Log in

Assessing the Employment Effects of California’s Paid Family Leave Program

  • Original Article
  • Published:
Eastern Economic Journal Aims and scope Submit manuscript

Abstract

Until recently, it was impossible to study how paid family leave mandates would impact employment in the USA. This changed in 2004 when California implemented the USA’s first paid family leave legislation. California’s program provides us with a quasi-natural experiment to study how the implementation of paid family leave affected employment in the state. This paper uses several difference-in-difference, fixed-effects regressions on a large panel of business establishments to study the impact California’s program had on establishment-level employment. Most model specifications show that the law is correlated with an increase in employment in firm establishments in the state.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or Ebook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1

Similar content being viewed by others

Notes

  1. A 2017 Pew Research Center study found that workers who did not receive their full wages or salary during leave for family, parental, or medical reasons had to make financial sacrifices such as cutting back spending, drawing from savings, or had to take a leave that was shorter than they would like in order to compensate for the loss in income. Of those with the lowest incomes (less than $30,000 a year) and did not receive full pay while on parental leave, about 60% had to go into debt, 48% had to go on some form of public assistance, and 46% got behind on paying their bills (Horowitz et al. 2017).

  2. These states include New Jersey (enacted in 2008), Rhode Island (2013), New York (2016), D.C. (2017), Washington (2017), Massachusetts (2018), Connecticut (2019), Oregon (2019). While the individual programs are different in terms of how generous their benefits are in terms of the number of paid weeks provided as well as the percentage of income the wage replacement provision provides, they all are similar to California’s program in that they are for the purpose of bonding with a new baby or taking care of an ill family member. They are also all implemented through an insurance-type program with contributions collected through payroll taxes (paid by employees, employers, or some combination of the two depending on the state) ("State Paid Family and Medical Leave Insurance Laws" August 2019).

  3. Among OECD countries in 2018, all countries (with the exception of the United States) provide mothers at least 14 weeks of maternity leave with partial wage replacement, with an overall average of just over 18 weeks. Most of the wage replacement provisions provided at least 50% of pre-birth earnings. ("OECD Family Database").

  4. While this dataset is unique in that it provides records of employment over time for a large subset of firms in the USA, it must be noted that because of the way that the EEOC collects data, the majority of establishments in this data have more than 100 employees, and relatively very few of the establishments have less than 50 employees (some establishments that have employment levels less than the 100/50 threshold still report their employment for various reasons). It is therefore the case that while this dataset can provide important insight into how larger firms are responding to a paid leave mandate, it can say very little about smaller establishments.

  5. The department of labor found that 40% received full pay when taking leaves that were longer than ten days compared to 60% for those that took shorter leaves (Klerman et al. 2012).

  6. In fact, according to a 2003 study of parental leave usage among both fathers and mothers, the FMLA had no significant effects on the instance and length of leave taken by fathers (Han and Waldfogel 2003). The authors conclude that the limited impact of the FMLA on the instance and length of leave among new parents may be due to the fact that it does not guarantee paid leave and that parents may be unable to take significantly more leave when a new child is born if the policy only provides job protection and does not include a wage replacement provision.

  7. In July 2014, the law was expanded to cover the care of a seriously ill parent-in-law, grandparent, grandchild, or sibling. However, the data used for this study only cover the period of time before this expansion and therefore do not include leaves taken for these purposes. http://www.edd.ca.gov/disability/PFL_Eligibility.htm.

  8. CAPFL was implemented as part of the state’s existing State Disability Insurance program with a tax rate of 1% of eligible wages.

  9. However, it does not provide leave-takers with job protection; therefore, in order for the employee to be guaranteed to return to their preleave job, it must be used concurrently with the Family Medical Leave Act or California’s Family Rights Act (Milkman and Appelbaum 2013).

  10. In the case of California’s Paid Family Leave Program, the supply effects are likely to be small, due to the nature and size of the benefit. If a worker utilizes the program, at most it represents 6% of their annual wages (and for most it would be much less, due to the weekly maximum imposed as well as the fact that many participants in the program would not take a full six weeks). It is also the case that only a small fraction of the labor force utilizes the program each year.

  11. The analysis unfortunately ends at 2009 due to availability of the data. The data that the EEOC provides are confidential, and gaining access requires the researcher to follow a strict, multistep approval process. Through this process, the author only had access to data up until 2009. Future research on this topic would be greatly enhanced if these data were more broadly available to researchers.

  12. Industry variables were created using self-reported two-digit North American Industry Classification System (NAICS) codes. (Some industries are not represented because of the EEOC reporting and data collection system. The dataset does not include those establishments or organizations that deal with Public Administration or Education, as those reports are kept separately).

  13. This model most likely suffers from two problems, both of which will cause a downward bias of the standard errors. (1) The number of observations (employment at the establishment level) is relatively large compared to the observations of the coefficients (state and year) and (2) serial correlation in the dependent variable. Therefore, the standard errors are clustered by state rather than year (Angrist and Pischke 2009).

  14. Rankings for the various business indexes were chosen from year 2007 and 2009. These years were chosen for a variety of reasons. The first reason concerned availability of data. Some indexes used in this analysis were not available before 2006. However, most importantly, in a study of business climate indexes and economic growth, Kolko, Neumark and Mejia found that “there is little variation over time, within states, in business climate indexes” (2013) meaning that the year of choice may not be very important. Another 2014 study on the SBSI in particular found similar results (Shukla and Shukla 2014).

  15. The psmatch2 package in Stata was used for this analysis, which utilizes full Mahalanobis matching to adjust for pretreatment observable differences between a group of treated and a group of untreated (Leuven and Sianesi 2003).

  16. The synth_runner package was used for this analysis which uses the synthetic control method outlines by Abadie et al. (2010) but allows for multiple treated units (Galiani and Quistorff 2016).

  17. The metropolitan areas above Douglas County was also added, including Carson City and the lower part of Washoe County in order to have enough establishments in the control group. This area is very close geographically to Douglas County and is connected to the Sacramento Metropolitan Area by Interstate 80.

References

Download references

Acknowledgements

The authors would like to thank M.V. Lee Badgett for helpful comments and encouragement. The author is also grateful for the constructive input given by the two anonymous reviewers. This research was conducted while the author was a Graduate Student Fellow at the Center for Research on Families at the University of Massachusetts-Amherst and is grateful for the support the center provided.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Samantha Marie Schenck.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Schenck, S.M. Assessing the Employment Effects of California’s Paid Family Leave Program. Eastern Econ J 47, 406–429 (2021). https://doi.org/10.1057/s41302-021-00193-9

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1057/s41302-021-00193-9

Keywords

JEL Classification

Navigation