GENERAL ARTICLES

He’s Ready To Take It

Who’s ready to take it? The man in the street. Take what? Stiffer war sacrifices. Who says so? The Gallup poll. How does it work? This article tells

WILFRID SANDERS January 1 1943
GENERAL ARTICLES

He’s Ready To Take It

Who’s ready to take it? The man in the street. Take what? Stiffer war sacrifices. Who says so? The Gallup poll. How does it work? This article tells

WILFRID SANDERS January 1 1943

He’s Ready To Take It

GENERAL ARTICLES

WILFRID SANDERS

SOMEBODY has been quoted as saying again that the public needs to be awakened; that it is apathetic to the war, and that it was “about time Canadians learned they have to make sacrifices.”

This is not the first time this statement has been made, nor, probably, is it the last.

But I say it’s spinach, and I hold a brief.

This brief is the record of the Gallup poll’s first year of operations in this country. And in my opinion it shows that Canadians during the past year have been more than prepared to accept all the sacrifices demanded of them.

We’ll look at the record in a moment, but first let’s take a look at the poll itself and see how it goes about this job of taking the nation’s pulse.

Let’s start with the case of Bill Gilhooley (the name’s a fake but the incident isn’t) who wheels a freight engine out of Ottawa to points east.

One day recently Bill had put his engine into the yards and was heading for a wash-up when a young man stepped up to him and said: “I’m a reporter for the Gallup Poll. I’d like to ask your opinion on a few questions.”

“Why?” asked Bill Gilhooley, asking the first question, which wasn’t cricket.

“Well,” said the reporter,“this is part of a nationwide survey to find out what people think about the important issues of the day.”

“Why?” said Bill, who had never played cricket anyway.

“So that the leaders of the country and others may know the will of the people.”

“Shoot, son,” said Bill at last. “If Mackenzie King wants to know what 1 think, he can have it. It’s more than my wife wants to know.”

The reporter wet his pencil, and started:

“Do you think people in Canada should be allowed to join the Communist party and enter candidates in future elections, or do you think the present law which outlaws the Communist party should continue in effect?”

Bill said he thought the present law was okay but added that he didn’t want the Russians to think from this that he, Bill Gilhooley, didn’t have a high regard for the fight they have put up against the Germans. But somehow, this was Canada, wasn’t it?

The reporter put a check mark under “Continue

present law” on his ballot, asked a few more questions, and went his way.

Bill went home, and told his wife about it. She said she thought he’d answered the questions “just right,” but was glad that the reporter had “forgotten” to ask Bill his name. The reporter, of course, wasn’t supposed to.

In the meantime the reporter had mailed a bundle of ballots to Toronto, including that containing Bill’s views on this and that. In the same Toronto mail were ballots from more than 200 towns, cities, villages, and farm communities across Canada, from coast to coast, all solemnly recording the views of the Bills, Toms, Dicks, and Harrys about the Communist party. In Toronto, girls in smocks leafed through all the ballots, punching keys in tabulating machines, which in turn punched slots in cards. Then these cards were put through a sorting machine, and, a few days later, twenty-six Canadian newspapers from coast to coast told their million or so readers that sixty-two per cent of the Canadian voters wanted the present law against the Communist party kept in effect; that twentythree per cent wanted the ban lifted and that fifteen per cent didn’t know whether it should be lifted or kept in effect.

In there somewhere was Bill’s opinion—an infinitesimal part of a fraction of one per cent, but there it was.

The Poll At Work

THIS was the Gallup poll, or, to give it its full title, the Canadian Institute of Public Opinion, at work.

At about the same time, seven thousand miles away, the Bill Gilhooleys of /. u:tralia were being asked their opinions on national and international issues. In Britain the same thing was going on; and in Sweden a Gallup reporter was making marks on a ballot headed, “Svenska Gallup Institutet.” In Oklahoma City a woman was giving her opinions to a Gallup reporter as she fed her laundry through the wringer.

So much for the mechanics. Now let’s look at the record of the Canadian’s attitude toward sacrifices demanded by the war.

Last March Gallup reporters across Canada went out and asked the Bill Gilhooleys and representatives of other groups this question:

“Suppose it was your job to decide income taxes for the coming year. How much would you have a man earning $1,500 a year ($30 per week) pay in Dominion income taxes next year?”

