The study presents estimates of indirect losses caused by morbidity for three population groups: the currently employed, those unable to work (not in the labor force), and those keeping house. It is the estimates relating to the currently employed that are comparable to those presented in this article. There are, however, some differences in definitions.

The estimates in the study are based on National Health Survey data on work-loss days for currently employed persons. The National Health Survey, which is based on a household sample survey of the noninstitutional population, defines the currently employed as persons who worked at any time during the 2-week period covered by the interview and those temporarily absent from work because of temporary illness, vacation, strikes, or bad weather (if they would be expected to return to work when the event causing their absence ended).

The estimates of income loss shown in this article cover the first 6 months of workers' illness, including illness of workers no longer on a payroll, those with long-term or permanent disabilities, and those in institutions. When these differences in definition are taken into account, the estimate of $9.8 billion for the loss of earnings by the currently employed in 1963, shown in the study, seems consistent with the estimated income loss of $10.2 billion that is shown for that in table 1.

Long-Term Disability

year

For some time the Office of Research and Statistics has considered preparing a series of estimates for loss of income resulting from long-term illness and for the extent of replacement of such income loss, comparable to the short-term series in this article. These estimates for long-term disability would include the loss and replacement of income after the first 6 months of disability for persons who would be working if not disabled. Estimating income loss for long-term disability presents more difficult problems than estimates of the loss for short-term illness. Additional problems arise primarily because of the indefinite attachment to the labor force of persons with longterm disability. For example, the estimated time

lost from work for short-term illness can be based on the average number of days of work-loss for the working population. For the long-term disability estimates, this approach is probably not possible and it is necessary to start with an estimate of the total number of disabled persons in the population. This estimate must then be adjusted according to the presumed labor-force participation rates. The age and sex composition of the disabled population must be analyzed to exclude persons who would otherwise have been keeping house, going to school, or living in retirement. Finally, annual work time must be estimated and adjusted for average time ordinarily lost from unemployment, strikes, and similar events. For persons disabled during a year, adjustment must be made to include only the workloss time after the first 6 months of disability. The institutionalized population, nearly all of whom have long-term disabilities, must be taken into account.

Other difficulties concern the earnings valuation to be placed on lost time for the long-term disabled. The income loss attributable to short-term sickness can probably best be based on current wage levels. For the "would-be" workers among the long-term disabled, consideration should be given to the possibility that the past earnings levels of these disabled were different (probably lower) than those of the active work force.

A start toward preparing income-loss estimates of this type for long-term disability is provided by the ORS series on the number of "would-be" workers who have been disabled for 6 months or more.2

The Public Health Service report on estimating the cost of illness, especially the sections concerned with persons "unable to work," presents additional methodology useful in estimating costs attributable to long-term distributions.3

The National Health Survey data provide much information needed to develop various components of the estimates. The 1966 Survey of Disabled Adults, now being conducted by the Social Security Administration, should also yield considerable additional data on the characteristics

2 Alfred M. Skolnik, "Persons Receiving Payments From Public Programs for Long-Term Disability," the Bulletin, October 1964.

Dorothy Rice, op. cit.

of the institutional and noninstitutional disabled population.

PROTECTION AGAINST INCOME LOSS

which are similar to the unemployment insurance laws of those States, cover most employees in industrial and commercial firms. They generally do not cover hired farm workers (except in California), domestic service workers, or employees of governments and nonprofit organizations.

Types of Protection

Most of the protection against loss of earnings from short-term nonoccupational disability is provided through the worker's place of employment. Some employers insure their workers against this risk by purchasing group policies from commercial companies under which cash benefits are paid during specified periods of disability, or they provide similar benefits by self-insuring. Others establish formal paid sick-leave plans that provide for continuation of wages (usually full wages) for a certain number of days. Still others combine the two methods and establish both sick-leave and group insurance plans that supplement each other.

Among other sources of employment-connected protection against income loss resulting from sickness are mutual benefit associations and union or union-management plans, often on a regional or industrywide basis. Workers and self-employed persons may also obtain protection through the purchase of individual sickness insurance policies from insurance companies or through membership in fraternal societies.

In California, New Jersey, New York, and Rhode Island, most employees are covered under a State temporary disability insurance law, and workers in the railroad industry are protected under a Federal sickness insurance law. In Rhode Island and the railroad industry, all benefits are provided from publicly operated disability funds. In California and New Jersey, employers have the option to "contract out" of the public plan by providing an approved private plan, usually one insured by a commercial company or financed on a self-insured basis. The New York law requires employers to provide sickness protection of a specified value for their employees by establishing a privately insured or self-insured plan or insuring with a State fund that itself has many characteristics of a private carrier. In California, New Jersey, and New York, union or union-management plans may provide the sickness benefits required by law. The coverage provisions of the temporary disability laws in the four States,

Number of Workers Covered

The extent of protection against short-term sickness is, as would be expected, considerably greater in the four States with temporary disability laws and the railroad industry than in other areas. As the following tabulation shows, of the wage and salary workers in private industry in the five jurisdictions with temporary disability laws, about 96 percent had some protection (including a small number of workers not covered under the laws but provided sick-pay protection by their employers). In the areas without laws, about 49 percent of these wage and salary workers had some type of protection.

