Workplace Fairness

Menu

Skip to main content

  • print
  • decrease text sizeincrease text size
    text

“The Algorithm Made Us Do It”: How Bosses at Instacart “Mathwash” Labor Exploitation

Share this post

Image result for Audrey Winn"

Instacart is messing with workers’ tips, again. The company’s workers are so fed up hundreds of them are out on strike this week.

Instacart—a gig economy company for same-day grocery delivery—has had problems with tipping date back to 2016. At that time, Instacart removed tipping from the app, before being shamed into reinstating a tipping policy the next month. Then, in 2018, the company altered its policy again by counting customer tips toward workers’ guaranteed $10 base pay—leading to situations where customers were paying almost the full base, with little contribution from Instacart. Now, Instacart is taking aim at the default tip amount. When customers finish their Instacart orders, the app had previously suggested a tip of 10%. This was unilaterally discontinued and replaced by a 5% default.

In response to the default tip change, Instacart worker and organizer Vanessa Bain penned an impassioned Medium post last month which inspired a walkout of more than a thousand workers demanding reinstatement of the 10% default. Instead of improving conditions in the workplace, their collective action was met with discouraging news. Two days after the walk-out, Instacart slashed workers’ “quality” bonus pay—one of the only remaining pay incentives on the app, and an incentive that has been alleged to make up to 40% of the average Instacart workers’ already low income (some estimates put this between 30 and 35%). The company also did not respond to the concerns workers aired in the Medium post.

Starting December 16 and extending to December 21, over 300 Instacart workers are expected to strike again to challenge Instacart’s incentive cut, tip default changes, and declining work conditions generally, with events scheduled each day.

Amid mounting outrage, Instacart has attempted to deflect criticism by vaguely citing data. “During the last year, we offered a new version of the quality bonus and found that it did not meaningfully improve quality,” the company told shoppers over email in November, after the first walkout. “As a result, we will no longer be offering the quality bonus beginning next week.”

Through this statement, the company blamed unverified, unexplained metrics for the cuts, not its own exploitative model. The metric is presumably based on data, but workers and consumers are never given insight into that data. While the jargon is new, the underlying reality is not: A closer examination reveals this is just a justification for good, old-fashioned exploitation.

By what metric does Instacart measure whether an incentive can “meaningfully improve” quality? For an improvement to be “meaningful,” what quantitative or qualitative factors must be present? Is there a specific “quality” that is being measured, and how does it take into account worker quality of life? Furthermore, how does the company justify the gap between its lowest- and highest-paid employees? The average Instacart executive compensation is $279,596 a year—with the most compensated executive making $790,000. In contrast, the average Instacart worker is making between $9.81 and $12.96 an hour.

By brushing off worker complaints through references to unexplained data that is available to neither workers nor consumers, Instacart is attempting to utilize an insidious rhetorical tactic: “mathwashing.”

Coined by tech-entrepreneur Fred Benenson, the term “mathwashing” can be used to describe attempts to use math terms like “algorithm” to gloss over a more subjective reality. In the case of Instacart, algorithms are being used to justify poor work conditions, since a faceless algorithm is more convenient to blame than the greedy bosses behind the decisions. Benenson is clear in describing why this is a problem.

“This habit goes way back to the early days of computers when they were first entering businesses in the 1960s and 1970s,” he stated, in an interview with Technical.ly Brooklyn. “Everyone hoped the answers they supplied were more true than what humans could come up with, but they eventually realized computers were only as good as their programmers.”

Though Benenson originally used the term to describe how Facebook’s trending topics were not neutral, but instead manipulated by Facebook’s data engineers, it arguably applies to Instacart and a lot of the “don’t blame the bosses, blame the algorithm” language that is common across the gig economy. While other companies like Uber and AirBnb have relied on this rhetoric, however, Instacart is a particularly egregious abuser.

Talking with TechCrunch in 2016, CEO Apoorva Mehta relied on jargon and abstract language to defend workers’ low wages. He praised his workers’ “NPS score” and noted that wages were “not a zero-sum game” because “the problem that we’re trying to solve is very hard.”

