Jump to ContentJump to Main Navigation
Unfinished BusinessPaid Family Leave in California and the Future of U.S. Work-Family Policy$

Ruth Milkman and Eileen Appelbaum

Print publication date: 2013

Print ISBN-13: 9780801452383

Published to Cornell Scholarship Online: August 2016

DOI: 10.7591/cornell/9780801452383.001.0001

Show Summary Details
Page of

PRINTED FROM Cornell University Press SCHOLARSHIP ONLINE (www.cornell.universitypressscholarship.com). (c) Copyright University of Cornell University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in Cornell for personal use (for details see www.cornell.universitypressscholarship.com/page/privacy-policy/privacy-policy-and-legal-notice). Subscriber: null; date: 21 January 2019

Paid Family Leave and California Business

Paid Family Leave and California Business

Chapter:
(p.55) 4 Paid Family Leave and California Business
Source:
Unfinished Business
Author(s):

Ruth Milkman

Eileen Appelbaum

Publisher:
Cornell University Press
DOI:10.7591/cornell/9780801452383.003.0004

Abstract and Keywords

This chapter turns to the impact of the paid family leave (PFL) program on employers. Companies have easily adapted to the implementation of PFL over the past decade, and their widely voiced concerns about the abuse of and negative effects on productivity and profitability have proven to be almost entirely unwarranted. In fact, many employers have experienced cost savings as a result of PFL. The new program indirectly subsidizes employers who previously provided other forms of wage replacement during family leaves, insofar as their employees draw on the state's PFL benefits instead. Only a few employers have experienced additional costs as a result of the introduction of PFL, but even they have benefited from improved worker morale and reduced turnover.

Keywords:   PFL program, employers, productivity, profitability, wage replacement, worker morale, paid family leave

Business opposition to the legislation that established California’s paid family leave (PFL) program was ubiquitous in the months prior to its passage in the fall of 2002. Despite the fact that, in the final version of the law, workers’ payroll tax contributions fully funded the program, with no direct costs to employers, business groups were unrelenting in their vehement opposition. The president of the California Chamber of Commerce singled the PFL bill out for special censure in the immediate aftermath of its passage, saying, “We’ve opposed a lot of bills, but this is one of the worst” (Broder 2002).

The business lobby repeatedly asserted that the new program was a “job killer” with the potential to wreak havoc on businesses in the state (Anderson 2002; Payne 2002). In a jointly written op-ed published in the San Diego Union-Tribune a year before PFL was scheduled to be fully implemented (but ten months after the law creating it had been passed), the presidents of the California Manufacturers and Technology Association, the California Chamber of Commerce, and the California (p.56) Business Roundtable denounced PFL as a “mega job-killing bill” (Stewart, Zaremberg, and Hauck 2003). The Chamber’s president predicted that if PFL was implemented, businesses would lose control over their employees and that the costs associated with replacing absent workers would be prohibitive (Edds 2002). And a spokesperson for the National Federation of Independent Business declared that PFL would be the biggest financial burden imposed on the state’s small businesses in decades (Geissinger 2002).

Among the criticisms of the program that business groups articulated was the fact that it did not require that employees have any minimum tenure in their jobs before becoming eligible to apply for leave, the lack of a “carve out” for small employers, and the notion that the program would transfer control over leave policies from employers to the state (Taylor 2002; Koss 2003). Another objection was that decisions about eligibility would be made by the Employment Development Department, the state agency that administers the PFL program, and not by employers (Girion and Garvey 2002).

Business also expressed alarm about the potential for fraud and abuse, noting that the law requires workers to certify that they have primary responsibility for caring for an ill family member, but provides no mechanism for either EDD or the employer to verify that this is true (Taylor 2002). “It’s so easy for someone to say, ‘Aunt Mary needs me to go take care of her,’ and the decision whether that person is eligible for paid leave or not is going to be made by the Employment Development Department,” one employer complained (Girion and Garvey 2002).1 The business lobby also suggested that workers would try to take advantage of the PFL program to extend their vacations or to take additional time off whenever they could, and that the program would be abused in a variety of other ways by workers who lacked a legitimate need for family leave (Girion and Garvey 2002).

Moreover, a spokesperson for the state Chamber of Commerce argued that the indirect costs of PFL—such as overtime pay and training for coworkers who would need to cover the duties of those on leave, and hiring temporary replacements for absent workers—would be a crippling burden on California employers and workers (Geissinger 2002). Some business groups also professed concern about the tax burden the new program would impose on workers who, in the final version of the law, (p.57) were fully responsible for financing PFL through payroll deductions. Prior to its passage, the California Chamber of Commerce estimated that PFL would cost $2.5 million a year, nearly four times the estimate advocates had put forward (Geissinger 2002). The president of the Ventura Chamber of Commerce predicted that adding PFL to the existing State Disability Insurance program would triple the payroll tax on employees (Taylor 2002).

These widely expressed fears regarding PFL’s potentially negative effects on business have not materialized, however, as we document in the rest of this chapter. According to employers themselves, in the vast majority of cases PFL has had minimal impact on business operations. Although some organizations do incur additional expenses when covering the work of leave-takers, many more enjoy cost savings thanks to PFL. For the latter, the state benefit has replaced paid leave benefits employers had previously provided, thus generating savings. Moreover, insofar as PFL reduces turnover, it also generates savings for employers by reducing the considerable expenses of recruiting and training new workers. Fears of widespread abuse of the program have also proved unwarranted. And the cost of the program—which in any case is borne entirely by workers—turned out to be well below even the advocates’ initial projections: the payroll tax in 2012 for the SDI and PFL programs combined—fully paid by employees—is only 1 percent of the first $95,585 of earnings.

