Page%1%of%16%% May 12th 2015 Do Workers Maximize Lifetime Earnings? • If we solve the ability bias problem, we can learn the return to schooling (on average by how much do earnings increase when schooling increases by 1 year). • However, this doesn’t tell us whether any given individual’s choices are optimal because of selection bias. • The idea of selection bias is that there is heterogeneity in the returns to a college degree. • And, it’s never possible to observe an individual’s returns since you never observe what he/she would have earned if he/she had made a different education choice. Selection Bias • For example, suppose ability is multidimensional. – Paul is a good plumber, but would make a lousy manager. – Marvin is a good manager, but would make a lousy plumber. – You need a college degree to be a manager but not a plumber. • Suppose that Marvin goes to college, but Paul does not. • Did they each make the right educational choice? %%ECON%139%%%SP%15 %%%Antonovics%%%7%%%5912915%%%1%Page%2%of%16%% • Observed college wage differential=WM-WP - Understates wage gain for Marvin - Overstates wage gain for Paul - Both may be making the right schooling choice, but it’s hard to know using the observed college wage differential. • In the model we just examined, workers who were relatively good at one type of job were relatively bad at the other type of job. • Contrast this with ability bias in which ability is one - dimensional. • Empirical work that attempts to correct for selection bias tends to find that people do make the right educational decisions. Social Standpoint • How much education is optimal from a social standpoint? • Concern motivated by 3 observations: 1. American workers are facing increasing competition from workers in other countries. 2. New technologies have increased the demand for workers with high levels of cognitive ability. 3. Although Americans spend as much per pupil as other industrialized countries on education, American students consistently score below students in other countries. Why Does Schooling Increase Wages? Two main theories: 1. Human capital theory: individuals acquire valuable human capital at school. -- Acquire useful skills at school 2. Signaling theory: schooling signals to employers that workers are more productive -- Employers need to know you are hardworking and determined, prove to the employers that you are able to suffer the 4 years in collegePage%3%of%16%%However, if it comes to what the government subsidy had helped, think of the first theory. The Signaling Model (Spence) • Two types of workers – High-ability workers: MP = 2 – Low-ability workers: MP = 1 High-ability worker twice productive than low-ability workers • Employers cannot observe worker’s type. • Education is less costly for high-ability workers – High-ability workers: c = e/2 – Low-ability workers: c = e High-ability worker twice less costly than low-ability workers • Let PVE denote the present discounted value of a worker’s future earnings. – In a perfectly competitive market, PVE equals the present discounted value of a worker’s marginal product. – PVEH=2PVEL Spence Model ¡ Now suppose that the firms implement the following policy: – If e≥e* then pay the worker a wage of $2 -- as high-ability worker – If e<e* then pay the worker a wage of $1. – as low-ability worker There’s no point for workers in getting schooling between [0, e*], because you end up getting$1 anyways. So either pick 0 or e*Page%4%of%16%%¡ Note that both high types and low types are being paid the present discounted value of their marginal product. ¡ What is the minimum level of schooling that will separate the high types from the low types? If e drops below e** both types will invest. Any < e** à separating two types of workers Summary of Key Features of Spence Model 1. Individuals have different innate levels of productivity. 2. Firms will want to pay workers their marginal productivity, but cannot observe productivity. Crucial. 3. Low-ability workers find it more costly to invest than high- ability workers. Crucial.Page%5%of%16%%4. Firm to set up “menus” of wage and education combinations such that they are able to sort out which workers are high- ability and which workers are low-ability. Thus, education increases wages because education serves as a signal of ability rather than actually increasing ability. Empirical Evidence on Human Capital vs. Signaling • Empirical evidence confirms that both components are important. • The debate is over how important. • Some Empirical Facts – The wages of college dropouts are approximately equal to the wages of high school graduates. Supports signaling. – College seems like a very costly way to get a signal of ability. Supports human capital theory. • Why do we care? – From an individual standpoint, it doesn’t matter. – From a social standpoint it does. – Subsidizing college education may not be worthwhile if signaling strong. Post School Human Capital Investments Basic Patterns in the data: • Highly educated workers earn more than less educated workers. • Earnings rise with age but at a decreasing rate. • The age-earnings profiles of different education groups diverge over time. – Earnings increase faster for educated groups Line up with Mince Model, earning increases even after people finish collegePage%6%of%16%%On-the-Job Training • General training: training that’s useful in many firms. Typing, taking blood pressure, making hamburgers, operating a forklift • Firm-specific training: training that’s useful in only one firm. Specific (firm specific) software, understanding particular administration policies • Profit maximizing condition with two periods of employment: Present discounted cost of hiring a worker has to be equal to the present discounted value of the benefit of hiring a worker. TC1 + TC2/(1+r) = VMP1 + VMP2/(1+r) • If training takes place in the first period of employment then: W1 + H + W2/(1+r) = VMP1 + VMP2/(1+r) VMP1 > VMP2 à training increases productivity Is Training Worthwhile? • Training is worthwhile as long as the cost of the training is less than the present discounted benefit of the