Most of the 32.6 million workers with protection in 1965 were covered by group policies written by insurance companies. About 22.6 million workers were covered by private commercial insurance, including 6.4 million under plans written in accordance with the State temporary disability insurance laws. The five government-operated funds provided protection for 6.5 million workers; union, union-management, and mutual benefit associations for about 1 million. The remaining workers were covered exclusively under formal sick-leave plans.

The extent of coverage under temporary disability insurance laws has not changed substantially since 1949 when the State of New York enacted its disability law. The Rhode Island law was enacted in 1942, the California and railroad

legislative programs in 1946, and the New Jersey law in 1948. No other laws establishing programs have been passed, but coverage in the existing programs has been extended to additional small groups of workers in some of the five jurisdictions by later legislation.

Interest may therefore be focused on the developments in the jurisdictions without temporary disability insurance laws, where the sickness protection is provided on a voluntary basis. The extension of sickness insurance coverage was greatest in the immediate postwar period, a period also marked by expansion of other types of employee benefits under health, welfare, and pension plans, for example. In the last decade, however, the extent of sickness benefit coverage has hardly changed, as indicated by the data below showing the number and proportion of private wage and salary workers (excluding railroad employees) covered by cash sickness programs in the States. without temporary disability insurance laws. These data exclude persons with protection only under group credit insurance arrangements since this type of insurance does not generally stem from an employment relationship. (Credit insurance is insurance purchased by lending institutions to protect their loans against the risk of nonpayment because of the borrowers' disability.) The number of workers with sickness protection has risen somewhat since 1954, but only in keeping with increases in the private labor force. The proportion of workers covered has not increased to any appreciable degree.

Insurance Plans and Sick Leave

Both of the two major types of protection against income loss-insurance plans and paid sick-leave plans-replace income that would otherwise be lost, but the method of replacement is of course quite different. Sick-leave plans usually

1 Beginning 1960, data include Alaska and Hawaii.

2 Data on premiums earned and losses incurred by commercial companies (including fraternal) as provided by the Health Insurance Association of America for the United States, by types of insurance benefit, adjusted to include accidental death and dismemberment provisions in individual policies that insure against income loss to offset understatement arising from the omission of current short-term income-loss insurance in automobile, resident liability, life, and other policies. For 1956-65, dividends deducted from earned premiums (2-3 percent for group: 1 percent for individual). Starting with 1956, all credit accident and health insurance classified under individual insurance.

3 Union-management trust fund, trade-union, and mutual benefit association plans.

'Company, union, and union-management plans under California, New Jersey, and New York laws.

period of 1 week. Some private plans use a shorter waiting period, perhaps 3 days, and may start benefits on the first day in case of accident. The benefit is usually a stated percentage of the worker's recent wages, often one-half to two-thirds, but is usually subject to some specified maximum

amount.

Each of the two types of protection has advantages and disadvantages. The sick-leave plans offer "first-day" benefits and usually full pay but may provide little protection in cases of more extended illness or disability, especially if the sick leave is not cumulative. The insurance plans usually provide better protection for the lengthier illness or disability that often imposes the greatest financial hardship. If the plan pays 26 weeks of benefits, it will usually maintain a certain income for the worker until programs geared to long-term disability start paying benefits. Insurance plans of course give little or no protection for the most frequent types of illness-those that last only a few days and then provide only partial wage replacement.

The fact that most illness lasts only 1 day, or a few, is important in interpreting the data on income replacement under sick-leave and insurance plans. Data on sick leave show replacement of a high proportion of income, probably about three-fourths for a typical group. Insurance plans show a much lower percentage, perhaps 20-40 percent under most plans. Because of the difference in the kind of protection offered, however, the extent of wage replacement is not necessarily an adequate measure of the comparative advantages of the two types of plans.

AMOUNT OF SICKNESS BENEFITS

Table 2 shows the amount of insurance protection against the risk of short-term illness provided through insurance companies and other private organizations. It includes separate data on private insurance written under voluntary arrangements and that provided in compliance with temporary disability laws in California, New Jersey, and New York. As for previous years, credit accident and health insurance has been included in the individual insurance category. Data on sick-leave plans and, in States without compulsory laws, on self-insured, unfunded employer-administered

plans are excluded from table 2 but are included in table 4.