In documenting the ways in which PFL has impacted California businesses, we draw on two types of original data, one quantitative and one qualitative. The quantitative data come from two telephone surveys we conducted with California employers, first in early 2004, after the PFL law had been passed but before the program was fully operational, and again in early 2010, more than five years after PFL was implemented. Each survey sample included over 250 establishments, and each was stratified by establishment size, so that we could test the prediction that the burden of PFL on small businesses would be particularly onerous.2 This chapter also draws on data from our surveys of California employees before and after PFL was implemented (described in chapter 5 and in the appendix). In addition to the employer and employee surveys, we conducted interviews during site visits to a convenience sample of about twenty workplaces around the state, in a variety of industries and sectors. At each site (p.58) we interviewed human resource managers or operations managers to learn about how they managed leaves on a day-to-day basis. This fieldwork provided a rich body of qualitative data complementing the quantitative data from the employer surveys. The timing was similar to that of the surveys: we did the first round of site visits in early 2004, after the PFL legislation had been passed but before the program was fully implemented; our follow-up visits took place in late 2009 and early 2010. Unlike the surveys, however, the fieldwork was longitudinal: whenever possible we did follow-up visits to the same workplaces, five and a half years after PFL was operational. This enabled us to examine the program’s impact on a select group of employers in considerable detail.3

California Employers’ Family Leave Policies before PFL

In the 2004 employer survey, we asked respondents about the extent to which they provided wage replacement in any form (paid sick leave, paid vacation, and paid parental leave, for example) for the types of family leaves—bonding or caring for a seriously ill family member—that the PFL program was about to make available. We were especially interested in comparing the benefits employers offered to “exempt” managers and professionals—those who are not covered by laws requiring overtime pay—with the benefits employers offered to other, “nonexempt” employees.4 On average, respondents to the 2004 employer survey reported that 60 percent of their nonexempt employees and 72 percent of their exempt employees had received partial or full pay while on leaves of more than one week to care for a new child or a seriously ill family member during the year preceding the survey. In nearly all cases (for 93% of nonexempts and 92% of exempts) this took the form of paid sick leave, paid vacation, or “paid time off” (a benefit combining paid sick leave and paid vacation); very few of these employers had formal paid family leave programs. As figure 4.1 shows, workers in larger establishments (fifty or more workers) were more likely to have such benefits than those employed by smaller ones, although, regardless of establishment size, benefit provision tended to be more generous for exempt than nonexempt employees. Not surprisingly, establishments with unionized workers were more likely to offer such benefits than those with no union presence. (p.59)

Paid Family Leave and California Business

Figure 4.1. Employer-provided benefits, by selected characteristics of workers and establishments, California, 2004. N = 263.

Authors’ 2004 Employer Survey.

We also asked these employers about their experience with the 1993 Family and Medical Leave Act. Not all of them were fully aware of its provisions: among those with fifty or more employees, only 80 percent indicated that their employees were covered by FMLA, while another 14 percent were “not sure.” (In fact under the law all establishments of that size are covered.) Among smaller establishments—many of which, though it is impossible to determine how many, were part of larger, covered firms—29 percent reported that they were covered by FMLA, while 39 percent of this group were “not sure.” Those who indicated that FMLA did apply to their worksites were then asked to evaluate the effect of complying with it on productivity, profitability, turnover, and morale. The vast majority of them, as figure 4.2 shows, reported that FMLA had “no noticeable effect” on their organization. In regard to employee morale, moreover, most respondents who indicated that FMLA did have an effect evaluated it as (p.60)

Paid Family Leave and California Business

Figure 4.2. Effects of FMLA compliance on establishment performance, California, 2004. N = 163.

Authors’ 2004 Employer Survey.

positive, although in the other three areas those who perceived a negative effect slightly outnumbered those who perceived a positive one.5

These results echo those of a national U.S. Department of Labor employer survey conducted in 2000, seven years after FMLA became law, in which large majorities of employers reported that the 1993 law had “no noticeable effect” or a “positive effect” on productivity (84%), profitability(90%) or business growth (90%). In that 2000 survey, similar percentages of employers reported that intermittent FMLA leaves had “no noticeable effect” or a “positive effect” on productivity and profitability (U.S. Department of Labor 2001, tables 6.4, 6.5 and A2–6.13; Waldfogel 2001), although such intermittent leaves remain controversial. In our 2004 employer survey, very few respondents reported any experience of employee abuse of FMLA. When asked, “Have you ever had an employee take a leave under FMLA on a fraudulent basis?” less than 1 percent of the FMLA-covered respondents answered “yes,” and more than half of that small group indicated that they had seen only one or two instances of such abuse in the previous year.6

We also asked managers about their experiences with FMLA in the first wave of site visits that we conducted in early 2004, prior to PFL’s implementation. Table 4.1 summarizes their responses. Confirming the (p.61) survey results, most interviewees reported that the effects of the federal law on their operations were positive or neutral. Several of them told us that FMLA pregnancy and bonding leaves, as well as medical leaves (as for nonemergency surgery) that could be scheduled in advance, were the easiest to manage, since advance planning was feasible—unlike leaves involving unanticipated medical crises, which took all concerned by surprise. But even in the latter cases, as a manager (case D) noted, “The problems are at the beginning; you have to scramble to figure out how to get the work done when someone first goes. After that, though, it’s not a problem.”

Several managers suggested that FMLA had not only helped them to retain valued employees but also had a positive impact on organizational morale in their workplaces. “The people who get the leaves appreciate it,” one commented (case G, the corporate headquarters of a food company). “In the long term, we get better productivity, because employees feel they are supported by the company,” another declared (case D, a computer engineering firm). “Overall, it helps with morale.” Similarly, the human resource manager of an apparel retail chain (case K) noted, “Yes, people have the burden of picking up the slack for someone else on leave, but they also know that someone will do this for them if they need it.” Another human resource manager told us, “It gives employees peace of mind, and plays into our [firm’s] culture of work-life balance” (case O, a law firm). “Turnover would be much higher in the absence of FMLA leaves,” an entertainment company human resource manager suggested (case F). “And other employees, coworkers, are sympathetic.”

The one issue that several managers did complain about involved the intermittent leaves available under FMLA, which in some cases created time-consuming and annoying problems for managers. Especially difficult were situations where an employee could obtain blanket medical documentation of a chronic health condition, such as migraine headaches or asthma. This led to chronic and unpredictable absenteeism. “The migraines occur regularly on Monday or Friday or when it will be particularly busy,” one frustrated manager explained (case S, a public utility). Another informant (a hospital manager at case C) told us, “The employee can just call up in the morning and not come in. It’s unchecked, uncontrollable absenteeism. It’s like they have a ‘Get Out of Jail Free’ card!” Such problems, although by all accounts limited to a handful of employees, not only were a nuisance (p.62)

Table 4.1. Case studies, employer experience with FMLA leaves, California, 2004

Case code

Type of business

Total employees

Covered by FMLA?

Employees aware of FMLA

Employees return after FMLA leaves?