Instacart’s process for deciding how to delegate orders is described by its website as a “Stochastic Capacitated Vehicle Routing Problem with Time Windows for Multiple Trips.” In describing delivery scenarios, Instacart’s website discusses using “time-based simulations” to replay “the history of customer and shopper behaviors with the existing algorithm and the new one.” The section shows colorful graphs and charts that fail to describe most of their variables, including one that simply lists “metric” instead of even pretending to have a quantity for measuring efficiency. The language is so loaded with jargon and italics that it is likely inaccessible to the average consumer or worker.

While this jargon conveys little, Instacart uses it to market the company’s “genius” design. To help readers understand that they are dealing with a company that is much smarter than themselves, Instacart includes a grocery-inspired illustration of Albert Einstein to accompany explanations of its black-box algorithim. Instead of leaving with a sense of awe, however, readers leave with a sense of having participated in a game of smoke and mirrors. The explanation reads less like a helpful primer and more like a desperate attempt to get consumers to believe anything other than the truth. Namely, that the company is the “despot” in control of its own algorithm.

This is not a marvel of technological innovation. It is a marvel of exploitation. You don’t need an advanced mathematics degree to know the score.

This article was originally published at InTheseTimes on December 18, 2019. Reprinted with permission.

About the Author: Audrey Winn is a Skadden Fellowship Attorney working and writing in New York City. She is passionate about workers’ rights, algorithmic transparency, and the inclusion of gig workers in the future of the labor movement.

Share this post

Trump’s Labor Dept. Has Declared War on Tipped Workers

Share this post

Image result for heidi shierholz

In October, the Trump administration published a proposed rule regarding tips which, if finalized, will cost workers more than $700 million annually. It is yet another example of the Trump administration using the fine print of a proposal to attempt to push through a change that will transfer large amounts of money from workers to their employers. We also find that as employers ask tipped workers to do more nontipped work as a result of this rule, employment in nontipped food service occupations will decline by 5.3% and employment in tipped occupations will increase by 12.2%, resulting in 243,000 jobs shifting from being nontipped to being tipped. Given that back-of-the-house, nontipped jobs in restaurants are more likely to be held by people of color while tipped occupations are more likely to be held by white workers, this could reduce job opportunities for people of color.

Employers are not allowed to pocket workers’ tips—tips must remain with workers. But employers can legally “capture” some of workers’ tips by paying tipped workers less in base wages than their other workers. For example, the federal minimum wage is $7.25 an hour, but employers can pay tipped workers a “tipped minimum wage” of $2.13 an hour as long as employees’ base wage and the tips they receive over the course of a week are the equivalent of at least $7.25 per hour. All but seven states have a subminimum wage for tipped workers.

In a system like this, the more nontipped work that is done by tipped workers earning the subminimum wage, the more employers benefit. This is best illustrated with a simple example. Say a restaurant has two workers, one doing tipped work and one doing nontipped work, who both work 40 hours a week. The tipped worker is paid $2.50 an hour in base wages, but gets $10 an hour in tips on average, for a total of $12.50 an hour in total earnings. The nontipped worker is paid $7.50 an hour. In this scenario, the restaurant pays their workers a total of ($2.50+$7.50)*40 = $400 per week, and the workers take home a total of ($12.50+$7.50)*40 = $800 (with $400 of that coming from tips).

But suppose the restaurant makes both those workers tipped workers, with each doing half tipped work and half nontipped work. Then the restaurant pays them both $2.50 an hour, and they will each get $5 an hour in tips on average (since now they each spend half their time on nontipped work) for a total of $7.50 an hour in total earnings. In this scenario, the restaurant pays out a total of ($2.50+$2.50)*40 = $200 per week, and the workers take home a total of ($7.50 + $7.50)*40 = $600. The restaurant’s gain of $200 is the workers’ loss of $200, simply by having tipped workers spend time doing nontipped work.

To limit the amount of tips employers can capture in this way, the Department of Labor has always restricted the amount of time tipped workers can spend doing nontipped work if the employer is paying the subminimum wage. In particular, the department has said that if an employer pays the subminimum wage, workers can spend at most 20 % of their time doing nontipped work. This is known as the 80/20 rule: employers can only claim a “tip credit”—i.e., pay tipped workers a base wage less than the regular minimum wage—if tipped staff spend no more than 20 % of their time performing nontipped functions; at least 80 % of their time must be spent in tip-receiving activities.

