Earlier this month, the Freedom Foundation formally challenged Spokane City Council President Ben Stuckart to debate the merits of mandatory paid sick leave laws. Stuckart has yet to respond to that invitation, though Blaine Stum, chair of the Spokane Human Rights Commission, recently emailed the Freedom Foundation a six-page letter with objections to our research paper on mandatory paid sick leave laws.

By way of context: Councilmember Mike Fagan cited Freedom Foundation research about the ineffectiveness of past paid sick leave laws at the city’s April 20 council meeting, drawing a response from Stuckart that he had “six pages of responses” to the Freedom Foundation paper that he would “save for the night we pass this policy.”

We contacted Stuckart following his comments to hear his objections to our research, noting that “the time for a discussion of the policy implications of such an ordinance is now, not after the outcome has been determined.”

Stuckart chose not to respond, leading to our invitation to debate on July 9. In the meantime, the council has continued to move full speed ahead towards adoption of an ordinance, with the officially sanctioned work group meeting for just six hours to discuss what an ordinance might look like.

In a July 17 interview on Freedom Daily, Fagan noted that the question of whether to adopt a paid sick leave law was never in doubt and expressed concern about the pre-determined outcome, especially given the lack of detailed information about Spokane’s economy and the effect sick leave regulations might have.

Instead of defending the effectiveness of mandatory paid sick leave regulations in a public forum, however, Stuckart (presumably) had Stum question the accuracy of our research on paid sick leave in a six-page letter. Interestingly, it does not appear to be the six-page document to which Stuckart was referring on April 20, as the letter’s introduction indicates it was written after the April 20 meeting. [Update 7/24/15: Though Stum’s initial letter to the Freedom Foundation was not the document to which Stuckart was referring on April 20, Stum forwarded an email in response to this post indicating that he had sent Stuckart a different document containing similar objections to our paper on the evening of the April 20 council meeting]

Nevertheless, Stum’s letter contains no substantive objections to our original paper, which was informally reviewed by numerous academic and professional economists prior to publication.

Our challenge to Council President Stuckart remains as issued.

A point-by-point response to Stum’s letter is provided below.

“In your examination of the Institute for Women’s Policy Research (IWPR) study on San Francisco’s earned sick leave ordinance, you suggest that the IWPR study should have focused on ‘the 34.9 percent of surveyed employers not already providing earned sick leave at the time of the law’s implementation.’ However, the IWPR study showed that 17% of employers surveyed had started offering earned sick leave since the law went into effect; something you note just a paragraph above the statement I have highlighted.”

A full reading of those two paragraphs in the original Freedom Foundation paper should make the point quite clear: The study should analyze those businesses actually affected by the San Francisco ordinance, i.e., businesses that had to create new or modify existing paid sick leave policies because of the ordinance.

Stum pretends to point out a factual inconsistency by claiming we confuse the 34.9 percent of employers that had to create new or modify existing paid sick leave policies with the 17 percent of employers that had to create new policies. The original report makes the distinction quite clearly, however, noting,

Of the employers surveyed, 34.9 percent were newly covered or affected by the PSLO in some way, and only 17 percent had started providing paid sick leave for the first time.

When we write in the next paragraph that, “the focus of study should be the effect of the PSLO on the 34.9 percent of surveyed employers not already providing paid sick leave at the time of the law’s implementation,” it should be quite clear from the context that we were referring to the 34.9 percent of businesses not already providing sufficient paid sick to comply with the ordinance.

Additionally, Stum raises no objection to the more substantive point that the effect of the ordinance on businesses not already in compliance with the law voluntarily should be the primary focus of study.

“You state, quite correctly, that ‘the data in the IWPR study does not allow for the connection’ between businesses that began offering earned sick leave when the law went in to effect, and the impact on those businesses. Yet you suggest that the data ‘offers a strong correlation between businesses providing earned sick leave for the first time and businesses reporting difficulties with the PSLO.”