Effect of FMLA on organization

A

Manufacturing

1,500

Yes

Most are

Nearly all return

“No effect at all”

B

Software design

850

Yes

Yes

Most return

Positive effect on morale

C

Hospital

8,500

Yes

Yes

Nearly all return

Some abuse, especially with intermittent leaves; this reduces morale

D

Computer engineering

16,000

Yes

Yes

About 75% return

Improves productivity and morale

E

Optometrist

3

No

N/A

N/A

N/A

F

Entertainment

2,000

Yes

Yes

“Most people do”

Has reduced turnover

G

Food–headquarters

250

Yes

Yes

7 of 8 did last year

Mostly good for morale, but some abuse by a handful of employees

H

Food–agricultural work

1,400

Yes

Yes

Most return

No effect on morale, profitability, productivity

I

Food processing plant

120

Yes

Yes

Most return

Reduces turnover, no effect on morale, profitability or productivity

J

Insurance

2,652

Yes

Yes

85% return

No effect on turnover, profitability, or productivity

K

Apparel retail chain

19,700

Yes

Yes

Most return

Positive impact on retention, negative for productivity and profitability

L

Computer chip design

12,500

Yes

Yes

Most return

No effect–a “nonevent”

M

Restaurant chain

550

Yes

Yes

90% or more return

“No big deal”

N

Biotechnology

97

Yes

Yes

Nearly all return

Positive effect on morale

O

Law firm

1,425

Yes

Yes

Over 90% return

Good for morale and recruitment

P

Construction/engineering

850

Yes

Yes

Nearly all return

“We have it down pat.” No effects

Q

Payroll processing

629

Yes

Yes

90% return

Overall, no effect

R

Nonprofit Research Firm

943

Yes

Yes

Everyone has returned

Positive for organization

S

Public utility

14,000

Yes

Yes

Nearly all return

Intermittent leaves that abuse law are bad for morale and productivity

(p.63) for managers but could also spark resentment and thus morale problems among the coworkers of the employees involved.

Although intermittent leave issues were a serious concern for some employers, the problem was not widespread among the establishments we visited, and in some cases it was associated with broader problems of dissatisfaction and morale. At the public utility (case S), for example, one (p.64) manager noted that intermittent leaves had become common in a department that also suffered from poor work organization and a work scheduling system that was the object of frequent employee complaints. After a new department supervisor was hired who put scheduling on a more rational and predictable basis, and who applied company rules in a fair and consistent manner, morale improved and absenteeism fell—as did the use of intermittent leave.

Most of the managers we interviewed felt that the positive benefits of FMLA outweighed its negative features, and some were positively enthusiastic. “You always have perfunctory whining from managers,” one human resource manager told us (case R, a nonprofit research firm). “But everyone has problems from time to time and we understand that employees need time off occasionally.” Another manager, who described himself as an advocate of work-life balance, defended the decade-old FMLA law against what he saw as ideological criticism from others in the business community. In his view, FMLA was a “nonevent” as far as productivity and profitability were concerned. “When we’re talking about bottom-line issues, I’ve never heard anyone say, ‘The real problem is FMLA.’ No one has ever said, ‘The share price of our stock stinks, and if we could only repeal this leave law we’d be doing better’” (case L).

PFL’s Impact on Leave-taking Rates

One might expect that once PFL became available, California workers would take family leaves more often, and for longer periods of time, than before this form of wage replacement benefit was available. Indeed, we saw some evidence of this in chapter 3, which documented the increasing take-up of PFL among men for bonding leaves. On the other hand, the deep recession that began in 2007 has sharply increased employment insecurity, and may have made workers more hesitant than before to go on any type of extended leave, particularly among those whose right to return to the same or a comparable job is not legally protected. The data on rates of leave-taking and leave lengths (in tables 4.2 and 4.3, respectively) should be interpreted with these factors in mind.

(p.65) Table 4.2 shows the rates of leave-taking that employers reported in our 2004 and 2010 surveys, and the variation in those rates by establishment size. In both years, more than half of the employers reported that no employees had taken parental leave in the previous year. In both the 2004 and 2010 surveys, more than 85 percent of small establishments (less than fifty employees) reported no cases of parental leave in the previous year; among those small establishments that did report at least one case of parental leave, the median rate of parental leave-taking was about 7 percent in both survey years, which translates into one employee per year in a typical establishment with twenty employees and three employees per year for one with forty to fifty employees. Of course, as the number of employees per establishment increases, the likelihood that at least one will go on a parental leave rises. Indeed, as table 4.2 shows, among establishments with 50 to 499 employees, 36 percent in 2004 and 38 percent in 2010 had at least one employee who took parental leave. As a share of the workforce at these worksites, however, median rates of parental leave-taking were miniscule: 1.4 percent in 2004 and 2.1 percent in 2010, which translates into about one employee a year for establishments with fifty employees and about ten a year for those with 499 employees. Among the largest establishments (five hundred or more employees), the likelihood that at least one employee will need to take parental leave during the year was still higher: 38 percent in 2004 and 82 percent in 2010. Here too, median rates of parental leave taking were modest, about 2 percent in both 2004 and 2010. What appears to be a higher rate of leave-taking in small establishments is largely an artifact of arithmetic: a worksite with only ten employees in which one takes parental leave will have a leave-taking rate of 10 percent.

As table 4.2 shows, between 2004 and 2010 there was a statistically significant increase—by a factor of two—in the number of parental leaves in the largest establishment category (establishments with five hundred or more employees). However, in all the other establishment size categories the change over time was not statistically significant, despite the establishment of PFL in the interim. The limited extent of change over this period may reflect the fact that birth rates decline in periods of economic downturn, or it may be an indication of workers’ greater reluctance to take leave during difficult economic times, or both. Even in normal economic (p.66)

Table 4.2. Parental and pregnancy leave-taking rates, by establishment size, California, 2004 and 2010

All employer respondents

Less than 50 employees

50–499 employees

500+ employees

2004

2010

2004

2010

2004

2010

2004

2010

Total N of all establishments

263

253

134

102

65

84

64

67

Number of establishments with any parental leaves in past year

101

92

13

8

37

32

51

52

Percentage of establishments with any parental leaves in past year (weighted)

15.4

8.8

13.9

7.0

36.0

38.1

38.1

82.1*

Median parental leave-taking rate (percentage of workers in establishments with any parental leaves in past year)

1.8

2.5

6.7

7.4

1.4

2.1

1.7

2.4

Number of establishments with any pregnancy leaves in past ear

111

16

40

55

Percentage of establishments with any pregnancy leaves in past year (weighted)

13.4

11.3

47.6

86.6

Median pregnancy leave-taking rate (percentage of workers in establishments with any pregnancy leaves in past year)

2.5

6.5

2.5

2.0

Source: Authors’ 2004 and 2010 Employer Surveys.

(*) p <.001.

times, however, only a small number of workers will become parents at any one point in time (although most will do so at some point in their working lives).