The protection provided by this rule is critical for tipped worker. For example, in a restaurant, the 80/20 rule prevents employers from expecting servers to spend hours washing dishes at the end of the night, or prepping ingredients for hours before the restaurant opens. Occasionally, a server might play the role of the host, seating guests when a line has formed, or filling salt and pepper shakers when dining service has ended—but such activities cannot take up more than 20 % of their time without employers paying them the full minimum wage, regardless of tips.

The Department of Labor (DOL), under the Trump administration, has proposed to do away with the 80/20 rule. Workers would be left with a toothless protection in which employers would be allowed to take a tip credit “for any amount of time that an employee performs related, nontipped duties contemporaneously with his or her tipped duties, or for a reasonable time immediately before or after performing the tipped duties” (see page 53957 of the proposed rule).

With no meaningful limit on the amount of time tipped workers may perform nontipped work, employers could capture more of workers’ tips. It is not hard to imagine how employers of tipped workers might exploit this change in the regulation.

Consider a restaurant that employs a cleaning service to clean the restaurant each night: vacuuming carpets, dusting, etc. Why continue to pay for such a service, for which the cleaning staff would need to be paid at least the federal minimum wage of $7.25 per hour, when you could simply require servers to spend an extra hour or two performing such work and only pay them the tipped minimum wage of $2.13 per hour? Or, a restaurant that currently employs three dishwashers at a time might decide they can manage the dish load with only one dedicated dishwasher if they hire a couple extra servers and require all servers to wash dishes periodically over the course of their shifts. Employers could pay servers less than the minimum wage for hours of dishwashing so long as they perform some tipped work right before or after washing dishes.

The department recognizes that workers will lose out under this change, stating that “tipped workers might lose tipped income by spending more of their time performing duties where they are not earning tips, while still receiving cash wages of less than minimum wage” (see page 53972 of the proposed rule). Tellingly, DOL did not provide an estimate of the amount that workers will lose—even though it is legally required, as a part of the rulemaking process, to assess all quantifiable costs and benefits “to the fullest extent that these can be usefully estimated” (see Cost-Benefit and Other Analysis Requirements in the Rulemaking Process).

The department claims they “lack data to quantify this potential reduction in tips.” However, EPI easily produced a reasonable estimate using a methodology that is very much in the spirit of estimates the Department of Labor regularly produces; DOL obviously could have produced an estimate. But DOL couldn’t both produce a good faith estimate and maintain the fiction that getting rid of the 80/20 rule is about something other than employers being able to capture more of workers’ tips, so they opted to ignore this legally required step in the rulemaking process.

Below we describe the methodology for our estimate. The simplicity and reasonableness of this approach underscores that by not producing an estimate, the administration appears to simply be trying to hide its anti-worker agenda by claiming to not be able to quantify results.

Methodology for estimating tips captured by employers

The remainder of this piece describes the methodology for estimating the total pay transferred from workers to employers as a result of this rule described above. To evaluate how this rule change would affect pay, we use data from the Current Population Survey (CPS), restricted to states with a tip credit (i.e., that allow employers to pay a subminimum wage to tipped workers), to estimate how much employers might shift work from traditionally nontipped to tipped staff. Doing so would allow them to spread out the total pool of tips received over more people for whom employers can pay less than the minimum wage, thereby reducing employers’ wage responsibility. We then estimate the change in total earnings that would occur for food service workers if that shift in employment took place.

The CPS is a household survey that asks workers about their base wages (exclusive of tips) and about their tips earned, if any. One problem with the CPS data, however, is that earnings from tips are combined with both overtime pay and earnings from commissions. Researchers refer to the CPS variable that provides the aggregate weekly value of these three sources of earnings (overtime, tips, and commissions) as “OTTC.” In order to isolate tips using this variable, we first restrict the sample to hourly workers in tipped occupations, to help ensure that we are not picking up workers who are likely to earn commissions.

