These Employers Pay Higher Salaries than Necessary

Some employers using online freelance marketplaces for the first time pay more than they have to for workers. Why? An information imbalance that job seekers can exploit, as explained in research by professor Christopher T. Stanton.
by Michael Blanding

Imagine you walk into a shop where you don’t know the prices. Maybe it’s a Turkish souk, redolent with smells of saffron and turmeric. Or maybe it’s a New Hampshire antique store, full of dusty shelves of enticing oddities. You pick out your silk scarf or ceramic tchotchke, and bargain the best price you can.

As you walk out of the shop, you can’t shake the feeling that you just got fleeced—forced to pay just slightly more than a local would pay. The fact is, you are probably right. No matter how vigilant you are, it’s hard to get a fair deal when the seller knows you are an inexperienced buyer. Now translate the same situation from a brick-and-mortar bazaar to an online peer-to-peer marketplace such as Upwork, which allows businesses to outsource work to freelancers, frequently overseas.

“Workers on the market have lots of feedback on their past jobs, and can also see how much experience the employer has on the market”

In this case, the employer is the shopper, and the freelancer has the information advantage—especially when working with a first-time employer. The result: The inexperienced employer tends to pay a higher hourly wage than what employers more experienced with the system negotiate.

“Workers on the market have lots of feedback on their past jobs, and can also see how much experience the employer has on the market,” says Christopher T. Stanton, an assistant professor in the Entrepreneurial Management unit at Harvard Business School. Employers, meanwhile, are comparatively in the dark, creating an “information friction” that the freelancers can exploit.

In a March 2015 working paper, Information Frictions and Observable Experience, Stanton, along with Catherine Thomas (PhDBE 2006) of the London School of Economics, does what a shopper could never do at the bazaar—measure the exact difference between the price paid by a first-timer and the price paid by an experienced haggler.

Stanton and Thomas obtained the complete database on all administrative support jobs between 2006 and 2010 on freelance platform oDesk (which merged with its biggest competitor, Elance, to become Upwork in 2015). After crunching the numbers, they found that the average freelancer was hired for an hourly wage of $3.56 by employers who had made at least five previous hires, $4.41 by employers who had made only one previous hire, and $5.01 by employers who were hiring for the first time.

Employers using a job site for the first time may pay
more than necessary for a candidate. ©iStock/Mufti

(Most of the workers on the site are from India or the Philippines, where minimum wages are around $1.50 an hour. These wages for administrative work are significantly higher than what might be implied by average per capita GDP.)

After controlling for workers’ job experience, English proficiency, and other factors, first-time employers still paid a premium of 10 percent over experienced employers—a significant leap for a job that might take 100 hours.

Freelancers aren’t necessarily charging higher prices to take advantage of inexperienced companies, says Stanton. They could also be trying to recoup their higher costs in working with a first-time employer, who might need more hand-holding navigating the platform and for which the worker might not be compensated. Using a demand model, Stanton and Thomas estimate that hand-holding premium accounts for about two-thirds of the price increase for first-time users, while markups represent the other third.

Charging any premium, however, counters the common marketing logic of charging first-time customers lower prices in order to get them in the door, and then raising prices later. Think of visiting a restaurant you might not otherwise have visited because you have a coupon for a half-price meal.

“The difference in a peer-to-peer market is that even though the platform’s objective is to get more people in, individual workers aren’t likely to capture that future return,” says Stanton. Since workers are bidding out to jobs individually rather than developing a relationship with a particular employer over time, he adds, “they don’t have the incentive to internalize those future returns by providing a discount.”


To account for this first-timer penalty, Stanton recommends that platforms like Upwork reduce fees for inexperienced employers. Presently, the company charges a flat 10 percent fee on all transactions. Cutting the fee for first-timers would essentially bring costs in-line with those of experienced users; while the platform would make less money on those individual transactions, it would in effect be providing a discount that could entice more users to try out the platform, and increase overall profits through repeat customers.

For employers using these platforms, meanwhile, there isn’t much they can do to avoid paying higher hourly wages on the first few hires. Until they get a sense of the variation in workers’ qualifications and a better feeling for finding the cheapest worker that can do a good job, they simply must absorb the higher wages.

Stanton advises novice employers to cut losses by dipping a toe in the water first.

“They should start with a small task, and build up experience before trying to do a more advanced project,” he says. “It familiarizes them with the market and the set of people that are likely to apply, so they have a good sense for which applicants are a good match.”

In other words, fake it until you make it. After a few hires, freelancers will see which employers have had some experience on the site, and are less likely to charge them a premium.

Now, if only there was a similar way to drive a harder bargain at the souk.

About the Author

Michael Blanding is a writer based in Brookline, Massachusetts

Post A Comment

In order to be published, comments must be on-topic and civil in tone, with no name calling or personal attacks. Your comment may be edited for clarity and length.