Again, Stum pretends to find a contradiction where none exists. We say no more than the data allows. There is strong correlation between the number of businesses affected by San Francisco’s ordinance (34.9 percent) and the number reporting it was difficult to understand (34.6 percent), administer (31.4 percent) and reassign or delay work because of the PSLO (42.7 percent). 

However, the data in the IWPR report do not allow us to confirm with certainty that the same businesses reporting difficulty were the same ones implementing new or modifying existing sick leave policies to comply with the ordinance.

Our summary of the evidence strikes the appropriate balance:

While the data in IWPR’s study does not allow for the connection to be made explicitly, it does offer evidence of a strong correlation between businesses providing paid sick leave for the first time and businesses reporting difficulties with the PSLO.

For his part, Stum provides no reason to believe that the businesses reporting difficulty with the ordinance were not actually the businesses affected by it.

“You also take issue with the IWPR’s statement that ‘workers with paid sick days are less likely than others to switch jobs.’ After examining the studies they cite, you conclude that ‘health concerns are conspicuously absent from a major meta-analysis of studies examining the causes of employee turnover, which could indicate that it is likely not a significant factor in overall turnover to warrant serious examination.’ There are several problems with this statement.”

Actually, as we shall explain in the ensuing paragraphs, our conclusion was that,

…mandatory paid sick leave advocates rely on faulty research, falsely assume turnover is necessarily harmful, confuse the important difference between individual employers providing the benefit voluntarily and all employers being required to, and present no evidence that mandatory leave policies succeed in reducing turnover in any meaningful way.

“First, studies suggest health related concerns and health related workplace policies are a predictor of job turnover. Access to health insurance has been associated with a phenomenon known as ‘job lock,’ meaning employees are less likely to leave their jobs when they have access to health insurance than when they do not. Certain illnesses, poor general health and disabilities have also been found to be predictors of involuntary and voluntary job separations.”

True, many studies have shown that employees are less likely to leave their jobs if they are getting health insurance through their employer. However, this is not necessarily a good thing and, even if it is, does not necessarily mean offering paid sick leave (a far less valuable benefit) has the same effect.  

“Second, you completely misread Cooper and Monheit’s (1993) study. According to your report, ‘the only time paid sick leave comes up in the report is in the finding that ‘other non-portable fringe benefits [paid sick leave being one] on an initial job also may inhibit job mobility.’’ This is categorically false. In tables 5 and 6, Cooper and Monheit estimate that access to earned sick leave reduces voluntary turnover by 5.58% for married men, 3.61% for married women, 5.75% for single women and 6.43% for single men. This is not hidden somewhere deep in the footnotes. It’s displayed prominently in the tables they base their conclusions on.”

Stum is correct that, in the literal sense, paid sick leave is referenced more than once in Cooper and Monheit’s paper. We simply referenced the one-sentence summary provided in the text of the study rather than the line in a chart the conclusion was based on. 

But if anyone is misreading Cooper and Monheit, it is ideologues like Stum. We explain in our reporta point Stum conveniently ignores:

The 1993 Blue Cross-Blue Shield paper examined “the relationship between employment-related health insurance and job mobility.” After analyzing now decades-old National Medical Expenditure Survey data from 1987, the authors found “support for the hypothesis that employment-related health insurance inhibits job mobility.” The only time paid sick leave comes up in the report is in the finding that “other nonportable fringe benefits (paid sick leave being one) on an initial job also may inhibit job mobility.” From this narrow finding, mandatory paid sick leave supporters extrapolate the broad conclusion that inhibited job mobility means reduced turnover and savings for businesses. Such is hardly the intent of the study. Far from advocating for mandatory paid sick leave policies as a way to decrease turnover, the study’s authors argue in favor of greater job mobility, contending:

“Perfect mobility in labor markets enables workers to determine their value in any employment situation and, therefore, to seek and obtain employment in industries and occupations where they are most productive… (Barriers to job mobility) can affect labor productivity and income and, ultimately, the volume and quality of goods and services produced in the economy.”

