3
THE UNIVERSITY OF CHICAGO
Graduate School of Business
Business 33032 Canice Prendergast
Spring 2008
Topic 2
Human Capital and Education
Most of the course is on the issue of how workers should be compensated. The primary determinant of someone’s pay is their productivity, largely reflecting their human capital. The objective of this section of the course is to understand the acquisition of human capital, who should pay for it, and how the returns to human capital have changed over the last number of decades.
We will do this in a number of steps. We being by simply providing some facts: how much more do people with different skills make, and do these differences depend on where you are in your career? Following this, we will describe a simple model of human capital collection, where we will treat the collection of human capital in exactly
Human capital refers to the intellectual capital that individuals hold and perhaps the bottom line of the lecture is that we should see human capital like any other form of capital. There are really two kinds of human capital that we will consider:
1. Formal education programs
2. On-the-job training programs
For the first part of the lecture I am going to consider formal education.
I: Some Evidence on Human Capital
Begin by looking at Figure 1. This measures how wages earned vary for workers of different education. There are a few things worth noting:
1. People with more education earn more (hardly surprising)
2. Age-earnings profiles are concave
3. People with more education have steeper age-wage profiles – in other words, much of the benefit of education is deferred.
Figure 1
Total Money Earnings (mean) 1981
Source: U.S. Bureau of the Census. Money Income in 1981 of Families and Persons in the United States, Current Population Reports. Series P-60, No. 137 (1983) Table 48.
What we want from our theory of skill collection is something that will help explain these phenomena. To do this we turn to a model of human capital acquisition.
II. An Investment Rule
To begin, we will treat human capital as if it is any other asset. What this means, in practice, is that we can ignore the concern that a firm pays for the skill collection of the worker, but where there is a danger of the worker quitting and taking her skills with her. Instead, what we will assume here is that the party that pays for the skills gets their benefits, and deal with cases where this may not be true below.
Let us deal with the human capital decision in terms of doing an MBA compared to finishing with a college education.
10
Suppose that if you do not get the MBA your income will be ytC in period t (C for college). Alternatively you can decide to go to college and earn ytM in period t (M for MBA). However, you also have to pay K dollars for a single year to do the MBA. Assume that the worker with the college degree works from year 0 to year TC, while the worker with the MBA works until TM. Why might TC be different from TM?
The present value of just going to college is
while the present value of doing the MBA is
To keep matters simple assume a similar retirement date for both workers. Then the worker should go to college if
An Example:
Assume that T = 2 and the wages earned are
No MBA: 30,000 and 34,000
MBA : 10,000 and 100,000
Let the interest rate be 10% and let the cost of an MBA be $35,000. Is the MBA worth it to the worker?
Observations:
10
1. Do people actually use this model when choosing to go to college?
2. The reason that we do a model like this is to get an idea of the factors which are likely to be important when people try to decide whether to get an education. What factors are important here?
What recent changes in the United States are likely to affect on skill collection?
1. Mandatory retirement being made illegal.
2. Birth rates have declined, so the position of women in the human capital market has changed.
The next thing that we do is to determine the internal rate of return on education. This is the critical interest rate, which if the actual interest rate exceeds this, you should not get an education. How would you compute the rate of return? Clearly, it is the interest rate r at which the costs and benefits are equalized.
Example Continued
Suppose that, again, t = 2 and it costs $100 to get an MBA. Wages are $50 higher for the MBA while the MBA is being done and $55 higher in the following year. What is the internal rate of return for education here?
So what if the interest rate is 8%? Should you get the education? How does r change with the cost of education, later retirement, age of the worker, etc.?
10
Estimates of r are in the neighborhood of about 7%, which is much the same as in other assets. Is this just coincidence? Probably not, as people have choices about what assets they should place their wealth in, and the returns should be equalized if the market is reasonably competitive. (This assumes, of course, that the misery of doing an MBA is no greater than the worries of holding stock, for example.)
Caveat: Suppose that you read in, say, The Economist, that the wages of college- educated people were 50% higher than the wages of high-school-educated workers. Could I extrapolate from there that the rate of return from education is about 12%, as college takes 4 years?
So you have to be very careful in interpreting data; always beware of self-selection. In the context of education, we will return to this later in the course, when we talk about signaling.