For hourly workers in these tipped occupations who work less than or equal to 40 hours in a week, we assume that the entire amount of OTTC earnings is tips. For hourly workers in tipped occupations who work more than 40 hours, we must subtract overtime earnings. We calculate overtime earnings for these workers as 1.5 times their straight-time hourly wage times the number of hours they work beyond 40. For these workers, we assume their tipped earnings are equal to OTTC minus these overtime earnings.

Some workers in tipped occupations do not report their tips in the OTTC variable; however, the CPS also asks workers to report their total weekly earnings inclusive of tips, and their base wage exclusive of tips. For those workers in tipped occupations with no reported value in the OTTC variable, but whose total weekly earnings is greater than the sum of their base wage times the hours they worked, we assume the difference is tips.

In other words, for hourly workers in tipped occupations we calculate tips in two ways:

1. For those who report a value for OTTC:

Weekly tips = OTTC for those who work ? 40 hours per week, and

Weekly tips = OTTC ? [(base wage) × 1.5 × (hours worked ? 40)] for those who work > 40 hours per week.

2. For those who do not report a value for OTTC:

Weekly tips = Total weekly earnings inclusive of tips – (base wage x hours worker).

In cases where tips can be calculated both ways, we take the larger of the two values.

Standard economic logic dictates that employers will spread out aggregate tips over as many workers they can—thereby reducing their wage obligations and effectively “capturing” tips. They will shift work from nontipped to tipped workers until the resulting average wage (combined base wage plus tips) of their tipped workers is at or just above the hourly wage these same workers could get in a nontipped job. For employers of tipped workers to get and keep the workers they need, tipped workers must earn as much as their “outside option,” since, all else being equal (i.e., assuming no important difference in nonwage compensation and working conditions), if these workers could earn more in another job, they would quit and go to that job. But for employers to keep these workers, they do not need to earn any more than they could earn in another job (again, assuming all else is equal), since as long as they are earning what they could earn in another job, it would not be worth it to these workers to quit.

To calculate the “outside option wage,” we use regression analysis to determine the wage each worker would likely earn in a nontipped job. We regress hourly wage (including tips) on controls for age, education, gender, race, ethnicity, citizenship, marital status, and state, and use the results of that regression to predict what each tipped worker would earn in a nontipped job. We set a lower bound on predicted hourly wages at the state minimum wage. We refer to the predicted value as the outside option wage—it’s the wage a similar worker in a nontipped job earns. We assume if a worker currently earns less than or equal to their outside option wage, their earnings cannot be reduced because if their earnings are reduced, they will leave their job and take their outside option.

However, if a worker currently earns more than their outside option wage, their earnings can be reduced by the amount the worker earns above the outside option wage, since as long as their earnings are not reduced below their outside option wage, they will have no reason to leave. We also assume that if their base wage is greater than the state minimum wage—i.e. if their employer is not taking the tip credit—their earnings will not be reduced, since the 80/20 rule applies only to tipped workers who are paid a subminimum base wage. We calculate new average tips earned as the aggregate tips of all tipped workers minus the aggregate amount, just described, by which their earnings can be reduced, divided by the total number of tipped workers.

Using this estimate of new average tips earned, we can estimate how much employers might shift the composition of employment by reducing the number of nontipped workers and adding more tipped ones. We assume that the total amount of tips earned remains the same— it is just spread out over more tipped workers (who are now doing more nontipped work). In particular, we assume that the new number of tipped workers is the number that, when multiplied by the new average tips earned, is equal to the total aggregate tips before the change.

We operationalize this by multiplying the sample weights of tipped workers by total aggregate tips divided by the difference between total aggregate tips and the aggregate amount by which earnings can be reduced. We then assume that the number of tipped workers added is offset one-for-one by a reduction in the number of nontipped workers who have food service occupations. We operationalize this by multiplying the sample weights of nontipped workers by one minus the ratio of the increase in tipped workers to the original number of nontipped workers. We find that employment in nontipped food service occupations will decline by 5.3% and employment in tipped occupations will increase by 12.1%, resulting in 243,000 jobs shifting from being nontipped to being tipped as a result of this rule. The work that had been done by those nontipped workers will now be done by tipped workers, with tipped workers spending less time doing work for which they receive tips.