The paper’s data and conclusion — “Informed health care policy should ensure that work-related health insurance does not pose a barrier to job mobility” — simply does not support the idea that mandatory paid sick leave policies save employers money by reducing turnover.

Third, several other studies have shown that earned sick leave is considered to be an important employee benefit and does in fact reduce turnover. For instance, Weick et al (2009) found that nurses across generations rank PTO, which includes earned sick leave, as one of the top five most important workplace incentives. The study further found that incentives played a role in job satisfaction, which was correlated with turnover intentions. Pavalko and Henderson (2006) found that women in caregiving jobs were more likely to remain employed and maintain work hours over a two-year period with jobs that provide access to three benefits (flexible hours, unpaid family leave, and earned sick leave/PTO).

A more recent study by Hill (2013) that focused exclusively on earned sick leave found that it decreases the probability of job separation by at least 2.5 percentage points or 25%. Numerous empirical studies have also found that work-family conflicts, which can arise with a lack of earned sick leave, impact the physical and mental health of employees, and decrease job satisfaction and organizational commitment. Finally, earned sick leave has the ability to increase schedule control, which is associated with job stability. I understand the need for more empirical research on the topic of earned sick leave and employee turnover. However I do not find it prudent to blithely dismiss both current theoretical and empirical research.”

1. We do not argue anywhere in our paper that an employer who decides to offer paid sick leave to his employees voluntarily will not experience some reduction in turnover. Indeed, it makes perfect sense that an employer may reduce turnover by increasing the compensation and benefits package for his employees, as long as other there are other employers not offering the same benefits.

However, as we explain in our original report a point Stum again ignores:

Even if an employer can reduce turnover by voluntarily providing benefits like paid sick leave, a policy mandating all employers provide the benefit would not have the same effect. The primary reason an employer might be able to reduce turnover by providing paid sick leave is if it makes him look good in comparison to his competitors. If all employers must provide paid sick leave, however, and workers can expect essentially the same sick leave benefits from any employer, then paid sick leave is effectively nullified as a factor in employees’ decisions about where to work.

A separate and smaller survey of San Francisco employers conducted by the Urban Institute indicated that the PSLO failed to reduce turnover for this very reason:

“Employers noted that turnover and retention seem less relevant to a mandated benefit, since now the same sick leave benefits are available across companies. As one small business owner observed, ‘The policies I had in place before were there to reduce turnover and get better employees—and they did have an effect. But now, since the new ordinance, employees will have the same benefit no matter where they work. There’s less of an incentive to stay and work for me.’”

Perhaps for this reason, no credible study has ever found that mandatory paid sick leave laws play any significant role in decreasing turnover.

2. None of the studies cited by IWPR or Stum examine the effect of government-mandated sick leave polices. Indeed, only two such studies have attempted to measure the effect of mandatory paid sick leave on turnover, both of which are addressed in the Freedom Foundation paper.

The study of Connecticut’s paid sick leave law by the Center for Economic and Policy Research (CEPR) found that 3.3 percent of surveyed employers reported reduced turnover following the law’s implementation. However, the structure of CEPR’s survey questions did not allow employers to report worsened or unchanged turnover.

In contrast, the second Seattle City Auditor study provided employers with a range of options, finding that 98.3 percent of employers reported no changes in turnover, 1.1 percent reported reductions in turnover, and 0.6 percent reported worsened turnover.

3. Theoretically, if all businesses are required to offer the same sick leave benefits, businesses that were previously more attractive to employees because they voluntarily provided sick leave benefits could become relatively less attractive to employees once all employers are required to offer the same benefits.

4. Lastly, if employers could truly reduce turnover by offering paid sick leave in a manner that produced net benefits for their business, they would likely happily do so without the necessity of a government mandate.

