Abstract

METHODS

The data source for our analyses was the 2009 Bay Area Employer Health Benefits Survey, which was conducted through telephone interviews with employee benefit managers. The survey collected information regarding 2009 sick leave policies and changes made in response to the 2007 ordinance to describe pre-2007 offerings and postmandate changes. We conducted all analyses at the establishment level.

The survey included interviews of 727 firms in San Francisco in 2009, with a response rate of 20%. We calculated the response rate using the Council of American Survey Research Organizations’ method. The sampling frame consisted of all for-profit San Francisco firms with more than 20 employees and a random sample of firms with 5 to 19 employees. We used data from the City of San Francisco Office of Labor Standards Enforcement and a larger sample that completed 1 question on health insurance offered to test for nonresponse bias in the sample. Although the low response rate is a concern, nonresponders were similar to the completed sample along observable characteristics,37,38 and all results are weighted for nonresponse. We created employer weights on the basis of firm size at site, profit status, location, and standard industrial classification industry group to make the sample representative of the population of firms in San Francisco.

Surveyed firms were asked details about their leave policies, changes to leave policies, other benefits, and business operations in response to the PSLO and difficulties understanding and complying with the PSLO. In addition, the survey collected information on firm industry, size, profit status, and employee characteristics.

An additional sample of firms with 20 or more employees in counties surrounding San Francisco were also interviewed (n = 282). These firms were only asked a subset of questions, as they were not subject to the PSLO. We have reported means from these counties (Alameda, Contra Costa, Marin, and Santa Clara) to compare sick leave offer status.

We divided the sample into 3 groups by employer size—small (< 20 employees), medium (20–99 employees), and large (≥ 100 employees)—and measured reported changes in sick leave offer status and type of leave offered in 2009 by employer size. Next, we grouped firms by changes they made in response to the PSLO: implemented a new paid sick leave policy because it did not have a policy before, made a major change in sick leave policy (increased accrual rates or expanded eligibility), or made no changes to sick leave policy. We further divided firms that did not make any changes into those that do and do not offer sick leave. We excluded firms from analysis of changes in sick leave if they did not know their sick leave offer status before the PSLO or they refused to answer (n = 28).

We described the characteristics of firms in each of these groups by creating indicators for a high proportion of part-time, female, Black, Hispanic, and new (hired in the past year) employees. We defined firms as having a high proportion of a worker type if they exceed the 75th percentile of San Francisco firms. We created indicators for whether a firm employs workers at less than $10.00 an hour (the minimum wage in San Francisco in 2009 was $9.79 an hour), whether the firm employs temporary or unionized workers, and whether the firm is nonprofit or part of a multiestablishment chain. We analyzed differences between firms with these characteristics and firms without these characteristics using the unpaired 2-sample t test.

We have described how complying with the PSLO affected employees, employers, and customers, according to employers’ reports, stratifying by whether the firm instituted a new policy, made a major change to an existing policy, or previously offered sick leave and did not make any changes. We used a 2-sample comparison of means to statistically test differences in responses of the group with a new or substantially changed policy and the group that already offered sick leave and did not make any changes to comply with the PSLO. Our a priori hypothesis was that the largest reported changes would be in the group that began to offer sick leave after the ordinance was implemented. Finally, we explored the association of firm size with effects on employees, employers, and customers. This characteristic was a focus under the hypothesis that small firms might be more likely to reduce other forms of compensation or have more difficulty implementing new leave policies.

RESULTS

In response to the PSLO in San Francisco, the proportion of firms with a sick leave policy increased from 73% to 91%, indicating that 18% of firms instituted a new sick leave policy between 2006 and 2009 (Figure 1a). These gains were concentrated in firms with fewer than 100 employees, where the proportion of firms offering sick leave rose by 20% in contrast to large employers whose coverage increased by only 9%. In 2009, 99% of firms with more than 20 employees offered paid sick leave in San Francisco. The corresponding rate for firms with more than 20 employees in counties surrounding San Francisco was 84% in 2009. In San Francisco, 52% of firms complied with the mandate by offering separate vacation and sick leave, 33% offered paid time off that could be used for either vacation or sick leave, and 6% offered sick leave only (Figure 1b). These proportions differed by firm size: larger firms were more likely to offer separate vacation and sick leave. Among the 9% of firms that did not offer sick leave, 1 in 5 did offer vacation time. The few remaining firms in San Francisco that did not offer sick leave in 2009 were likely not in compliance with the law.