The loss in pay is calculated as the difference between current aggregate food service tips and new aggregate food service tips using the new employment weights just described for tipped and nontipped workers and the new average wages for tipped workers. We assume average wages for nontipped workers do not change. We estimate that there will be a transfer of $705 million from workers to employers if this rule is finalized.

Finally, it should be noted that our estimate of the transfer from workers to employers is likely a vast underestimate for three reasons. First, tips are widely known to be substantially underestimated in CPS data, thus it is highly likely that we are underestimating the amount of tips employers would capture as a result of this rule change. For example, we find that 47.6% of workers in tipped occupations do not report receiving tips. Similarly, using revenue data from the full-service restaurant industry and updating the methodology from Table 1 here to 2018, we find that tips in full-service restaurants are $30.5 billion, which is roughly twice the amount of tips reported in food service in the CPS. This means the amount employers will really capture is likely roughly twice as large as our estimate.

Second, we only estimated losses in food service. However, about 26.0 % of tips earned in the economy are not earned in restaurants or food service occupations. Combining these two factors together means what employers will really capture may be 2.5 times as large as our estimate. Third, our estimates assume that getting rid of the 80/20 rule will only have an effect if the employer is already taking a tip credit. This ignores the fact that some employers may be incentivized to start using the tip credit if the 80/20 rule is abolished, knowing that without the rule they will be able to capture more tips. Accounting for this factor would increase our estimate further.

The piece was also published at the Economic Policy Institute’s Working Economics Blog.

This article was originally published at In These Times on December 3, 2019. Reprinted with permission.

About the Author: Heidi Shierholz is Senior Economist and Director of Policy at the Economic Policy Institute. From 2014 to 2017, she served the Obama administration as chief economist at the Department of
Labor.
About the Author: David Cooper is a Senior Economic Analyst at the Economic Policy Institute.

Share this post

DoorDash changes its tip-stealing policy after outcry, this week in the war on workers

Share this post

Delivery app DoorDash decided to change its ways after getting some very bad publicity this week around its policy of taking tips that customers thought were going to workers. The way the policy went, “dashers” got a set rate for a given delivery, and if that rate was $6 and a customer tipped $4, well, the dasher still got $6 but DoorDash only had to pay $2. Now, tips entered in DoorDash will go to workers.

“The new model will ensure that Dashers’ earnings will increase by the exact amount a customer tips on every order. We’ll have specific details in the coming days,” the company’s chief executive tweeted. But the devil may be in those specific details, because Dashers aren’t convinced the company will do right by them.

On a forum for DoorDash workers on Reddit, some Dashers greeted the news with concern that DoorDash would simply pay them less to make up for the revenue it expected to lose after no longer being able to subsidize labor costs with tips.

“I’m worried that the orders will guarantee less now, but we get all the tips,” wrote a Reddit user named Dmillz648. “Meaning a previously guaranteed 10-dollar order might now only guarantee 5 bucks, and you get a 2 dollar tip, meaning you got 7 bucks for that order.”

If so many people who order delivery didn’t fail to tip, there’d be less of an issue (seriously, tip your delivery person!), but this is very much on the company as well.

Here are the tip policies of some other delivery apps.

 

This blog was originally published at Daily Kos on July 29, 2019. Reprinted with permission.

About the Author: Laura Clawson is labor editor at Daily Kos.

Share this post

Follow this Blog

Subscribe via RSS Subscribe via RSS

Or, enter your address to follow via email:

Recent Posts

Forbes Best of the Web, Summer 2004
A Forbes "Best of the Web" Blog

Archives

  • Tracking image for JustAnswer widget
  • Find an Employment Lawyer

  • Support Workplace Fairness

 
 

Find an Employment Attorney

The Workplace Fairness Attorney Directory features lawyers from across the United States who primarily represent workers in employment cases. Please note that Workplace Fairness does not operate a lawyer referral service and does not provide legal advice, and that Workplace Fairness is not responsible for any advice that you receive from anyone, attorney or non-attorney, you may contact from this site.