One can view workers from the perspective of the portfolios they hold. For most workers, this portfolio, broadly defined, evinces a heavy concentration of assets in the firm that employs them. They stand to incur wage reductions if they move to some other company midway in the contract because much of the human capital that they hold is not fully transferable. In many cases, they also stand to lose substantial pension value, and may lose other benefits as well, such as health insurance, longer vacation periods earned through seniority and so on. This just means that they are party to a risky venture, where the downside outcome poses substantial losses. It turns out that the broad implications of bankruptcy for workers are well indexed by pension losses, which means that an understanding of the pension implications of bankruptcy also conveys the thrust of non-pension losses as well. Risk is not a free good. As in standard models of finance, firms are expected to pay for the risks they impose on workers, and indeed, are expected to pay premia in excess of those implied in financial models if they expose workers to non-diversified risks. The latter premium makes sense only if firms believe that heavily exposed workers are more likely to work hard to ensure the continued viability of their employer. If this incentive is sufficiently strong then higher productivity finances the additional risk premia. Workers can choose jobs among employers that pose different levels of bankruptcy risks, and those least willing to undertake risk take jobs with relatively low levels of it, and receive commensurately lower pay., As such, there is no reason to believe that workers’ losses in some bankrupt companies raise special public policy concerns. Most workers at risk experience the upside potential; a small percentage experiences the downside. There is no reason to believe that the contracts ex ante are inefficient. But if the government may have to assume responsibility for near destitute older workers, then this prospect might give rise to various mandatory insurance schemes. Even if these insurances are not market priced, the participants at least have to pay some of the costs of the exposure they incur. This may be one reason why institutions like unemployment compensation and pension insurance arise. The compensating differential model used to explain the risk-reward tradeoff inherent in defined benefit pension plans also can explain their demise. Beginning in the mid 1980s, the federal government altered its traditional bonding role in the contract and, instead, permitted firms to unilaterally terminate defined benefit plans and take excess assets (those in excess of termination benefits) into corporate profits. This had the result of exogenously increasing the probability of termination far greater than bankruptcy risks. Workers were (justifiably) more leery of the pension promise, which reduced the amount that they were willing to pay for it. The plans, therefore, are valued by less than it cost the companies to provide them, which creates the incentive to move towards pension plans that are self-bonded, such as 401k plans. 401k plans have dominated pension growth for the past twenty years and are fast becoming the pension plan of choice in the (non-union) private sector. While they eliminate the bankruptcy risks inherent in defined benefit plans, they create their own set of bankruptcy exposures. While most workers covered by these plans do not invest in company stock, about one in ten have at least half their account balances concentrated in the securities issued by the firm that employs them. In many large companies, 401k plans are heavily invested in company stock. These investment patterns can give rise to the potential for catastrophic losses for workers late in their career. In order to protect themselves from ex post bailouts, taxpayers may want to give some thought to enforcing some set of minimum diversification rules in these plans. Reprinted by permission of the publisher.