The 2004 survey did not include questions about pregnancy leaves (for which wage replacement was already available through SDI at that time). But since PFL cannot be used for wage replacement for pregnancy leaves, there is no reason to expect significant change in such leaves over time, except as a result of the economic downturn. In any case, the rates for pregnancy leaves in 2010 were generally similar to those for bonding leaves. (p.67)

Table 4.3. Median length of family leave (in weeks) by leave type, gender, and exempt/nonexempt status, California, 2004 and 2009–10

Baby bonding

Ill family member

2004

2009–10

2004

2009–10

N (all leave takers)

65

98

33

53

Exempt

   Male

3

3

3

4

   Female

11

12

4

4.5

Nonexempt

   Male

3

3

3

3

   Female

12

12

3

5.5*

Source: Authors’ 2004 and 2009–10 Employee Surveys.

(*) p <.05.

In addition to its effect on the frequency of leave-taking by employees, one might expect the availability of PFL to increase the length of leaves, since workers can afford to take more time off with partial wage replacement than in its absence. However, as table 4.3 shows, the length of baby bonding leaves that respondents to our employee surveys reported did not change significantly between 2004 and 2009–10. For men, the median leave length, whether they held exempt or nonexempt jobs, was three weeks in both periods. The median length of baby bonding leaves taken by new mothers in nonexempt jobs also was the same in both periods, at twelve weeks. Only mothers in exempt jobs reported taking longer leaves in the later year, and this change was small and statistically insignificant (the median was twelve weeks in 2009–10 and eleven weeks in 2004). The median length of leave to care for an ill family member did increase for women in both exempt and nonexempt jobs, and for men in exempt jobs, but these changes were not statistically significant. The only statistically significant increase occurred among nonexempt women taking leaves to care for a seriously ill family member.

PFL’s Effects on Organizational Performance

The 2010 employer survey explored the impact of PFL on various aspects of organizational performance. To an even greater degree than the results (p.68) reported for FMLA in the 2004 survey, the data show that PFL had a minimal impact on the operations of California businesses. As figure 4.3 shows, nine out of ten respondents reported either positive effects or no effect of PFL on their establishments. When asked, “What effect has it [PFL] had on this location’s business productivity?” 89 percent of employer respondents reported either a “positive effect” or “no noticeable effect.” Similarly, when asked, “What effect has it had on this location’s business profitability/performance?” an even higher percentage of employer respondents, 91 percent, reported that PFL had either a “positive effect” or “no noticeable effect.”7 As for employee turnover, 93 percent of the employers surveyed in 2010 reported that it had a “positive effect” or “no noticeable effect.” And nearly all respondents said that the effect on morale was either positive or not noticeable.

Contrary to the claims of business groups that the PFL program would place an especially severe burden on small businesses, as table 4.4 shows, among the small minority of employers that did report negative effects, the larger establishments—those with more than five hundred employees—were overrepresented. This may reflect the fact that in larger establishments the organizational representatives who responded to our telephone survey tended to be more removed from the day-to-day management of family

Paid Family Leave and California Business

Figure 4.3. Effects of PFL compliance on establishment performance, California, 2010. N = 176. Totals may not add up to 100% due to rounding.

Authors’ 2010 Employer Survey.

(p.69) leaves than those who responded on behalf of smaller establishments. It follows that some of the data shown in table 4.4 for the larger establishments may reflect ideologically rooted views of PFL rather than on-the-ground experience with leaves.

Our 2010 survey also explored another concern business groups had raised when the PFL legislation was being debated, namely that the program would be subject to abuse, with workers filing illegitimate claims. However, asked if they were “aware of any instances in which employees that you are responsible for abused the state Paid Family Leave program,” 91 percent of all employer respondents surveyed replied “No.” And among the other 9 percent who were aware of abuse, it was a relatively rare occurrence: 27 percent of them knew of only one instance of abuse, and 99 percent knew of no more than five such instances. Once again belying the concern that small businesses would be disproportionately affected by PFL, respondents from the largest establishments were more likely to report abuse than their smaller counterparts: 25 percent of establishments with five hundred or more workers reported abuse, compared to only 7 percent of establishments with fifty to 499 workers, and 9 percent in establishments with less than fifty workers (N = 177).

These survey findings about the impact of PFL on performance were confirmed in our fieldwork, when we returned in 2010 to the companies we had visited six years earlier to explore their experiences with the new state program.8 Of the nineteen employers we visited in 2004, fifteen agreed to be reinterviewed in 2010. We replaced the other four (cases A, D, N, and Q in table 4.1) with similar companies in the same industry (these are denoted A’, D’, N’, and Q’ in tables 4.5, 4.8, and 4.9). As table 4.5 shows,

Table 4.4. PFL’s effects on organizational performance, by establishment size, California, 2010 (%)

“Positive effect” or “no noticeable effect” on:

Less than 50 employees

50–499 employees

500+ employees

All employer respondents

Productivity

88.8

86.6

71.2

88.5

Profitability/performance

91.1

91.2

77.6

91.0

Turnover

92.2

98.6

96.6

92.8

Morale

98.9

95.6

91.5

98.6

Source: Authors’ 2010 Employer Survey.

N = 175

(p.70)

Table 4.5. Case studies, employer experiences with PFL leaves, California, 2010

Case code

Type of business

Total employees

PFL’s effects on organization

A’

Manufacturing

4,894

Small positive effect on profitability and morale. Employees really enjoy being able to spend more time with their new children. PFL has saved the company money since they now top off PFL.

B

Software design

2,781

N/A

C

Hospital

11,284

Minimal impact on productivity. Noticeable improvement in worker morale.

D’

Computer engineering

20,000

No effect on operations or productivity. Mothers love the program. Costs of PFL covered by payroll deductions, so there is no cost to the company.

E

Optometrist

4

No impact to date on the business, but concerned about potential future extended absences in this small business.

F

Entertainment

1,300

No negative impact. Boosts morale and company’s “family oriented” image. Employees expressing gratitude for help with arranging leaves and “for caring.”

G

Food–headquarters

1,334

No effect on productivity, profitability, turnover, or absenteeism. “It’s a blip on the radar screen.”

H/I

Food–processing and agricultural work

1,000

No effect on operations. Workers less “stressed” over leave now that it is paid.

J

Insurance

2,000

No effect on the business.

K

Apparel retail chain

561

Not an issue for the company. PFL helps them retain employees.

L

Computer chip design

10,552

The cost of replicating PFL program as part of self-insurance SDI has been less than expected due to low take-up. Boosts morale.

M

Restaurant chain

630

No adverse impact. Coworkers like the opportunity to pick up extra shifts. Mothers like the additional income.

N’

Life sciences/biotechnology

108

Long leaves can impact productivity–not PFL per se but long medical leaves. “It’s a cost of doing business, but it’s not a problem.”

O

Law firm

270

“Stress” leaves incur resentment from coworkers. But for caregivers, it’s better to take time off rather than working while distracted.

