If you have been a soldier for some length of time perhaps you feel confident that you know what “we believe” or what “our regiment believes” on certain subjects. Doubtless you’ve talked to hundreds of men representing all shades of opinion within your group. That is why you believe you know what “we” think. On a few questions you may even be ready to say what the soldiers in “this man’s army” believe. You may not hit the answer right on the nose, but if you have heard enough comment from fellow soldiers, you’re probably pretty certain that you’ve caught the drift of their opinion.

Men have been trying for centuries to find ways of gauging public opinion. American history is peppered with incidents in which somebody guessed wrong about what people were thinking—all the way from George III who thought our ancestors would take a Stamp Tax and like it down to a man named Adolf Hitler who had us sized up as a people too soft to go to war. Yet methods of measuring public opinion are neither magical nor mysterious. They never have been. It’s an age-old idea that the best way to find out what people are thinking is to ask questions of them.

When the local sage sitting on a cracker barrel at the general store tells you his community is in favor of a new roof for the schoolhouse, does he know what he is talking about? Probably he has heard and noted the views on the subject expressed by the stream of local citizens who have dropped in at the store for weeks past. It’s an old-fashioned way of opinion polling, but it can be pretty sound.

Likewise, when your Congressman returns to Washington and announces that he will support the majority view in his home district on the new tax bill, is he just uttering words or does he have something to go on? More than likely he bases his remarks on talks with a large number of people back home. He may not be able to say what the majority view on any particular bill is within a few percentage points, but he believes that he has discovered the views of the majority or of those whose support is most important to him.

Nothing is more unjust or capricious than public opinion. William Hazlitt

Similarly, political reporters of big city newspapers have long made a practice of touring the country before elections in an effort to find out the drift of opinion. Afterward they base their reports to the public on a large number of chats and interviews with persons in various social and economic classes. Often their predictions prove correct. Sometimes they fall wide of the mark, for the chances of error in their method of sampling the population are great. They cannot be sure that the persons to whom they talk really represent the rank and file of the voters.

How accurate was the old way?

Such rule-of-thumb methods of finding out what the public thinks about its problems allow almost anyone to set himself up as an expert on public opinion. They give anyone with an ax to grind a chance to claim public support. The announcements often made by representatives of special interests, for instance, that “business believes” thus and so, that “labor feels,” that “the farmers demand,” or even that the “American people insist,” may possibly be based on a poll of some sort. Or they may actually be unsupported statements which greatly exaggerate or twist the real opinion of the public. For unless we are told how and on what basis such conclusions were reached, we have no way of testing how reliable these sweeping statements are. And unless reasonably accurate means are used to find out what the public’s opinion on current issues is, even the sincere and honest forecaster of opinion is likely to make serious mistakes in his predictions.

The proverbial wisdom of the populace in the streets, on the roads, and in the markets, instructs the ear of him, who studies man more fully than a thousand rules ostentatiously arranged. Anonymous, 1804

Equally tricky are attempts to predict the trend of public opinion on the basis of the public’s views in years gone by. Public opinion is not fixed or static. It changes with the times, and big shifts in opinion may result from new situations or the effect of recent events. A trend of opinion throughout several generations may rapidly give way to new forces. Pearl Harbor blacked out isolationism in a few hours.

From EM 4: Are Opinion Polls Useful? (1946)