In the voluntary segment, group insurance continued to show considerable growth with an 8.3percent increase in benefit payments to $543 million, following the rise of 10.4 percent in 1964. These increases represented substantial improvement over previous years since the average annual compound growth rate for the 4-year period 195963 was 3.7 percent. However, the other major voluntary insurance category, individual insurance, changed comparatively little in 1965.

While sickness insurance written under disability insurance laws increased somewhat, as discussed later, the growth was relatively greater for voluntary group insurance. Of the total of $666.7 million in group commercial insurance benefits paid in 1965, 18.5 percent was paid in accordance with statutory provisions, compared with 19.3 percent in 1964.

Public Provisions

Table 3 shows the benefits paid under the five temporary disability insurance programs. To the extent that the protection under these programs is provided through commercial companies or other private arrangements, the data in table 3 overlap those in table 2.

In 1965, total benefits under the public provisions increased only slightly, by $10.9 million or 2.4 percent, to a total of $466.7 million. One factor affecting the growth in these benefits has been the steady decline in benefits under the Federal program for railroad workers that reflected the decline in railroad employment. Sickness benefits under the railroad program reached a peak of $66.2 million in 1959 and has subsequently fallen in each year. The further decline of $5.2 million in 1965 reduced the railroad payments to a low of $40.8 million.

The year 1965 saw the reversal of the trend that began in 1958 toward declining participation of private plans in the temporary disability insurance programs. From 1957 to 1964 the share of benefits provided by private plans fell from 58.3 percent to 42.0 percent, with a corresponding increase in the part paid directly by State funds. The proportion paid by private plans rose slightly in 1965, however, to 42.3 percent of the total benefits paid.

The earlier decline in private-plan participation was caused mainly by the declining importance of insured private plans in California. These plans had almost disappeared by 1964, providing only a fraction of 1 percent of the benefits paid under the California law in that year. In New Jersey the participation of insured private plans has shown small relative declines in recent years. In 1965, there was little change in the amount of benefits paid by insured private plans in California and New Jersey and a moderate increase in New York, where private plans continued to dominate the program. Self-insured plans in these three States have maintained, in recent years, a relatively stable position as providers of benefits under the disability laws.

The disappearance of insured private plans in California is attributable mainly to a series of legislative enactments that affected the financial experience of and the relationship between the State fund and private plans. A comprehensive analysis of these events is given in a recent study of the California program.*

The 1946 legislation that established the temporary disability insurance program provided for the participation of private plans meeting certain requirements of law. The plans were required to pay benefits in some respect better than those provided by the State fund. The benefits of the public and private plans were to be financed by employee contributions of 1 percent on the first $3,000 of annual wages. In practice, a close and somewhat delicate relationship exists between the State fund and the private plans; changes in the provisions or experience of either often have important effects on the other.

The State fund began its operations in 1947 with a substantial balance transferred from the unemployment insurance account. Also, in the early years the contributions income of the State fund greatly exceeded benefits and the fund accumulated large reserves. This "over-financing" was no doubt an important factor in encouraging the California legislature to increase benefit amounts on several occasions, without making provision to finance the additional costs. By the late 1950's the State fund was sustaining substantial annual losses and was rapidly being de

'Nathan Sinai, Bert S. Thomas, Benjamin W. Wheeler, Disability Insurance in California. School of Public Health, the University of Michigan, 1965.

1 Programs under the Railroad Unemployment Insurance Act and the laws of Rhode Island, California, New Jersey (beginning 1949), and New York (beginning 1950). Excludes hospital benefits in California and hospital, surgical, and medical benefits in New York.

? Under the laws of California, New Jersey, and New York.

Employers may self-insure by observing certain stipulations of the law. Includes some union plans whose provisions come under the law.

Includes State-operated plans in Rhode Island, California, and New Jersey, the State Insurance Fund and the special fund for the disabled unemployed in New York, and the railroad program.

pleted. The private plans, which at their peak covered about half the covered workers, were hard-pressed to provide the required benefits, at the same 1-percent contribution rate, for many of the employee groups. As a result, plan coverage declined. The financial problems of the private insurers probably caused them to limit their selection of employee groups to the best risk groups those with the lowest disability rates. This action, in turn, aggravated the financial problems of the State fund.

Legislation enacted in 1961 raised the taxable wage base and made other provisions designed to restore the fund to actuarial balance, but later experience indicated that these revisions were not adequate. One of the 1961 provisions tightened the requirements concerning selection of risks by private insurers. These new requirements, combined with the other difficulties, resulted in the virtual disappearance of private plans.

In 1965, the California legislature enacted amendments to the disability law that raised the taxable wage base to $7,400 annually-the highest base for any social insurance program in the United States and provided a small, temporary increase in tax rates. They also froze at $80 week