You also dispute the idea that there are high costs associated with employee turnover. However, a recent literature review of 22 case studies for workers earning less than $50,000 annually suggests that the cost of replacing a workers is 20% of salary. Black Frank (2000) estimates that the cost of replacing a single cashier can range from $2,200 to over $4,000. Hinkin and Tracey (2000) estimate the cost of replacing a front office worker at a hotel to range from $5,600 to nearly $6,000. When comparing these costs to the costs of earned sick leave via the Bureau of labor Statistics, it’s clear that turnover is a much more costly affair for employers.”

Certainly turnover can be expensive. However, that does mean that it is always undesirable, as we explain in our original paper:

A 2010 paper observed that “many managers hold important misconceptions about turnover,” one of which is “that turnover is uniformly bad.” While acknowledging, “Employee turnover can certainly be problematic,” the authors observe that, “turnover is a complex phenomenon that comes in many shapes and sizes. It is not always harmful, and in some cases may even be beneficial for organizations.” Harboring misconceptions about turnover is harmful, they contend, because it “may lead managers to enact ineffective retention strategies that fail to reduce turnover” or “are not cost-effective.” By assuming that turnover is necessarily detrimental and expensive, paid sick leave advocates are over-simplifying a complex issue and advocating an ineffective solution.

Stum ignores these nuances entirely. And again, the cost of turnover is irrelevant if mandatory paid sick leave regulations do nothing to diminish it.

“You suggest that the IWPR’s data proves the earned sick leave ordinance did not have any impact on employee presenteeism. There are two problems with this assertion. First is that it lacks context: As Jessica Milli with the IWPR noted, the data collection took place during the H1N1 epidemic. Second, as research has consistently shown, be nefit utilization is influenced by workplace atmosphere, along with several other factors. This does not mean earned sick leave policies are ineffective at reducing employee presenteeism. It means that a supportive, flexible work environment is critical to ensuring employees are actually able to use earned sick leave.”

1. For starters, if any report lacked context it was IWPR’s. The finding that employee presenteeism had not declined was tucked away in an appendix at the back of the report and was not mentioned at all in the text of the report. Milli’s claim about the H1NI outbreak was not made until September 2014, three years after the initial publication of the IWPR report and a month after the publication of the Freedom Foundation paper in August 2014.
2. Although the H1N1 outbreak may have influenced the survey results to make it appear workplace illness was higher than it was, there is no way to know for sure.
3. Furthermore, several other studies of different paid sick leave laws conducted well after the H1N1 outbreak also found no decreases in presenteeism following implementation of the mandate, including two studies conducted by the University of Washington on behalf of the Seattle City Auditor. The only study that claimed to find a decrease in presenteeism – the Center for Economic and Policy Research’s study of Connecticut’s paid sick leave law – did not give employers the option to report unchanged or increased presenteeism, as the IWPR and U.W. studies had. As we note in our paper,
   

   Without knowing how many employers would have reported worsened presenteeism or no change at all, CEPR’s finding that 18.8 percent of employers reported decreased presenteeism is of little use.

4. Stum implies that the lack of a decline in presenteeism could be explained by employees not using their sick leave due to unsupportive work environments. However, the study he cites as proof merely notes that an employee being “positive about social-medical support during sick leave” is one of many factors which could influence how often and for how long the employee takes sick leave, but notes that the literature on the subject “is scarce or ambiguous.”

It is alternatively possible that some employees utilize their sick time as vacation time and end up coming to work sick anyway. In one survey, half of U.S. workers admitted to calling into work sick when they were not actually ill. Either way, the existing studies do not offer definitive explanations for why workplace illness fails to decrease after passage of a mandatory paid sick leave law.

“Much of the commentary of the Colla study has centered on the fact that 25.3% of employers reported worse profitability as a result of earned sick leave policies. The problem with touting this finding is that the survey did not seek to determine whether other variables may have caused decreased profitability; it merely asked whether employers thought these policies contributed to decreased profitability. This is an important distinction, as the survey was conducted in 2009 during the height of the recession.”

