledge of chemistry and machine drawing. In addition, goods ranging from cars and radios to refrigerators and office calculatorsboosted the need for a wider spread of skills. The productivity gains permitted by the new technologies could not have been realised without improved education.

Yet the return to education in terms of the extra wage premium somebody could expect for completing high school fell sharply between the turn of the century and the late 1920s even as productivity and average earnings rose steadily. The return to a year of secondary education for the average male declined from about 22 per cent to 12 per cent. The reason was the more than commensurate increase in the supply of skills thanks to the high school movement. The decline in returns to education continued in the years after the Second World War, resulting in a very noticeable reduction in income inequality.

Goldin and Katz have recalculated the increase in the US's stock of educational capital between 1910 and 1950 based on new data. They find that the increase in years of schooling of the workforce was greater than earlier work indicated. "The increase in the US educational stock was a major contributor to American growth across much of the 20th century. Education's role in economic growth may have been larger than previously thought," they conclude.

The pattern has changed during the past 30 years. Productivity growth, in the US and elsewhere, has slowed sharply even though average years of education have continued to climb. At the same time incomes have started to become more unequal. There is widespread evidence that the returns to education for individuals are once more high and increasing. In the UK, according to research by the IFS**, the average premium a college degree adds to weekly earnings is between 12 and 22 per cent for men and 38 to 42 per cent for women.

The explanation for this pattern - slower growth, greater earnings inequality - is probably an increase in demand for skills unmatched by an increase in supply.

There is good reason to believe information technology requires cleverer people to work with it. But the supply of educational attainment has not increased in the same way as it did earlier this century.

The clear conclusion is that boosting the national investment in education will have the desired results. Past experience suggests a more skilled workforce does boost economic growth when technology changes in such a way that extra skills are required to make best use of it. In addition, boosting the supply of skills tends to reduce earnings differentials. The big increase in inequality that the US and Britain have suffered is the result of the restriction of educational improvements to a privileged minority.

Unfortunately, confirming the diagnosis does not make the cure easy. For instance, there is the issue of how to improve the education system and the quality of its output. The other important point is that these kinds of shift in levels of skill in the workforce take decades to achieve. If our school system improved substantially tomorrow the full effect would not show until today's infants joined the workforce in 2020. All credit to the Labour government for making education its top three priorities, but we will not see the results until well into the next century.

* Education, the Wage Structure and Technological Change by Claudia Goldin and Lawrence Katz, NBER Reporter, National Bureau of Economic Research, Spring 1997.

** Higher Education, Employment and Earnings in Britain by Richard Blundell et al, Institute for Fiscal Studies, details on http://www.ifs.

Returns to education in Britain

% increase in weekly wages due to higher education qualifications

Men Women

first degree higher degree first degree higher degree

max. estimate 22.2 17.7 42.1 53.3

adjusting 12.2 10.3 40.3 46.5

for other


* early ability, region, type of school, family background Source: IFS