On Aug. 25 the front page of The Wall Street Journal got my attention. The headline spread over six columns: “Markets Reel in Global Selloff — Wild Ride Leaves Investors Gasping.” On the left a sub headline warned, “Risks Rise in U.S. As World Staggers.”
The message was clear: This event is important. You need to read about it. And so I did, following the headlines into the paper for two-and-a-half full pages of news and commentary about the market decline and what it meant, and at least three more full pages in the Money & Investing section, with an extended discussion of how the volatility was affecting investors and businesses.
I learned that a couple in Oregon called their broker to liquidate their stocks and that investors were piling into U.S. Treasuries as they scrambled to protect their money.
Five days later everything changed. The lead article on page one read, “The big question worrying investors now is whether last week’s rally is sustainable or just the prelude to another storm.”
I can’t count how many similar articles I have read like this over the years — perhaps hundreds. The format goes something like this: The market has moved sharply/gently down/up; investors are concerned about whether the market will continue/reverse; we interviewed three experts who will briefly summarize their optimistic/pessimistic outlooks.
The time periods change, the reporters change, the experts change, but the message is always the same: “What is happening in the market now and may happen in the very near future is significant; there is no agreement about what it all means, but you need to be concerned.”
The influence of the media on our thinking is well documented. Companies wouldn’t spend trillions of dollars on advertising if it didn’t influence consumers to buy their products, and newspapers wouldn’t print stories if subscribers didn’t read them. Repetitive messages in the media are reinforced by social interaction (“How was the market today?”) and our own inherent biases (finding patterns where they don’t exist). They shape our thinking and our behavior.
Just as the media has gone through a radical transformation, so has its influence. That influence is now being delivered without our knowledge, through the most benign and utilitarian format: the search engine. The search engine is the Yellow Pages, the White Pages, Encyclopedia Britannica, Whole Earth Catalog, Library of Congress and a whole lot more rolled into one convenient location. It is the most popular, most used and perhaps most valuable tool on the Internet. It is how we find stuff. How could that be dangerous?
The first inkling that a search engine might not be as neutral a tool as we imagined came with the emergence of its most successful exploiter: Google.
Google’s managers stumbled upon the insight that the connection between searches and the searcher had enormous commercial value — leading to their domination of an entirely new form of advertising. They gave us access to information, and we gave them access to what we were looking for. Our privacy was compromised, perhaps permanently — but that is nothing compared to how our growing reliance on search results is changing our perception of reality.
If researchers Robert Epstein and Ronald E. Robertson are correct, the rankings of search results that we see after we type a question are influencing our attitudes, preferences and behavior — so much so that elections can be manipulated by changing the search results — and they go to great lengths to show how easy it is.
Their findings were detailed in a recent report about what they call “The Search Engine Manipulation Effect,” published in the Proceedings of the National Academy of Sciences. They wrote: “Biased search ranking can shift the voting preferences of undecided voters by 20 percent or more.”
Given the billions of dollars spent by candidates to sway far smaller percentages of votes, this effect, if replicated by other studies is nothing short of enormous. Few of us, certainly not I, thought very much about how the results of our Google searches were influencing our thinking and conclusions.
The study’s authors emphasize that there is no evidence to suggest that Google or other search engine providers are knowingly altering our search results to manipulate elections (or anything else). But they write: “Because the majority of people in most democracies use a search engine provided by just one company, if that company chose to manipulate rankings to favor particular candidates or parties, opponents would have no way to counteract those manipulations.” This is scary stuff.
Epstein and Robertson appear to have identified a totally new form of media influence. Since Google’s founders didn’t predict or plan for the scenario where search changed the face of advertising, it is likely that the unprecedented influence search results appear to have over our perceptions and beliefs are unintended consequences of otherwise good intentions. Perhaps, but it doesn’t make the problem any less significant.
In retrospect, of course, it seems so obvious. In the years before Internet search existed, there was a monster of a book called the Reader’s Guide to Periodical Literature. If we wanted to research any subject that might be referenced in a magazine or journal, this was the source. The process was time-consuming, and by today’s standards, incredibly inefficient. But that inefficiency forced us to broaden our scope in the search for information. The only thing restricting the end result was time and the mental energy we had for the project.
Google search results are so quick and consistently good we have learned to trust them. Epstein notes that the result of that trust leads 90 percent of us to not go past the first page of results; and 50 percent of the time “our clicks go to the first two items.” Speaking for myself, it is so simple and convenient to get information from a Google search, there’s not a lot of motivation to question what’s behind it all.
There are dangerous implications from that choice: First, our excessive trust in search results means that we don’t think too much about what we are not seeing; and second, all of our searches are mediated by an algorithm whose variables are unknown and invisible to us — which means we are allowing our search for important information to be directed more by Google and its algorithm than by our own independent efforts.
Markets are efficient pricing mechanisms because the majority of participants are making their choices independently. Diversity of input is what creates a “smart crowd.” When that diversity breaks down; when too many participants make their choices from a limited number of shared sources, the smart crowd starts behaving like a mob and the pricing mechanism loses its efficiency.
Regardless of Google’s intentions, with nearly 70 percent of all searches going through their algorithm, the end result is that our individual and collective pursuit of knowledge is being influenced by one for-profit company’s version of “relevant.”
Am I going to stop using Google for search? Not likely. And neither are you, or the millions of others who have benefitted from this powerful and convenient tool. But when there are important decisions to be made, the only one who should decide what is relevant and what is not is you, not Google.
Research without relying solely on Google Search takes longer, is less focused, less convenient, even a little messy. But the process takes us ultimately to a place where we can perceive the world in a way that would not be possible without the effort. It all begins with a simple mindset: The media exist to serve us, not vice-versa.