At the beginning of the notes, I mentioned that there were a number of different characteristics of the data that we would like to explain. Among these were (i) earnings profiles being concave and (ii) those with more education having steeper age-earnings profiles. How can we explain these with our theory?
10
III. Changes in the Return to Skills over Time
The next thing we look at is how the returns to college education have changed over time. To see this, consider Figure 2. This measures the extra income earned by a college graduate compared to a high school graduate. There are two lines here; (i) all college grads versus all high school grads, and (ii) similar figures for those with only 1-5 years of experience. Why is the picture for those with 1-5 years of experience so much more volatile?
There are two main findings from this graph:
1. A big decline in the college premium in the 1970s
2. A big increase since the 1980s
Why?
The next piece of data that is interesting is to look at is how income distribution has changed over the last 20 years. This is given on Figure 3. This shows how the income of those in the 0-10 decile (the poorest 10%) has changed over time and similarly for those in different parts of the income distribution. The most striking result is that those in the lowest part of the distribution have had their wages reduced by 25-30% in real terms over the last 20 years.
The role of trade in explaining the increase in inequality:
What % of US GNP is traded?
How much has this increased in the last 20 years?
How much of US trade is with Less Developed Countries?
How much has trade with LDCs increased in the last 20 years?
10
IV. On-the-Job Training
Much of the human capital that all of us collect has nothing to do with formal training programs. Instead it is made up of on-the-job training programs, where workers collect skills while they are employed.
Examples:
(i) Training by senior workers or by doing
- how to operate machinery - how to deal with particular clients
- how to process forms
(ii) Formal training programs - computer courses
We have to distinguish between two forms of training: (i) general training and (ii) specific training.
General training: this refers to skills that are useful in many companies, such as an MBA.
Specific skills: this refers to skills that are useful only as long as you remain with your current employer. Can you think of examples of specific skills?
The key distinction between these two forms of training is the bargaining power that it gives workers. This is what generates the problem for firms, as workers who have greater bargaining power may extract some of the returns to skill collection.
One must make a number of considerations when designing training programs. These are (i) how do I pay the worker? and (ii) who pays for the training? We will see that these two are really the same question and that the type of skills collected plays an important role in determining the answer to this question. Consider a general case where a worker is trained in period 0 and then works until period T. There is a cost k to training the worker and his marginal productivity in period t is yt. I am going to ignore the issue of whether it is efficient to train the worker, instead assuming that it is. The question we address here is how the worker should be paid. The total value of the worker’s productivity is given by
Let us consider a case where the firms are in a competitive market and earn no profits (the basic idea here is robust to allowing this assumption to drop). Then is the worker earns a wage wt in period t the firm designs its wages such that
10
The question we address is how the worker should be paid over his lifetime.
A. General Training
Here the worker can use the general nature of his skills to get a wage that is close to his productivity in the firm. Let’s consider the extreme case where there are no moving costs and the skills are completely general. Then the worker can earn yt = wt in any period. To put this in more standard terms, if the worker’s skills are general, he can bid his wage up after training so that the firm cannot earn much money from him after training.
Does this mean that firms are unwilling to train workers in general skills? (Key point)
Hence one solution to the problem associated with general skill collection is to offer a wage schedule that increases rapidly with age. In effect, you are making the worker pay for his own skill collection by offering a low wage while he trains. An example of this is apprenticeships, where trainees get badly paid but are well paid after they finish their training.
What can firms do when this is a significant problem?
B. Specific Training
Here the firm is not so worried about the worker using his skills to bid up wages, because the skills he collects are useless elsewhere.
Here the firm has more discretion. It could offer a wage profile like that for a generally trained worker or it could offer a flat wage profile.
What kind of wage profile should you offer?
There are a number of influences:
1. Can workers pay for their skills?
2. Should workers pay for their skills?
3. How likely is a layoff relative to a quit?
Suppose that a flat wage profile is offered. Then the firm is making a lot of money from the worker after training but the worker is earning little extra. Hence there is a danger that the worker will quit. One solution to this is to offer a wage profile that increases after training. However, if the wage profile is very steep, the worker may be in danger of being laid off by the firm. If this is the case, one solution is to offer a flat wage profile, as the firm is then earning money from the worker as he ages, and hence is unlikely to lay him off.