Preston J. Miller - Former Vice President and Monetary Adviser
Published March 1, 1993 | March 1993 issue
With my tongue pressed only lightly against my cheek, let me propose a new law: Whenever any government program is defended on how it affects the number of jobs in the economy, a little red flag must be waved. The flag would alert the public that, although the program may change the composition of jobs, it is not likely to change the total number of jobs. In most instances defense of or opposition to the program based on jobs is a substitute for arguing for or against the program based on its particular merits and costs. Let me make my case by suggesting how programs should be justified, explaining how total employment is determined and illustrating how typical jobs-based justifications are misleading.
Government programs, or other market interventions, need to be justified by demonstrating that they will lead to improvements in the well-being of the citizens. As Murray Wiedenbaum, former chairman of the Council of Economic Advisers, put it, policymakers must be able to answer yes to these two questions:
The correctly applied principles behind these questions are that unless proven otherwise, freely operating markets can best deliver the goods. These principles also acknowledge the big difference between "can" and "will." Just because a government program can improve things, there is no guarantee that in practice it will.
Many government programs are not justified by applying Wiedenbaum's principles. Instead, they are justified because they will increase employment. In most of these cases that justification is not valid, and examining how total employment is actually determined helps to make that clear.
Over time, total employment in the economy is determined essentially by how many people want to work. That depends on the size of the population, cultural attitudes toward work, and policies, such as unemployment compensation, which affect the relative returns to working and not working. Over time, wages adjust so that firms are willing to hire those who want to work.
In the shorter run, economists disagree about what determines employment and whether countercyclical monetary and budget policies can do much about it. But those who do believe in the efficacy of these policies rely principally on these policies' ability to affect the economywide demand for goods and services.
Let me now turn to three policies that have been debated in terms of their employment effects and tell you why I think these discussions may at times have been misleading. The three are environmental policies, free- trade agreements and state/local industrial policies. None of these policies changes the real factors that determine the level of employment over time and none need change the level of demand in the shorter run. However, each would affect the composition of employment.
In recent years we have heard how strict environmental laws would either cost jobs or create jobs. Although it's true that they would cost some existing jobs, they would also add some new ones. But they would do little to the total. If we outlawed carbon fuels, we could still have everyone working, as they did in the 1600s. Many just would be making wagon wheels and buggy whips.
I am not belittling government's role in protecting the environment. There is a market failure in dealing with pollution. But when the government proposes a new environmental policy, it needs to demonstrate that the benefits of the programas it is actually implementedwill exceed the costs. Arguing about its employment effects is a distraction.
The North American Free Trade Agreement (NAFTA) is a second policy that has been debated based on its employment effects. Again, the policy would cause some people to lose jobs, but new jobs also would be created. There is no reason to believe total employment would be affected. Nations can decide not to trade at all and still have full employment, as some island economies have done, or nations can decide to trade freely and have full employment, as Great Britain did in the 1800s. Free trade allows more specialization and allows the total amount of goods and services to increase in the trading countries. Thus, freeing up trade would not change the total amount of jobs, but it would change the types of work people were doing.
I believe governments' policies of restricting trade are not justified when Wiedenbaum's two principles are applied. There is no market failure that warrants such government intervention. Thus, policies which move us toward free trade are beneficial. The real issue is whether NAFTA is moving us in that direction. Although NAFTA makes trade easier in North America, it makes trade more difficult between countries from outside and from inside the continent.
Finally, let me discuss state and local government programs designed to either create or save jobs. One such program by New York City used a system of incentives to keep firms from moving out of the city. The trouble was that subsidies to weak firms had to be paid for by higher taxes on healthier firms. Not surprisingly, a study concluded that the healthier firms moved while the weaker ones stayed, and, if anything, employment in the city decreased. This program also fails the Wiedenbaum test. An appeal to jobs once again sold a program that was not wise.
If my law is not passed, citizens should wave a red flag in their minds when they hear claims that government programs will create jobs. Citizens should ask themselves:
Can private institutions solve the identified problems?
If they cannot, will government involvement, in practice, make the outcome worse? If the answer to either question is yes, citizens should ask the government to wave a white flag and retreat.