The question was repeated for various income groups, from $1,500 to $25,000 a year. It was asked nearly five months before the last budget substantially increased the tax burden on the average citizen. If the public “needed to be awakened to the necessity of making sacrifices” at that time, one would think its apathy would have been reflected in the answers received to the question above. The

Who’s ready to take it? The man in the street. Take what? Stiffer war sacrifices. Who says so? The Gallup poll. How does it work? This article tells

table below, however, tells a different story. It compares the tax the public would levy (based on the mathematical average of all suggested tax schedules) with the tax the citizens were actually paying at that time. It should be remembered that at that time, according to official estimates, five times as many Canadians were paying five times as much income tax of all kinds, as before the war. Moreover, at the time the question was asked, hundreds of thousands of ordinary citizens were busy making out their returns for the March 31 dead line, and income tax matters were accordingly very much in the forefront of the average man’s mind.

Here is the table: Tax Which Tax As Set Family of Public By 4 Earning: Would Set Government $1,500 $23 Nothing $2,500 $191 $30 $5,000 $875 $525 $25,000 $8,200 $9,500

Notice an interesting thing about these results. On the average or lower incomes, the public’s tax schedules were well above the actual levels at that time. When it comes to a man earning $25,000, a big income in Canadian terms, the public set a tax below that actually paid. So that the table not only gives some indication of how far ahead of the Government the public was in setting the burdens of war, but it also indicates that apparently the public had no idea of what taxation the wealthy man was paying, and thus set the tax below the actual figure. Maybe there is something here for those who profess to detect a “soak-the-rich” spirit abroad to ponder over.

At this point some cynic might rise to say: “Well, it’s onethingto ask a man what taxes he would levy, in theory, to help win the war, and another to get his reaction if he actually had to pay it.”

Question Carefully Worded

IN THE first place, the wording of the question made no direct appeal to patriotic or warwinning motives. In the second place, it just so happened that a few months after this poll, a new tax schedule was announced from Ottawa, greatly increasing the tax burdens on our average citizens. This extra burden made its first impact on the national pay envelope after Sept. 1 last. After the Bill Gilhooleys of the country had had a month to feel this new drain on their income, the pollsters went out to see how the public was adjusting itself, and found that about eighty per cent of the taxpayers were taking the new taxes somehow in their stride, with no weeping, no wailing and only a minimum of teeth gnashing. Of the remaining twenty per cent, twelve per cent thought they were down to bare necessities even before the new taxes, and eight per cent had not yet got down to brass tacks by way of making tangible plans.

Thus, before the 1942 budget, the majority of persons were in favor of higher income taxes on the middle and lower income brackets, and after the higher levies had been put into effect, the vast majority of the people were adjusting themselves without complaint.

Another example. Early in October this question was put to the Canadian people:

“At present the Government’s policy is to try Continued on page 29

Editor, Canadian Institute of Public Opinion

Continued from page 12

to pay about three quarters of the war’s cost out of its present income. Do you think this is a good policy, or do you think more of this expense should be left to be paid after the war is over?”

Since the higher taxes had been collected since Sept. 1, and this question was asked in the first half of October, nobody can fairly claim that the public did not realize what the new levies would mean.

Here were the results obtained from this question:

Approve present policy.. . 67%

Pay more after the war .. 20%

Undecided.............. 13%

If those who were undecided were left out, this would mean that seventy-seven per cent of those with an opinion, or over three out of every four persons, want to pay for as much of the war now as possible.

Nothing here to indicate unwillingness to sacrifice.

This same public, this man on the street, this little guy whom critics have accused of apathy, voted through the poll by a majority of fifty-four per cent for a compulsory savings scheme, months before it was law, and was willing to have ten cents deducted from every dollar earned for this purpose, without interest.

When the first hint of a sugar shortage was abroad nearly seven out of every ten Canadians interviewed said, again through the poll, that he or she thought the best thing

to do was to ration sugar. Some months later rationing came.

When the rubber shortage became page one news we heard quite a few stories about hoarding of tires; about how people were still driving their cars to work, with only one person in the car; about people who went on long and unessential pleasure trips by car. Such stories were eagerly told, and eagerly listened to, gathering color and momentum as they went. But just about then the Bill Gilhooleys and the rest of the country, all classes, were being asked this one:

“If the rubber shortage became very serious do you think the Government should take the tires from automobiles that are not necessary to the war effort or for making a living?”