Open in a separate window

FIGURE 1—

San Francisco sick leave offering by (a) firm size in 2006 and 2009 and (b) type of leave offered in 2009: Bay Area Employer Health Benefits Survey, California.

Note. PTO = paid time off. Offering sick leave includes separate vacation and sick leave, sick leave only, and paid time off that can be used for vacation or sick leave. We used employer weights on the basis of firm size at site, profit status, location, and industry to make the sample representative of the population of San Francisco firms.

Furthermore, San Francisco firms offering paid sick leave make the benefit eligible to a larger share of employees than did firms outside San Francisco. Within San Francisco firms offering paid sick leave, 76% of employees in small firms (< 20 employees), 91% of employees in medium firms (20–99 employees), and 86% of workers in large firms (≥ 100 employees) were eligible for the benefit. In firms that offered sick leave benefits outside San Francisco, 73% of employees in medium firms and 70% of employees in large firms were eligible for the benefit (small firms were not surveyed; medium and large firm comparison: P < .001).

In addition to the group that initiated a new sick leave policy, 17% of firms made a major change to their sick leave policy to comply with the PSLO (Table 1). These changes included increasing accrual rates or expanding eligibility to employees who did not previously qualify for sick leave (including shortening a waiting period to accrue sick leave or expanding to additional employee classes). More than half of firms (57%) did not make any major changes to their existing paid sick leave policy in response to the mandate.

Firms in the accommodation and food services sector, a low-wage sector in San Francisco,39 were significantly more likely to have added a new sick leave policy in response to the PSLO than was the average San Francisco firm (27% of accommodation and food services vs 17% in other sectors [P = .03] and 35% of restaurants [a subset of the accommodation and food services sector] vs 16% in other sectors [P = .001]). Firms that employed a high proportion of part-time workers, new workers (hired in the past year), and Hispanic workers were less likely to offer sick leave both before and after the law. However, they were also significantly more likely to have implemented a new policy to comply with the mandate. Nonprofits, firms that were part of a multiestablishment chain, and large employers (≥ 100 employees) were significantly less likely to implement a new policy after the PSLO but were much more likely to have offered sick leave in 2006. Firms with unionized workers had a baseline sick leave offer rate similar to that of nonunionized firms. However, by 2009 nonunionized firms were significantly more likely to offer sick leave (92% vs 80% in unionized firms; P = .011). Firms with a high proportion of temporary and part-time workers were significantly more likely to have made a major change to sick leave policy, perhaps to extend eligibility for benefits to these workers.

DISCUSSION

We have provided evidence of employer-reported responses to and difficulties with implementing a paid sick leave ordinance. The San Francisco PSLO increased the offer rate in firms that did not offer sick leave in 2006: 18% of all San Francisco firms implemented a new policy in the first 2 years after the PSLO (66% of those that previously did not offer a policy). Most firms (57%) did not make any changes to comply with the law. A small proportion (9%) did not offer paid sick time in 2009 and were likely not in compliance with the law. The remainder (17%) made minor changes such as increased accrual rates or expanded eligibility to comply with the PSLO. These findings document that a paid sick leave mandate can increase workers’ access to paid, job-protected illness leave.

Our findings are consistent with the expectation that a greater proportion of firms with part-time or low-wage workers would have to implement a new sick leave policy. Smaller firms were also less likely to offer sick leave at baseline (and after the mandate went into effect). The group that implemented a new sick leave policy in response to the PSLO reported stronger effects on business outcomes than did firms that did not make any changes. Thirty-nine percent of these firms reported reducing some form of compensation, 49% reported having difficulty complying, and 32% reported worse profitability. In addition, firms with new policies were more likely to report adverse effects for consumers, including raising prices and worse customer service. These proportions were all significantly higher than those reported in the group making minor or no changes to comply. Nevertheless, support for the policy was consistently high (> 70%) across all firms, including those that implemented a new policy. Additionally, the group implementing a new policy reported increased employee morale.

Although 58% reported some difficulties (understanding, administering, or scheduling) during the first 2 years of implementation, most businesses did not report worse overall outcomes. Although only 19% of firms reported any reduction in profits, the proportion of firms was larger (32%) among those that had to implement a new sick leave policy. This implies that even in this high-impact group, the majority of firms did not report lower profits because of the mandate. Moreover, this occurred during the recent economic downturn, when profits were generally falling. This might have led to an overreporting of profit reduction attributed to the policy. For example, 14% of those that did not make any changes to sick leave policy also reported worse profitability.

