Thursday, May 3, 2018


The recent surge in fake news has created the demand for government regulations, mostly out of concern that it influenced the last U.S. presidential elections. Regulations would be warranted only if fake news actually affected the outcome and if it could also be inexpensively identified and eliminated.
Fake news very likely had no effect on the election outcome under the reasonable assumptions that both major parties used it to the same degree and their targeted audiences were equally likely to evaluate it properly. The amount and success of fake news originating in Russia is not known but its existence should prompt policies aimed at it alone, not all U.S. communications. That is because the cost of discovering and eliminating fake news is very high, requiring large and costly banks of computers and sophisticated algorithms. The data in the computers conjure concerns over privacy. Importantly, the interpretation of the algorithms risks that the individuals doing the work will let personal ideological and political preferences influence it.
I have had some personal experiences that illustrate this risk. In 1975, I presented a lecture on the future of the international monetary system to the graduating class of the Diplomatic Academy of Chile in Santiago. To my surprise, I found no evidence of the public protests against the Pinochet regime that I had expected after seeing them regularly on CBC.
To understand the difference between the TV news coverage and reality, I asked Canada’s ambassador to Chile, who said other Canadian visitors made the same observation. He then told me that a few days earlier his people had learned that a CBC TV crew had landed in Santiago without following the normal protocol of informing him of its arrival. He discovered that the TV crew had a pre-arranged meeting with a Chilean group in a poor part of the city. The Chileans arrived armed with anti-Pinochet placards. They waved these placards and chanted slogans while the CBC crew filmed them from an angle that suggested a larger crowd. He said the crew paid the villagers who then stored the placards for later use and dispersed. The day after the performance, Canadians saw on TV a supposedly big demonstration in Santiago protesting the country’s president.
In the 1980s, I saw fake news about South Africa, where I taught economics for a semester at the University of Cape Town. There I befriended a Dutch doctor and his wife. They were interns at the famous Groote Schuur hospital, and their parents in the Netherlands were worried that they might get caught in the deadly riots that Dutch newspapers regularly described. One of their letters included a newspaper picture of a scene allegedly showing the aftermath of such a riot. The scene was of the hospital’s courtyard. The same, landscaped yard the couple saw each day and where staff often ate their lunches and took naps on the grass. It was no riot scene. Something had been faked.
These two episodes illustrate the extent to which supposedly objective media can and do produce fake news slanted by the political views of their managers and employees. We should expect the same from any proposed agencies charged with the objective task of suppressing fake news. For these reasons, it seems better to prevent the likely small cost of fake news by relying on the common sense of the common people rather than rely on the decisions of a few employees of government agencies.
Herbert Grubel is emeritus professor of economics, Simon Fraser University.

The article was published in the Financial Post on May 2, 2018