P

Construction/engineering

810

No major impact on business or on turnover. Enables company to give new career hires rotational experience. “We use longer leaves to our advantage.” Positively affects morale–coworkers help one another out, and PFL provides time to arrange childcare, reducing absenteeism.

Q’

Payroll processing

500

No effect, a “nonevent” for the company.

R

Nonprofit research firm

819

Overall it’s positive, especially for retention. But parents more likely to take time off due to PFL income, and leaves are longer, especially for fathers. Gives employees a chance to rotate and get new skills and experience. Boosts morale.

S

Public utility

19,000

N/A

T’

Public-sector hospital

8,756

N/A

(p.71) most of the managers we interviewed reported that PFL had had positive effects on their employees, especially in regard to morale (see cases C, F, P, and R). Several noted that employees themselves had very positive views of the program (cases A’, D’, M), and that employees’ stress levels were reduced thanks to the availability of paid leave (cases H and M). Nearly all of these employers told us that the PFL program had had no negative effects on their business operations (cases D’, E, F, G, H, J, K, M, P, Q’) or a positive effect (cases A’, P, and R). Only case C reported a minimal negative effect on productivity from family leaves. (At case N, managers also stated that long leaves impact productivity, but this involved medical leaves rather than family leaves.)

(p.72) Covering the Work of Employees on Family Leave

The majority of employers we surveyed in 2010 reported that they typically covered the work of employees on leave by assigning the work temporarily to other employees. As table 4.6 shows, this was the primary method of covering the work of exempt workers on leave, cited by 77 percent of respondents. For an even larger proportion—90 percent—of these employers, this was the most common method of covering the work of nonexempt employees on leave. This practice was especially common in larger establishments, while in smaller ones the work of exempt employees was often put on hold, or employees were asked to do some work while they were out on leave. Many employers reported that they hired temporary replacements to cover the work of nonexempts during their leaves; this method was used a bit less frequently to cover the works of exempts.9

The widespread practice of assigning the work of an employee on leave to co-workers raises concerns that this might lead to resentment among

Table 4.6. Method of covering the work of family leave-takers, California, 2010 (%)

Method(s) used to cover the work of leave-takers

Exempt workers

Nonexempt workers

All establishments

Number of employees in establishment

All establishments

Number of employees in establishment

< 50

50–499

500+

<50

50–499

500+

Assign work to others

76.9*

72.3

100.0

98.0

89.8

88.1

96.8

100.0

Hire temps

39.3

39.9

36.8

27.5

48.7

52.4

32.3

46.2

Hire replacements

0.4

0

0

3.9

2.4

0

12.9

1.9

Put work on hold

59.1**

67.5**

15.8

31.4

8.0

6.0

16.1

20.8

Workers do some work while on leave

45.5**

52**

13.2

5.9

6.6

6.0

9.7

1.9

Some other method

15.7

16.2

13.2

11.8

9.7

6.0

25.8

18.9

Source: Authors’ 2010 Employer Survey.

Note: Totals may add up to more than 100% because employers could report more than one method.

N = 138

(*) p <.05;

(**) p <.001, comparing exempt and nonexempt workers.

(p.73)

Paid Family Leave and California Business

Figure 4.4. Impact of coworkers taking leave, California, 2004 and 2009–2010. N = 472 in 2004; N = 454 in 2009–10. Totals do not add up to 100% due to rounding.

Authors’ 2004 and 2009–10 Employee Surveys.

the latter group. We explored this question in the employee surveys in both 2004 and 2009–10; the results are summarized in figure 4.4 and table 4.7. In both years, about 60 percent of respondents had a coworker who went on leave. About a third of workers in each period reported working more hours as a result, and over half reported that they took on more duties. Although the great majority of workers reported that this had no impact or a positive impact in both years (90% in 2004 and 94% in 2009–10), the proportion reporting a positive impact increased from 16 percent in 2004 to 25 percent in 2009–10. Thus the concern that coworkers might resent having to fill in when a colleague goes on family leave turns out to be highly exaggerated. On the contrary, the view of employers that PFL improves worker morale is confirmed by these data on workers’ views of leave-taking by coworkers.

Confirming the survey results, most of the employers we interviewed on our 2010 site visits told us that to cover the work of employees on leaves, they most often assigned the tasks involved to other employees, with the second most common method being hiring of temporary replacements to cover the work. These findings are summarized in table 4.8. In most of these companies, the work of employees on leave was covered by coworkers (cases A, B, D’, F, H/I, M, N’, O, P, and R). Some employers reported (p.74)

Table 4.7. Rates and impacts of coworkers taking leave, California, 2004 and 2009–10

2004

2009–10

Percentage of respondents with a coworker taking leave (%)

56.7

60.9

Effects of coworker leave on respondent’s job (%)

   More hours

32.5

33.0

   Extra shiftsa

22.7

N/A

   More duties

52.6

51.2

Overall impact of coworker taking leave on respondent (%)

   Positive

16.1

24.7*

   Negative

10.1

6.4

   Neither positive nor negative

73.7

68.9

N

264

278

Source: Authors’ 2004 and 2009–10 Employee Surveys.

(a) 2009–10 survey did not include a question about extra shifts.

(*) p <.05, comparing 2004 to 2009–10.

the use of cross-training, opportunities for employees to learn new skills, and successful backup systems as among the key reasons they did not have to hire temps (cases B, H, K, and P). In a few cases, employers did regularly use temps (cases A, C, R); N’ made occasional use of temps as well. And case H used overtime during peak demand periods to cover the work of those on leave. But assigning coworkers to cover the work was by far the most common approach.

At the same time, this apparent homogeneity of approaches obscures the rich variety of arrangements that were used for covering the work of employees during both brief and more extended absences. In our field work, we found that every single employer we visited had developed systematic, and often ingenious, methods for covering the work of employees when they were absent for any extended period, whether it was for a family leave or for a vacation, or any other reason. As one might expect, there is considerable variation depending in part on the nature and urgency of the tasks that need to be covered.

In some settings, like a factory assembly line or a hospital, full coverage for all positions is needed 100 percent of the time. A hospital we visited (p.75)

Table 4.8. Case studies, covering the work of employees on PFL leaves, California, 2010

Case code

Type of business

Total employees

Covering the work of employees on PFL leaves

A’

Manufacturing

4,894

Blue-collar work is covered by temps and white collar work by coworkers.

B

Software design

2,781

Coworkers cover the work. Many are cross-trained.

C

Hospital

11,284

Request a “traveler” for RNs, techs; for nurses, may also use trainees.

D’

Computer engineering

20,000

Coworkers cover; in the vast majority of cases, jobs are divided among work team.

