The big business news of the week is IBM's abandonment of its conventional pension plan so that, going forward, its employees will bear the investment risk previously assumed by the employer-promisor of what had been a defined benefit plan. Viewed over the long term, we have seen a shift from no employer-provided pensions to defined benefit plans (promising retirement income, often in a manner based on final employment income and years of service) and now, apparently, to defined contribution plans (employer promises certain matching payments, perhaps, but employee bears risk of performance). Some of this evolution can be attributed to tax law (or perhaps itself caused changes in tax law); employment places with no pension plans sacrificed an obvious tax benefit, defined benefit plans allowed employers to benefit from appreciated investments and their favored tax treatment, and (now) defined contribution plans make use of 401(k) and other vehicles. But it is not entirely clear why employees would rather bear the risk of poor investment returns. In principle, the more attractive a pension plan is to employees, the lower the wage they need to be paid, so that employers ought simply to provide the best or the most tax-favored plan as viewed by a majority of, or in some industries perhaps marginal, employees. Some easy answers to the question of why we see some changes in the absence of legal change are that wages are sticky or that employees (especially young ones) underestimate the value of pension contributions or that they do not expect pension plans to be solvent when their retirement dates arrive or that they do not expect to stick around at their current jobs.
Might health care be next? For most of our history, employees could not get the same tax benefits if they took taxable wages and paid for their own medical care or health insurance. Employer-provided health care, or health insurance, thus had a huge advantage. The advantage is arguably smaller for retirees, and we have indeed found a lesser inclination to provide health care for retirees, and there have been recent employer withdrawals from the provision of this retirement benefit. Employer provided health care might also be more efficient - or less so. More efficient to the extent that a group can bargain with health care providers, but less efficient because of moral hazard. It might, for example, be efficient to have individuals pay for their own health care and then join groups (through insurance companies or employers) for catastrophic and expensive care. But the deductibility of health care costs when employer-provided, makes this less attractive than it might otherwise be.
Returning to retirement plans, even if tax law gives individual savers all the advantages it offers them through employer-provided plans, there is the fear that many individuals will simply save much less than they would have, or were forced to, through employers' defined benefit plans. Over time this might build up a stronger interest group, or voting bloc, in favor of more generous social security payments (and higher taxes). It might also bring about mandatory retirement savings plans, at the workplace or at home. Eventually, it is likely to bring about increased regulation of private firms that manage retirement accounts, as is currently occurring in Chile. At their core, then, pensions, health care, child care, and a number of other goods and services need to be thought of in a deep way (perhaps through explicit collective decisionmaking) as properly belonging in one of three sectors: government,employment, or household. Libertarians tend to like the last, but it is hard to get far from questions of government regulation (of insurance companies, say, and through tax incentives) once large numbers of individuals and organizations see where their interests lie.