Some eighty-four per cent of the people at large, as measured through the poll’s cross section, said yes, the Government should take such tires. Eight per cent said they should not, and the other eight per cent were undecided.

“Sure,” says our cynic. “Lots of people—most people—don’t drive cars. They should worry !”

Then let’s separate the car drivers from the non-car drivers, and see what the answers look like:

Car Non-car Owners Owners

Should take tires 84% 84%

Should not take

tires.......... 10% 7%

Undecided....... 6% 9%

To a further question, nearly all

those who were in favor of the proposal said they thought the Government should pay the owner a fair price for the tires, and not confiscate them without remuneration.

Impatient For Action

SLIGHTLY over a year ago these average Canadians were asked what they thought was the most important problem facing the Government at that time. The greatest number (thirty-five per cent) said conscription. Next largest number (twenty-three per cent) said increasing war production, while financial matters, such as avoiding inflation, taxation, and so on, were, as a group, third on the list.

Again, last March, the matter of hours of work in war plants was put to the people. “How many hours,” they were asked, “do you think workers in war industries should work each week?”

To this question fifty-seven per cent of the voters interviewed answered in terms of over forty-eight hours per week. Average number of hours selected was fifty-four hours per week. Y et a great many, perhaps the majority, of war plants operate on an eight-hour day, or forty-eighthour week.

As for strikes in war plants, seventy-eight per cent wanted these completely and effectively outlawed for the duration.

The public, far from being apathetic, is actually impatient for faction. Most of them felt that Parlia‘‘“ment should not have adjourned when it did. Moreover, just before the adjournment, the Gallup poll found that more people thought the then current session of Parliament had “wasted too much time in political talk,” than thought “it had done a good job.”

In May last forty-nine per cent of the electorate thought quarrelling among political parties had interfered with Canada’s war effort; thirty-four per cent thought it had not, and eighteen per cent were undecided.

Another indication of the place the public gives the war on its list of “things to be done” is supplied by the overwhelming vote against holding provincial and, to a lesser extent, federal elections in wartime. In a survey made last October the poll found seventy-one per cent of the electorate with an opinion on the matter, in favor of postponing provincial elections, and sixty-three per cent in favor of postponing federal elections in wartime. Before the April plebiscite, a majority of the voters disapproved of the Government’s proposal to hold one on the grounds that it was an unnecessary expense.

It must, of course, be admitted that the strange phenomenon of a public which is fighting a grim war for democracy, being willing to jettison a prime democratic function such as an election, for the duration, may be interpreted in several ways— as a grim commentary on the health of democracy, or as further evidence of the public to sacrifice certain standards of living to the end of winning the war. However, there it is. Last August the Institute re-

ported that more than six out of every ten adults in Canada’s Pacific province felt an air raid on the Pacific coast was possible before the end of the summer. Their fears were not justified, but nobody can say that this was apathy.

A month earlier the poll reported that over a third of the population in areas most likely to be bombed, if bombing there was (that is, cities of 10,000 population and over) were not satisfied with their local air-raid precautions.

The above is just a sample of how the Canadian public view domestic issues. On international affairs their eagerness for action is just as evident. They are in no mood for pussyfooting with Germany, and when asked what they would do with Germany after the war, thirty-eight per cent of the answers could be grouped under the heading of “end Germany as a nation.” Another twenty-five per cent wanted her put in a strait jacket. This poll, of course, was taken in the heat of war, and here again, I’m not arguing that the public is right in its present attitude. This instance, like the others mentioned, is cited merely to illustrate our major point: that Canadians do not need to be “awakened” or to have the war brought home to them.

On what grounds does the Gallup poll, on the basis of asking a few thousand Bill Gilhooleys, report the results cited above?

People And Percentages

SUPPOSE you were at the exit of a big city hockey arena as the crowd was coming out, and you wanted to find out what the proporportion of women to men at the game was. As the people pass through the exit you count the number of men and women. Of the first 300 you find that 225 were men and seventy-five women, or, to say the same thing in another way, that seventy-five per cent of your sample were men, and twenty-five per cent women.

