Identifying how collusion works promises to boost competition where it is least expected — in regional communities
Big data innovations are helping shine a light on issues affecting regional Australia, and a University of Melbourne academic is at the forefront of the transition.
Support from the Samuel and June Hordern Endowment helped Dr David Byrne undertake groundbreaking research into “tacit collusion” in Perth’s petrol market.
Tacit collusion can occur when, without explicitly talking to each other, companies raise prices in tandem. A dominant company might raise its prices, for example, and then other companies, seeing what it is doing, raise theirs accordingly.
It is different to explicit collusion, which is illegal.
Along with a colleague from the University of Sydney, Dr Byrne secured access to information on the price of petrol at every single pump in Perth from 2001 to 2015.
The result was a complete set of data, a relative rarity in the world of economics. “For years, economists have studied collusion after it’s already established,” says Dr Byrne, a senior lecturer in the Department of Economics and a director at the University’s Centre for Market Design.
Having super rich micro data on prices that researchers haven’t previously had access to allows us to study the evolution of pricing behaviour over time.
“Perth is the context for this, but these data sets are now emerging in retail markets around the world, especially regional ones.”
Dr Byrne’s research found evidence of “tacit collusion” among petrol companies in raising prices. He does not suggest that any explicit collusion took place. But it didn’t have to – companies could see what dominant market players were doing, and adjusted their prices accordingly.
Dr Byrne found that the companies doubled their profit margins between 2010 and 2015, from 5 per cent to 10 per cent, costing a typical petrol consumer around $150 to $200 per year.
“It shows how companies, without phone calls, are able to co‑ordinate,” he notes.
Thanks to the emergence of big data sets like this one, researchers are beginning to paint detailed pictures of how regional markets work. This is something of a breakthrough in economics, where data has often been concentrated where people are – in cities.
“It’s like we have a new microscope allowing us to gain insights into what regional market competition looks like,” he says enthusiastically.
Dr Byrne is currently partnering with the NSW government on a research project analysing petrol data in all urban and regional markets across the state. The research makes use of the state’s FuelCheck website, which was designed to give consumers up‑to‑the‑minute information on petrol prices at every pump in NSW.
Dr Byrne says that high petrol prices are a major problem in regional markets, where there’s less competition than in the cities.
“There’s very good reason to believe, based on research I’ve done in Canada on petrol, that we will potentially find even larger issues in rural places, where markets are even more concentrated.”
The rise of big data might be ushering in research breakthroughs, but it’s also creating all sorts of opportunities for collusion-like activities. Information that’s available to researchers and consumers is also available to companies.
Dr Byrne notes that, in a legal context, collusion can currently only be proved if it can be established that firms communicated with each other.
“The definition in the law is that the medium for communication is words. Phone calls, emails … that sort of stuff.”
So, what happens when companies start using big data and algorithms to change their prices automatically based on those of their competitors? We could very well see “battling algorithms”, operated by companies and regulators, duking it out to adjust prices or detect collusive behaviour. And while that might play out on a global stage, the impact on regional Australia would be profound.
We already worry in rural markets that tacit collusion exists because there’s so few petrol stations who all watch each other.
“This would only exacerbate some of these problems and bring them to even larger markets.”
Established in 2011 by the Aranday, Yulgilbar and Myer Foundations, the Samuel and June Hordern Endowment supports activities that increase linkages between the Faculty of Business and Economics and rural Australia.
Funding from the endowment helped Dr Byrne’s research in two very different ways. Crucially, it helped the researchers promote their findings more broadly, and talk directly to governments. They presented workshops on their case study to regulators far and wide: from the Australian Competition and Consumer Commission to the US Federal Trade Commission in Washington DC.
Thanks to this profile building, the case study was featured at a 2017 OECD conference for global antitrust agencies.
“If you just wait for governments to find the research they’ll eventually find it, but it could take years. We just went out and got things rolling. [We are now] informing an ongoing debate around big data and algorithmic collusion.”
The funding also helped the researchers buy “auxiliary data” to firm up their findings by contrasting the Perth data with other markets.
“This funding is contributing to solving a massive societal challenge that will affect not just Australians but people around the world,” Dr Byrne says. He both recognises and relishes the challenge algorithmic collusion poses to researchers and regulators.
It’s a hard problem, but universities hit their stride when people are working on the hard problems that societies are facing.