Why might the adverse effects of the mandate be small? First, some firms modified the composition of compensation: 16% reported reducing some other benefits. The proportion reducing vacation time or pay raises and bonuses was (as expected) higher for firms implementing a new policy (39%). This shows that firms usually have the flexibility to adjust compensation policies to meet the requirement without adding additional cost to their bottom line, confirming previous findings.35 Second, some firms passed on any added costs to customers (8% of all firms and 18% of those implementing a new policy). This is consistent with evidence from employer mandates in San Francisco and elsewhere.39–41 Third, there may be beneficial effects on productivity. The increase in employee morale (especially among those newly implementing a policy) is consistent with that perspective. Finally, the fact that San Francisco firms overwhelmingly support the ordinance—72% of businesses overall and 71% of businesses implementing a new policy—substantiates the idea that any adverse effects on businesses were small.

San Francisco’s PSLO guarantees accrual and job-protected use of paid sick days for workers’ and their families’ health needs. Similar policies have now been enacted in many jurisdictions, and Congress and policymakers in 20 states and cities are currently evaluating paid sick days policies.42 The movement to ensure minimal access to paid sick leave has been likened to the campaign to enact the minimum wage: an effort to establish a floor below which no employer or worker may fall. When paid sick leave policies are targeted at vulnerable workers, such as mothers earning low wages, and workers who have a lot of face-to-face contact with the public, such as restaurant employees, these campaigns present a compelling image.43,44 Congressional proposals to create emergency paid sick leave policies to reduce the spread of the H1N1 virus in fall 2009 cited the potential importance of paid sick leave in protecting public health, and public opinion polling shows very high levels of support for paid sick leave policies.45,46

Although a majority of San Francisco firms (73%) offered some kind of sick leave before adoption of the ordinance, workers gained a significant employment right in having access to paid sick leave established as a legal requirement, with protections against punitive responses from employers for use of that leave. This policy institutionalization can indemnify workers against cuts in voluntary benefit provisions, such as those occurring in an economic downturn. In some cases the policy can also lead to reductions in other forms of compensation. Survey results show that in 2009 among firms with 20 or more employees, 16% in the counties surrounding San Francisco lacked paid sick days, compared with only 1% in San Francisco. Many firms that already offered leave made changes that are meaningful for workers, such as increasing sick leave accrual rates or expanding eligibility to access the benefit. Overall, the net benefit of the sick leave mandate depends on the value of the sick leave to workers, their families, and the general public weighed against the possible reduction in other employee benefits and added costs to consumers and employers.

The strength of this research is a new data source of employer reactions to and opinions concerning a recently mandated sick leave policy. Besides the usual concerns regarding nonrandom response and recall bias, limitations of the study arise from the city-specific nature of the policy and the generalizability of policies enacted in San Francisco, a high-income city. San Francisco may be somewhat unique in having had relatively high (paid or unpaid) sick leave at baseline. We found that 73% of San Francisco firms offered paid or unpaid sick leave in 2006, whereas an analysis of the 2006 National Compensation Survey reported that only 59% of all US firms provided paid sick leave.47 If this is the case, the level of impact of a paid sick leave mandate on employers in the United States overall might be different than that reported here. However, the significant increase in sick leave offerings among employers in the hospitality industry and the strong support for the PSLO they voiced may indicate that sick leave policies could help low-wage workers elsewhere. Additionally, our analysis was derived from employers’ reports; the impact on employees and customers might be better evaluated using information from employees or customers themselves.

We have offered an initial glimpse of some likely impacts of paid sick leave mandates on the configuration of employers’ paid leave programs. Because this is an area of active policy development, research is critically needed on many other aspects of paid sick leave policies. There is no comprehensive analysis of the impact of paid sick leave on health outcomes and health care expenditures for workers and their families, for instance. Minimum paid sick leave standards may have the potential to reduce the spread of communicable disease2–4; this role warrants more attention from public health experts. In addition, the impact on employer outcomes such as implementation costs, changes in employee retention and productivity, and business location decisions are important outcomes for future empirical analysis.

Acknowledgments

Human Participant Protection

References