In the News
Labour markets: How useful are ratings in the Gig Economy?
Using any of the gig economy platforms gives you the chance to rate the person you have hired. This means that they can build up a reputation for their work, be it as a driver, a plumber or a provider of accommodation. The idea of the ratings system is to build up trust between strangers, and so enhance the chance of the workers' being picked by future users of their service. A thought-provoking article in the FT questions this system.
The companies at the heart of the gig economy like to talk about its fluidity. Workers flit from one task to another, using platforms like Uber and Upwork to find work.
The issue is that, as they make that flit, they leave their ratings behind them, and have to start to build their reputation again. Research suggests that in some cases this is not very important - when we pick a driver on Uber or on Deliveroo we base our choice more on timing and convenience than on their rating.
However, if we are picking a plumber or builder, the case is different and we are far more influenced by ratings and comments made by previous service users. So there is a suggestion, supported by unions as well as academics, that workers should be given ownership of their online reputations so they can transfer them to new platforms.
There is another interesting aspect to this: how reliable are those ratings that customers give? Behaviourally, we are reluctant to publicly give low ratings, even when we are not very happy with the service we receive.
The FT refers to some recent research under the title 'Reputation Inflation' which highlights how the distribution of feedback scores in various online marketplaces seems implausible. For example, "...the median seller on eBay has a score of 100% positive feedback ratings, and the tenth percentile is 98.21% positive feedback ratings. On Uber and Lyft, it is widely known that anything less than 5 stars is considered “bad” feedback: Athey et al. (2018) find that nearly 90% of UberX Chicago trips in early 2017 had a perfect 5 star rating."
One platform conducted some research, asking hirers to rate workers privately as well as publicly — they were promised the private ratings would not be given to the worker or to other would-be hirers. The difference was remarkable. Almost 30 per cent of hirers who privately said they definitely would not use the same worker again still rated them publicly at 4 or more stars out of 5.
The conclusion is that hirers felt there was a cost to giving publicly bad feedback — either they feared it would rebound on them in some way (perhaps in a poor rating of them as a customer), or they simply did not want to harm the worker.
Surely this makes the system worthless?