E

Optometrist

4

Would be difficult to cover work in this small business.

F

Entertainment

1,300

Work is reassigned to other team members in a department.

G

Food–headquarters

1,334

N/A

H/I

Food–processing and agricultural work

1,000

In the fields, crews just run with fewer people. We use overtime in the peak season.

J

Insurance

2,000

Very few employees have taken leave.

K

Apparel retail chain

561

New hires cover sales work. High turnover means a job is available when employee returns from leave.

L

Computer chip design

10,552

Covered by coworkers.

M

Restaurant chain

630

They offer extra shifts to the rest of the workforce (but avoid paying overtime).

N’

Life sciences/biotechnology

108

Mostly covered by coworkers, occasionally by temps.

O

Law firm

270

Developed a successful “back-up” system to avoid using temps.

P

Construction/engineering

810

Work is reassigned to available coworkers.

Q’

Payroll processing

500

Many leave-takers are “virtual employees”–is good for moms.

R

Nonprofit research firm

819

Temps cover nonexempts’ work; coworkers cover work of exempts.

S

Public utility

19,000

N/A

T’

Public-sector hospital

8,756

N/A (not covered by PFL)

(p.76) (case T’) provides one example of a situation where coverage is imperative, and work is highly skilled. The hospital maintains a “voluntary extra shift list” of nurses and nursing assistants, who indicate the days they are available to work overtime. The hospital offers incentives for this by paying well above the legally mandated rate for overtime. Another approach was used at the public utility (case S), where the work is also skilled and often highly time sensitive. Here, managers rely on voluntary overtime among coworkers to cover the work of employees who are absent. In one manufacturing firm we visited, where one might expect hiring replacement workers or temps to be the main approach, the machine operators worked in teams, enabling coworkers to be assigned to cover the work of absent team members.

At the other end of the spectrum, as in office settings where deadlines are typically more flexible, even absences of several months can be covered by coworkers, with less urgent tasks put on hold and others taken over by the staff that remain. In one worksite we visited, the corporate headquarters of a food manufacturing company, it is standard practice to put some work on hold until the absent employee returns, while coworkers cover the most urgent tasks. At a nonprofit research firm, similarly, coworkers cover the duties of absent support staff; in some cases people are moved from one office to another to balance workloads.

In retail chain stores and in agricultural work, where turnover is very high and hiring is therefore virtually continuous, leaves can be handled even more easily, since an employee’s temporary absence can be covered by one of the many new hires that are constantly arriving; and when a leave-taker returns to work there is sure to be a position available. At the other extreme, for jobs that require extensive training so that very few people are able to take over for someone who is absent, managing leaves presents a greater challenge and managers must be more inventive. At a law firm we visited, where continual coverage of support staff positions is essential, management keeps several floaters on the regular staff, drawing on temps as a last resort. “Every day is a juggle,” one manager said.

We also visited a small retail department within an entertainment company where merchandise is highly specialized. Here there were only four full-time workers, but two other individuals with extensive training and experience in this setting were available on an “on call” basis. These two workers were not interested in working full time, but made themselves (p.77) available when the regular staff members were absent and occasionally have replaced a full-time worker on an extended leave. Similarly, a small branch of a restaurant chain that employs twenty-four nonexempt workers has four students on the payroll who work part-time and fill in when full-time staff members are unavailable.

At a biotechnology company we visited, extensive cross-training ensures that many professional and managerial employees can cover for one another during absences as needed. When this is not practical, the company employs contractors and consultants who periodically work for the firm and are generally familiar with its operations. A large engineering/construction firm with a great deal of project-based work on tight deadlines adopted a different approach. Because professionals already routinely work very long hours, and because the price of engineering mistakes can be enormous, this firm does not ask coworkers to do additional work to cover for an absent engineer. Instead, management identifies another engineer within the firm who is working on a project that is winding down and assigns him or her to cover for the absentee. We observed a similar approach at a large high-tech firm that maintains a “redeployment” pool that includes both exempts (such as software engineers) and nonexempts (including factory workers as well as administrative support staff) whose positions had been eliminated. Managers regularly draw on this “redeployment” pool to find replacements for employees on leave.

The smallest business we visited, an optometrist’s office, was the least well equipped to cover leaves. This business only has three employees (apart from the owner), one of whom is a highly skilled technician. When this individual is absent, the optometrist fills in himself and takes fewer clients. Very small businesses like this one do face special challenges since an inevitable effect of their size is that very few coworkers are available to cover the work when someone is absent.

In virtually all the establishments we visited, managers had crafted solutions of one sort or another to the problem of covering the work of absent employees. Most were able to do so with little difficulty, although sometimes the costs in premium overtime pay or fees to temp agencies were significant. However, having contingency plans in place for such events is a business necessity, entirely apart from the question of family leave. Managers constantly face the possibility that an employee may quit precipitously, (p.78) become seriously ill and unable to work, go on a military leave, take an extended vacation or unpaid leave, and so forth. Under all these circumstances, many of which occur on a regular basis, the work of absent employees must be covered. As a result, virtually all employers have established mechanisms for ensuring that work will be performed during employee absences—mechanisms they can deploy when employees take leave to care for a new child or seriously ill family member, just as for other absences.

Costs and Benefits to Employers of PFL

Prior to 2004, business groups raised the specter of huge hidden costs for employers if California implemented its PFL program—costs that, they claimed, could cripple California businesses and force them to lay off workers or move out of state. Our data show that this fear was unwarranted. Asked if the introduction of the PFL program had resulted in “any cost increases,” 87 percent of respondents to our 2010 employer survey indicated that it had not. Some employers (9% of those responding to this question) indicated that the PFL program had generated cost savings for their organizations, by reducing employee turnover, reducing their own benefit costs, or both, when employees used the program instead of (or in combination with) employer-provided paid vacation, sick leave, or disability benefits.

Those employers who cover the work of employees on leave by assigning it to coworkers—which as we just saw is by far the most common approach—incur minimal extra costs as a result of any additional leaves that have resulted from the passage of PFL. Those who replace leave-takers with temps or by assigning overtime do incur some extra costs, even subtracting the savings from not paying the wages that employees on leave would have earned had they remained on the job. Indeed, among the 13 percent of respondents to the 2010 employer survey who indicated that PFL had generated cost increases, most reported additional hiring and training expenses to cover the work of employees who were out on leave. But for most employers these costs tend to be modest.