It is a mathematical fact that if you assumed, from your count of the 300, that the whole stadium contained seventy-five per cent men and twenty-five per cent women, you would be within seven per cent of the actual figure—in other words, the arena contained somewhere between sixty-eight and eighty-two per cent men, and somewhere between eighteen and thirty-two per cent women.

But perhaps this isn’t accurate enough for your purpose, so you count another 250 persons, find you have counted 429 men, and 121 women, and that your total sample was therefore seventy-eight per cent men and twenty-two per cent women. You can now be sure (and again this is a mathematical fact) that you have the percentage of men and women at the game right within five per cent. If you want to whittle down your margin of error still farther you can count still more.

There is one other thing, however, you must watch out for. Perhaps the Y.W.C.A. or some other women’s organization occupied a block of seats near the exit. That would mean the first 550 persons to leave the arena would contain a greater percentage

of women than prevailed over the whole arena and that your cross section was not accurate. That is something you would have to check at the box office before you started your count.

Now what has this got to do with public opinion measurement? Suppose, that instead of finding out the number of men and women at the game, you wanted to find the percentage of fans who thought the referee was right in penalizing that offside, and the percentage who thought he was wrong. You would go through exactly the same procedure as before, except that instead of counting the men and women, you would ask the first 550 persons to express their opinion. Your margin of error would be the same as it was in the first case.

But here again there is a possible error to guard against. Supporters of the one team might all have been sitting near the exit, in which case you would not get a fair cross section of opinion throughout the whole stadium by questioning the first 550 persons.

To get the opinion of the whole nation the poll uses the same technique as we used at the hockey stadium—except that it makes sure it gets a fair cross section. Obviously, if we get the opinion of too many farmers, or too many Liberals, or too many wealthy persons, we will not be getting the opinion of the nation as a whole. So the poll makes sure that all types and groups of citizens are represented in its cross section in the same proportion as they are represented in the whole population. What it has, then, is a miniature of the whole population.

The poll tries to keep its margin of error down to four per cent. Since elections provide the best test of the poll’s accuracy, let’s look at the record. The 1942 Congressional election in the United States was the 114th election in which thet Gallup poll made a forecast. The average error in those 114 forecasts was between three per cent and four per cent. The first prediction in 1936 was out six per cent. The forecast of the result of the Dewey-Bennett fight for the governorship of New York State last November was out one half of one per cent. However, when your error is less than four per cent you’re lucky.

The biggest problem facing the Institute at election time remains the question of turnout. This was demonstrated in the U.S. Congressional elections last fall, when less than half the voting population of the United States went to the polls. This was, in proportion to population, the lightest vote since the last war. It has been well established that the majority of the stay-at-homes were Democrats. This resulted in the poll’s underestimating the Republican gain by just over the four per cent margin striven for. But the poll’s estimate of which of the 435 Congressional seats at stake would go Republican, and which Democratic, was not based on surveys in each of the 435 Congressional districts, but entirely on the forecast of the national strength of the two parties. As a result, the slightly abnormal error in forecasting the

vote, threw the estimate out of seats by thirty-one in 435, as compared with an error of six seats in the 1938 election. To ensure against such a situation it would have been necessary to have made 435 separate surveys—one in each Congressional district.

What this example boils down to is that, it is one thing to determine how the whole adult population feels about a certain issue, but another thing entirely to predict how many of them will take the trouble to implement that opinion at the polling booth. The technique of judging turnout is being closely studied, and is being steadily improved.

The Institute does not consider the forecasting of elections its most important job. This job, it believes, is the accurate reporting at frequent intervals of public opinion on issues

of the day, between elections. Elections, however, do serve as a check on the accuracy of the poll’s methods.

Another question which requires continuous care is the phrasing of questions, so as to eliminate, as far as is humanly possible, misunderstanding or bias. Every question the poll puts out is gone over carefully word by word, by experts, and then tested by actual interviews before it is put out on a national scale. This is typical of the case necessary to get a proper cross section of the nation’s opinion.

Of course, Canada has its factions, its minorities, its isolationists and racial bigots. But viewed as a whole the nation can be, and is proving it can be, capable of a cohesive, definite public opinion.

And it is a broad, aggressive stream.