1. Every study that has examined the question has found a noteworthy number of employers have experienced increased costs or decreased profitability following passage of a mandatory paid sick leave law, though the percentage obviously changes based on the time, place and number of employers voluntarily offering sick leave before the mandate. The finding has been consistent in studies conducted both during and after the recession.
2. However, Stum is correct that the Colla study merely surveys employers’ opinions. Some employers responding to the survey may have incorrectly attributed cost increases caused by other factors to the new sick leave mandate, just as some may have inaccurately attributed increases in employee morale to the sick leave mandate. Undoubtedly, the research could be much stronger. However, it is worth noting the eight of the 10 studies that examined the effectiveness of existing paid sick leave laws were based on surveys of employers and/or workers. Three of the eight studies were conducted by paid sick leave supporters, three by city auditors and two by sick leave opponents. The two remaining studies were so methodologically unsound as to be entirely useless. If Stum wishes to reject all studies that survey employers, advocates of paid sick leave will be left without much of the research they typically rely on.

“In analyzing the Colla study, you also appear to selectively interpret which findings are deemed to be ‘significant.’ You note, for instance, that ‘11.9% percent [sic] (of business surveyed) reported raising prices and 7.4% reported decreased customer service.’ In the next paragraph, you go on to say that the ordinances [sic] beneficial effects on productivity are ‘not a significant factor’ because 11% of businesses effected [sic] reported an increased in employee morale and 6% experience reduced presenteeism.”

Context, context, context. The comments Stum refers to were made in a list of reasons provided by the authors of the Colla study as to why only about a third of businesses affected by San Francisco’s sick leave ordinance reported decreased profitability:

1. Some firms reduced other forms of employee compensation. According to the study, 25.8 percent of affected firms reduced compensation, 20.3 percent reduced vacation time, and 9.5 percent reduced pay and bonuses.
2. Some firms passed on their costs to consumers. 17.9 percent of affected firms reported some “negative consumer effect,” while 11.9 percent reported raising prices and 7.4 percent reported decreased customer service.
3. Thirdly, the authors speculate that “there may have been beneficial effects on productivity” based on the finding that 11 percent of affected firms reported higher employee morale and 6 percent reported decreased workplace illness. Our report notes that these findings “(do) not appear to be a significant factor” in explaining why only a third of affected businesses reported reduced profitability for several reasons. First, the Colla study’s authors note that businesses already offering paid sick leave at the time the ordinance was implemented were almost as likely to report a decrease in workplace illness as businesses adopting new sick leave policies, meaning it is impossible to conclude from the Colla study that any measureable decline in workplace illness occurred. Second, the fact that 11 percent of affected employers reported increased employee morale is relatively insignificant compared to the far larger percentages of employers that reported concrete negative consequences.

“In your discussion the Auditor’s report, you seize on the fact that 12.5% of businesses suggested they may move from the District of Columbia due to the requirement to provide earned sick leave. You characterize this finding as ‘startling’ and suggest that ‘most of the newly affected businesses expressed a willingness to potentially cease their operations in the District because of the mandate.’ There is no data in the auditor report, or any reports on D.C.’s implementation of sick leave, to support such a statement and as researchers know quite well intent is not always followed by action.”

1. On the contrary, our summary of the D.C. Auditor’s survey is quite nuanced and balanced. When asked by the Auditor if the paid sick leave requirement would cause them to move out of the District, 12.5 percent of employers responded that it would. The Auditor also determined that 50 percent of surveyed employers were already offering sick leave benefits at the time of the ordinance’s implementation and that 31 percent were not offering sick leave and were out of compliance with the ordinance at the time of survey. Consequently, the only businesses newly affected by the paid sick leave requirement were the 18 percent that began offering paid sick leave for the first time between 2008 and 2012. With this in mind, we conclude that,

If only 18 percent of surveyed employers were affected by the law, and 12.5 percent of employers said the law could cause them to move, then most of the newly affected businesses expressed a willingness to potentially cease their operations in the District because of the mandate.

