'Critical industries' and the pork barrel
Even if it were possible to objectively identify "critical industries", which is doubtful, political forces would likely turn any organized attempt at industrial policy into a pork barrel. There is no denying that many government procurement programs, especially Department of Defense procurement, have fostered the development of key private industries. There is no denying that subsidization can help an industry develop. The question is whether we can accurately determine which industries are "undervalued" by the private market. The debate over which industries offer the greatest potential for spillovers is likely to have few objective criteria. And criteria will be plentiful enough that other considerations, such as electoral calculations, might come to play an increasingly important role in industrial policy.

The natural political outcome we would expect is that large industries get the most support. Large industries are easily defined as critical since, by definition, many people depend on them for supplies and employment. But large industries are not always the most deserving of new capital. Success in the past does not guarantee high rates of return in the future. Industrial policy could easily lead to industrial stagnation. Government industrial policy in Japan and Korea has probably played a significant role in the problems we have witnessed in those countries in recent years. Investment rates have been very high, but capital keeps getting funneled to low return activities. While an industrial policy might work well in the short run, when there might be agreement about the industries that are most essential, the problems with an organized industrial policy would grow more severe with time as needs changed.

In summary, to say that defense spending created some new and valuable private industries does not imply that industrial policy will provide good value. The amount of industry development that was generated per dollar of spending was probably quite small. It is not enough to say that the market is not perfect in allocating capital in order to justify industrial policy. One needs to be able to believe that government allocation of capital would be more efficient than private allocation. History teaches us many sobering lessons on this point.

Current strength of US economy
The U.S. economy appears remarkably strong entering the new millennium, based on traditional indicators such as GDP growth, unemployment, employment, market capitalization, and inflation. Everyone has seen the glowing articles about the current state of the overall U.S. economy, which has been variously described as a "Miracle Economy", a "Jobs Machine", and other colorful monikers. Based on those descriptions alone, we can be pretty sure that on average, things look quite good for the U.S.

To say that things are good on average does not mean that every sector and every worker are benefiting from economic expansion. There is variation around the mean. Plant closings still happen, workers are laid off or otherwise unemployed, and families do suffer. It is impossible to determine the separate contributions of technological change, international trade, government policy, and changes in consumer preferences in generating these adverse outcomes for individuals. More importantly, a full employment economy implies that these adverse outcomes are only one side of the coin. The other side is the growth of the new industries and new jobs that take place in other sectors.

Conclusion
In summary, there are three important questions in thinking about the role of trade in the US economy are as follows. First, what level of foreign indebtedness or foreign debt service would constitute an undue risk for the U.S. economy as a whole? Where is the danger zone? Second, what factors are most likely to be contributing to our ongoing indebtedness? Perhaps people are saving too little because they overestimate the value of their Social Security benefits. We should keep our focus on the incentives to save, rather than the incentives to trade. And most importantly, how can we better assist people who are hurt by industrial transformations-whether they are precipitated by trade, technology, or something else? To prevent transformation that is driven by either trade or technology would be to throw the baby out with the bath water. Supporting the continuation of declining industries is not a desirable solution for protecting jobs, since such support comes at the expense of emerging industries which offer higher returns to workers and investors alike.