On the other hand, many of the employers who had provided their own paid leave benefits to employees prior to the creation of the state PFL program are now enjoying cost savings. Among employers responding to the (p.79) 2010 survey, 60 percent reported that they coordinated their own benefits for exempt employees with PFL, and nearly as many (58%) did so for nonexempt employees. Similarly, about a third of the twenty employers we visited in our 2010 fieldwork coordinate company benefits with the state PFL program (A’, B, H/I, N’, and O) and realized a cost saving in that way, as table 4.9 shows. Case B requires employees to use the state program while the others strongly encourage its use. About 40 percent of these employers reported that they allow employees to use accumulated paid vacation or other paid time off at their discretion to top off the state benefit or to extend their leaves (cases C, D’, J, K, M, P, and Q’). A few employers told us that their employees are more likely to use company-provided benefits than PFL (cases F, L, and P). Surprisingly, however, one company

Table 4.9. Case studies, coordination of employer-provided benefits and PFL, California, 2010

Case code

Type of business

Total employees

Coordination of employer-provided benefits with PFL

A’

Manufacturing

4,894

HR encourages everyone eligible to use the program as that is a cost saving. Older workers often prefer to use accumulated sick leave and vacation pay to avoid PFL paperwork. But parents, who are younger and have had time to plan leaves during the pregnancy, typically use PFL, topping it off with sick leave and vacation pay.

B

Software design

2,781

The weeklong waiting period is fully paid by the company, and then the company coordinates with PFL benefits. Birth mothers on PFL can use vacation pay and/or up to 6 weeks of sabbatical pay if eligible. For fathers, company provides a one-week parental leave at 100% pay, and an additional 3 weeks leave coordinated with PFL.

C

Hospital

11,284

PFL is topped off from any available extended sick leave or paid-time-off benefits employee has accumulated.

D’

Computer engineering

20,000

Does not actively coordinate benefits with PFL, but company benefits can be used to extend leaves. Exempts have unlimited paid sick days; nonexempts must document time they are out of the office.

E

Optometrist

4

No coordination of benefits.

F

Entertainment

1,300

No coordination of benefits. Most people (especially men) prefer to draw on accumulated vacation pay rather than PFL.

G

Food–headquarters

1,334

Employees who take PFL are placed on “no pay status” and may not coordinate PFL with accrued paid leave; but company does coordinate its benefits with SDI, which yields cost savings.

H/I

Food–processing and agricultural work

1,000

PFL is coordinated with company benefits for those who are eligible for the latter (only one segment of the workforce).

J

Insurance

2,000

Employees may draw on accumulated paid time off at their discretion to top off PFL. No direct savings for employer.

K

Apparel retail chain

561

Full-time employees, both exempt and nonexempt, can top off PFL with accumulated paid time off. No direct savings for employer.

L

Computer chip design

10,552

Company’s own family leave program is easier to access than state program, take-up is low, no coordination of benefits.

M

Restaurant chain

630

Hourly employees may use any accrued paid vacation to top off PFL. No savings for employer.

N’

Life sciences/biotechnology

108

PFL is coordinated with company benefits.

O

Law firm

270

For nonexempt employees company tops off PFL benefits so workers get up to 100% of pay (% depends on years of service). For lawyers, primary caregivers get 100% pay for the 6 weeks of PFL; secondary caregivers get 2 weeks.

P

Construction/engineering

810

No coordination. Senior employees who have accumulated a lot of paid time off typically use that and don’t bother with PFL.

Q’

Payroll processing

500

For minor (i.e., brief) leaves HR has some discretion and sometimes just lets exempts use sick leave for the whole period to avoid the paperwork hassles. For longer family leaves, California PFL can be concurrent with a vacation, and vacation leave can be used to top off.

R

Nonprofit re-search firm

819

There is no company-provided PFL program, but a payroll service provider administers the program and helps employees fill out forms. Some “whining” from non-California employees who don’t have PFL. Providing the same company benefits across states is a problem even in regard to who to include. Company allows employees to take time off to care for a domestic partner, which was a contentious issue in their more conservative Arlington, Virginia, office.

S

Public utility

19,000

N/A

T’

Public-sector hospital

8,756

N/A

(p.80) (p.81) (case G)—a corporate headquarters—explicitly prohibits employees from drawing on paid vacation or other paid time off while using PFL.

Turnover and Retention

Turnover can be quite costly to employers. Replacing a worker who quits typically requires advertising the position and may also involve agency fees and candidate travel for interviews. A large company with high turnover may need to hire a staff of professional recruiters. As candidates apply for the position, employees must spend time reviewing resumes, carrying out preliminary telephone interviews, interviewing top candidates, calling references and doing background checks, and carrying out drug tests or physicals if required for the job. Once hired, a new employee’s information must be entered into the payroll system and the employee enrolled in health insurance and other benefits. Often there is an orientation for new employees on “company time.” And it takes time—from a few days to a few months, depending on the job—for a new employee to become fully productive. These are among the costs an organization may experience when it fills a vacancy. Insofar as PFL reduces turnover, then, employers may reap considerable savings.

(p.82) In our 2004 employer survey, we collected data on the costs organizations incurred when filling a vacant position. To calculate the costs for filling a vacant nonexempt position (typically an hourly paid position) we collected information on (1) time spent by exempt employees screening applicants and doing paperwork; (2) agency fees, advertising costs, and related expenses; (3) training time for new nonexempt employees; and (4) average time for the nonexempt employee to achieve full proficiency. We made the simplifying assumption that proficiency increases in equal increments, so that a worker achieves 10 percent proficiency in 10 percent of the time it takes to achieve 100 percent proficiency. In this case, the employee’s proficiency is 50 percent, on average, over the time period to achieve full proficiency. The employer, however, is paying full wages over this period. Thus, half the earnings of a new employee over the period when they are learning the job are a cost to the employer of filling the position. We multiplied time spent by exempt and nonexempt employees in these activities by the average wage of exempts and nonexempts respectively in the organization. We added these costs to find the total cost of replacing a nonexempt employee. We did the same calculation for exempt employees. Those costs varied with establishment size, as table 4.10 shows. Turnover costs for nonexempt workers varied from $4,149 (in 2009 dollars) to almost $8,043 in the largest organizations, while for exempt workers the range was from $12,625 to $18,331.10

Table 4.10. Turnover costs for employers by firm size (in 2009 dollars)

Nonexempt

Exempt

N

Average cost ($)

Average wage ($)

N

Average cost ($)

Average wage ($)

Firm size < 50

89

5,394

15.24

51

12,625

27.47

Firm size 50–499

58

4,149

12.03

49

13,912

21.70

Firm size 500+

51

8,043

17.95

48

18,331

29.92

   Total

198

5,328

15.02

148

12,887

27.02

Source: Computed from authors’ 2004 employer survey.