While Stum is correct that “intent is not always followed by action,” we make no such argument. The Auditor’s survey is typically cited by advocates of mandatory paid sick leave, including the city of Spokane, as proof that businesses will not leave a city because of a sick leave ordinance. Our contention is simply that the Auditor’s survey supports no such conclusion.

2. Additionally, Stum does not respond to our observations that, since the Auditor’s survey was conducted several years after the enactment of the ordinance and would have left out any businesses which left the district during the intervening period, it “most likely understates the true impact” of the law. Similarly, Stum fails to acknowledge our observation that the Auditor’s report, “did not take into account the full range of business responses to the law, like raising prices, decreasing employment or limiting other employee benefits.”

“In relation to the impact of the law on employees coming to work sick, you infer that had respondents been given different question wording on presenteeism, ‘some employers may have responded, as did some in San Francisco and Seattle, that the sick leave law had worsened presenteeism in their workplace.’ You then proceed to dismiss CEPR’s finding as ‘of little use.’ We have no way of knowing how different question wording or answer structures would have affected these results, but to cavalierly declare the results ‘of little use’ because two other surveys used different wording and structure is not an example of strong research. If you have conducted a thorough review of how the wording and structure of these questions bias results, please feel free to share.”

Actually, the original Freedom Foundation report cited and linked to the questions from the CEPR survey. Regarding presenteeism, employers were asked: “Has the paid sick leave law had any of the following effects on your organization? … Reduced the number of employees who come to work sick.”

Employers were not asked about or given the ability to report unchanged or worsened presenteeism.

Contrary to Stum’s contention, our original report was not concerned with the wording of the questions, per se, but the range of potential answers provided to employers. When employers in Seattle and San Francisco were allowed to choose between reporting improved, unchanged or worsened presenteeism, the results indicate there was no net change.

The fact that 18 percent of surveyed employers reported improvements in presenteeism means little without knowing how many would have reported worsened or changed presenteeism had they been given the option.

“There are some serious problems with using this study to legitimize claims of negative impacts on businesses. As the study itself notes, businesses who responded were ‘those most likely to hold strong feelings about the law.’ With a response rate of only 20% and a small sample size, this means the sample of businesses was not representative of businesses that had begun to offer sick leave after the law went in to effect, let alone all businesses.”

Our original report was intended as a review of the research on existing mandatory paid sick leave regulations. The Employment Policies Institute’s surveys fit under this banner and comprise only two of the 10 studies examined that have been conducted by an organization that opposes sick leave mandates.

We clearly note the limitations of EPI’s survey of Connecticut employers:

The study comes with several caveats. First, EPI notes that “the response rates for the survey were low – roughly 20 percent.” Second, instead of using a random sampling like the CEPR study, EPI worked with the Connecticut Business and Industry Association and the Connecticut Restaurant Association to identify and survey businesses most likely to be affected by the mandate. Consequently, EPI notes that its study “makes no claim of being representative of the broader Connecticut business population.” Still, the survey provides a more detailed look at the consequences of mandatory paid sick leave for the businesses most likely to be affected.

The survey zeroed in on 86 Connecticut businesses that had begun offering paid sick leave for the first time because of the new law and found that about half took some form of mitigating action like raising prices, reducing compensation or benefits, or limiting expansion.

Our conclusion notes,

Overall, although EPI’s survey is limited in scope, it indicates that mandatory paid sick leave policies can produce appreciable negative consequences for businesses forced to provide sick leave benefits for the first time.

Stum provides no reason to believe our assessment of the survey’s value and limitations is inaccurate.

“In discussing impacts on employer operating costs your report conspicuously leaves out two facts: 1) most employers did not track costs related to implementing new earned sick leave policies and 2) of those employers who did systemically track costs related to providing earned sick leave, the average cost was only four tenths of one percent of total revenue. This is relatively close to the costs of providing earned sick leave calculated by the Bureau of Labor Statistics. According to data from the National Compensation Survey, the cost of providing sick leave to all workers in private industry is 0.8% of total compensation, or $.26 per hour worked per employee. For the service industry, the cost is 0.5% or $0.07 per employee per hour worked. That means for a restaurant that employs ten people, the cost would be an estimated $1,456.”

1. Stum’s use of the word “conspicuously” frames his observation as an accusation without providing any reason for the significance of the omission. The second Seattle City Auditor report alone totaled 108 pages. Much was omitted for the sake of brevity. His implication appears to be that the cost of providing mandatory paid sick leave benefits is relatively low. Certainly, paid sick leave mandates are not likely to be as impactful as large increases in the minimum wage. The point is, contrary to the claims of many labor activists, there is a cost involved.
2. Additionally, the administrative costs of complying with ordinance and potential litigation can increase the costs on employers beyond the direct cost involved in providing the benefit.
3. Lastly, the fact that many employers “do not plan to identify costs or do not have the resources to do so,” as the Auditor’s report puts it, does not tell us anything about the cost of compliance. As the Auditor’s report explains, the employers that did track the cost of complying with the ordinance were, “self-selected and too few to provide statistically representative data… Calculating cost is not a straightforward procedure, and different employers may use different approaches depending on their situations and accounting practices.”

“Rather notably, your commentary on Seattle’s law does not include any discussion regarding safe leave despite the fact that 33.5% of employers had expanded policies to include safe leave. This is important because domestic violence often spills in to the workplace.”

Most mandatory paid sick leave ordinances, including Seattle’s, permit accrued time to be used for “safe time.” When evaluating the impact of Seattle’s ordinance, the Seattle Auditor’s report looked at the consequences of the ordinance as a whole and often did not differentiate between sick and safe leave. For instance, while it is true the Auditor found that 33.5 percent of employers expanded their policies to include safe leave, the Auditor also found a total of 37 percent of employers changed their sick and/or safe leave policies in response to the ordinance. We examined the impact of Seattle’s ordinance through the lens of the more comprehensive number – 37 percent – rather than try to parse out the effect of the sick leave expansion versus the safe leave expansion with data that often only considered the impact of the ordinance as a whole.

As with many of his other objections, Strum’s point on safe leave is nonsensical.

“The Seattle survey also noted 15 employers went out of business in between the first survey and the second. There are no reasons given for these closure. Despite that, you suggest that anywhere between 60-300 businesses may have closed throughout the city in between surveys due to the implementation of the ordinance. After making this unverifiable assertion, you state, ‘… it is more than a little misleading for the study’s authors to claim that the ordinance did not cause businesses to close.’ I find it more misleading to assume something which the data provided cannot tell you.”

Yet again, Stum mischaracterizes our argument, which is stated quite clearly.

The study’s authors note that (1) 0.6 percent of surveyed employers reported closing or moving operations out of Seattle because of the ordinance and (2) 11,174 Seattle businesses were covered by the ordinance. As we note in the original report,

…if the 0.6 percent ratio in the survey holds up for all 11,174 covered employers, then the ordinance caused more than 60 businesses to move or shut down locations in Seattle in the year following the ordinance.

We point out further that,

During the process of attempting to follow up with the employers from the first (baseline) survey (conducted just prior to the implementation of the ordinance), the study’s authors note that “15 were determined to have gone out of business.” No further mention of these employers is made. These 15 employers represent 2.7 percent of the 551 businesses from the original survey. If 2.7 percent of employers from the baseline survey closed in the first year following the ordinance, and the same rate applies to all 11,174 Seattle businesses estimated to be covered by the law, then as many as 300 businesses closed in the interim period between the two surveys. Any or all of these could have potentially closed due to the ordinance.

Given that at least 60 and as many as 300 businesses closed because of the paid sick leave mandate, our conclusion stands:

Given these realities, it is more than a little misleading for the study’s authors to claim that the ordinance did not cause businesses to close.