(p.83) Calculating the annual earnings of employees in each category and comparing this with the cost of turnover, we found that it cost employers between 17 percent and 21 percent of the employee’s annual salary to replace a nonexempt worker who leaves her or his job, and between 22 percent and 31 percent of the employee’s annual salary to replace an exempt employee. On average, it costs 17 percent of annual salary to replace a nonexempt employee and 23 percent to replace an exempt employee. As these figures make clear, turnover is quite costly to employers, and retention can provide organizations with significant savings.

The overwhelming majority of employers (93%), as figure 4.3 shows, reported that California’s PFL program had no effect or a positive effect in reducing turnover. Data from our 2009–10 employee survey shed additional light on this issue (see discussion in chapter 5). Among respondents to that survey who used PFL when they took a family leave, 83 percent returned at the end of their leaves to the same employer they had worked for previously compared with 74 percent of those who did not use PFL, although this difference was not statistically significant. California’s PFL program may contribute to reduced turnover and thus provide another form of savings to employers. This is especially valuable for smaller employers that may be unable to afford high levels of wage replacement for workers who need to take a family leave but wish to retain those workers.

Conclusion

Despite widespread fears expressed by opponents of the program that PFL would create a heavy burden on the state’s employers, our data suggest that they have had little or no difficulty adjusting to it. Five and a half years after PFL began operations, the vast majority of employers reported no effects or positive effects on their productivity or profitability (or, in the case of non-profit organizations, on their performance). Predictions that small businesses would find PFL especially burdensome were not borne out; on the contrary, among the few employers that did report negative effects, large businesses predominated. Cases of PFL abuse have also been rare, contrary to some predictions.

(p.84) For many employers, PFL has been a source of cost savings, either due to reduced turnover or because they coordinated their own benefits with the state PFL program. Most employers reported that they covered the work of those out on leave by reassigning it to other employees, at little or no cost. A few employers did report higher costs due to the need to hire temporary replacements for employees on family leave, or for overtime pay. To the extent that the PFL program reduced turnover, however, the cost savings from improved retention helps to balance these additional costs.

Notes:

(1.) In fact a leave to care for “Aunt Mary,” even if she were seriously ill, would not be eligible for PFL, since aunts are not among the family members the program’s caregiving leaves cover (only seriously ill parents, children, spouses, and registered domestic partners of a covered worker are included in the program).

(2.) All results reported in this chapter are weighted (using the establishment weights described in the appendix) to adjust for the overrepresentation of large firms in the sample, and to adjust for nonresponse. The 2004 and 2010 samples were independent of one another; this was not a longitudinal study. For details on the survey methodology, see the appendix.

(3.) These interviews ranged from thirty minutes to two hours in length each, and at most worksites we interviewed at least two managers. This was a convenience sample. Because of the difficulties in gaining access to worksites, and the fact that “family friendly” employers tend to be more receptive to requests for interviews than other employers, the establishments we visited are not a representative sample of California employers (unlike the surveys). Rather the fieldwork was biased in favor of employers with a relatively high level of concern with work-family balance, most of whom offered unusually extensive family-leave-related benefits. Although a few small businesses are included, most of our cases are relatively (p.135) large establishments; in this respect as well they are not representative of the overall population of employers in the state. Indeed, as table 4.1 shows, all but one of the case study establishments we visited in 2004 were covered by the FMLA, which does not cover small employers. About a third of the employers we visited were members of the UCLA-based Human Resource Round Table who agreed to participate in the study as a result of a request from its director. Another set of site visits (about a fourth of the total) were arranged with members of One Small Step, a San Francisco Bay Area employer group focused on work-family issues. Access to the rest of the cases came through contacts from our personal networks. Only one of the many “cold calls” we attempted resulted in a successful interview. We added one more case in 2010, case T, a public-sector hospital that we had not visited in 2004. Cases H and I are divisions of a large food company, where we conducted separate interviews in 2004; in 2010 information on both divisions was provided by the same respondent. We also conducted similar fieldwork interviews in New Jersey prior to the introduction of paid family leave insurance there; our findings from that research are reported in Appelbaum and Milkman 2006.

(4.) Nonexempt employees are those covered by basic employment laws like the 1938 Fair Labor Standards Act, which established the minimum wage and the forty-hour work week as well as mandating extra pay for overtime hours. Many states have passed similar legislation setting higher standards than those required by federal law. Managers and some other upper-level employees are “exempt” from such laws, hence this terminology.

(5.) The Ns for figure 4.2 were 96 for the question on profitability, 95 for the questions on productivity and turnover, and 92 for the question about morale.

(6.) N = 169.

(7.) For nonprofits, which make up about one-fifth of the overall sample, the question was asked about “performance”; the rest were asked about “profitability.”

(8.) Four of the twenty establishments did not have any experience with the PFL program. The public sector hospital (case T) is not covered by the PFL program at all (public-sector workers are not covered unless they actively opt-in through the collective bargaining process). Cases L, P, and S all chose to be self-insured, offering somewhat superior benefits to those in the state program.

(9.) Most of the differences shown in table 4.6 by establishment size were not statistically significant, with the following exceptions: small establishments were significantly different from medium and large ones in the frequency with which they assigned work to others (p < 0.05); small establishments were significantly different from medium ones in the frequency with which they put work on hold (p < 0.001); small establishments were significantly different from medium ones in the frequency with which they had workers do some work while on leave (p < 0.05); small establishments were significantly different from medium ones in the frequency with which they hired replacements (p < 0.05); medium establishments were significantly different from large ones in the frequency with which they hired replacements (p < 0.001); and small establishments were significantly different from medium ones in the frequency with which they reported using some other method to cover the work (p < 0.05). Note that the significance results denoted by asterisks in table 4.6 are not by firm size but rather for the comparisons between exempt and nonexempt workers.

(10.) Prior to computing these estimates, we imputed missing data for cases where information was missing on some elements needed to calculate the cost of turnover. Imputation was performed by first separating the data into three establishment-size strata (fewer than 50 employees, 50–499 employees, and 500+ employees), and then, within each stratum, imputing missing values as the median value for the variable. We used median values (rather than means) so that the final results would err on the conservative side and not be distorted by a few high values. The imputation process had a small effect on median turnover costs ($4,174 unimputed for nonexempts compared with $3,965 imputed; $12,578 unimputed for exempts compared with $12,794 imputed). We then calculated average cost of turnover for (p.136) exempt and nonexempt employees to facilitate comparison with the average annual earnings. We weighted average costs of turnover to account for overrepresentation of some firm sizes in the sample. Average turnover cost unimputed for nonexempts is $4,474 and for exempts is $11,834. The unimputed turnover cost for nonexempts is 14% of annual earnings; for exempts, it is 20% of annual earnings.