Reuters Poll Finds More People Support Apple in Fight Against FBI

| Analysis

Reuters released Wednesday the results of a poll that found more people—just less than half—supported Apple in its decision to fight a government order to create a backdoor in iOS. Those results contrast sharply to a Pew poll released earlier in the week that found a majority of people wanted Apple to do the government's bidding. The differences in those results come down in part to the way the questions were phrased.

Reuters/Ipsos Poll

First off, let's look at the Reuters/Ipsos survey results, which shows that 46 percent of people agreed with Apple's decision to fight the court order. 35 percent disagreed. The poll broke down results by political affiliation, just 37 percent of Republicans agreeing with Apple, while 45 percent disagreed. For those identifying as Democrats, 45 percent agreed with Apple, while 35 percent did not.

An interactive graph from Reuters:

What Are Words For?

When it comes to polling, words matter, and in this case how Reuters/Ipsos asked the question is important. From the poll question, which were published as a PDF:

Apple is opposing a court order to unlock a smart phone that was used by one of the shooters in the San Bernardino attack. Apple is concerned that if it helps the FBI this time, it will be forced to help the government in future cases that may not be linked to national security, opening the door for hackers and potential future data breaches for smartphone users. Do you agree or disagree with Apple’s decision to oppose the court order?

That question is biased a little in Apple's favor—Reuters itself pointed out that Pew included the (unsupported) FBI claim that the data on the iPhone in question was "an important part" of its investigation, while Reuters did not. On the other hand, Pew's question was heavily skewed towards the FBI by asking:

In response to court order tied to ongoing FBI investigation of San Bernardino attacks, Apple should/should not unlock iPhone.

When asked that way, it's no surprise that Pew found 51 percent of respondents thought Apple should unlock the phone. The only problem(s) with that is:

A.) Apple can't unlock the iPhone.

B.) Apple isn't being ordered to unlock the iPhone, it's being ordered to create a backdoor to allow the FBI to brute force attack its way into the iPhone.

Reuters's question is far more accurate, and that accuracy led to results that show more people support Apple than don't. It's not a clear majority, but it's a gratifying sign of progress in public opinion in this important issue.

In a related note, Apple CEO Tim Cook gave a passionate interview to ABC on Wednesday where he made the case for why it is important that Apple not create any back doors into iOS. We posted that interview in full.

Popular TMO Stories


Lee Dronick

  When it comes to polling, words matter, and in this case how Reuters/Ipsos asked the question is important

Yep, and I have seen polls that say Apple is wrong in this case.


The confusion surrounding this important issue is unfortunate.  I think the general press/media does not understand the issue themselves due to their lack of tech knowledge, and this is making the problem worse when it comes to polls and other data gathering.

The tech media outlets (at least the ones that are not Android or Microsoft shills) seem to understand what the FBI is really asking and how bad it would be.  The more I read general news and media outlets, the greater the confusion and misreporting.

I am totally happy that Apple has decided to fight this, and I hope the win.


Apple is right to fight and should never give in, and here’s why…



I just don’t believe polls.
They deliberately or accidentally skew the results by phrasing the question in a particular way. They also have no way of getting a truly representative sample. Calling mobile phones is not allowed, so the sample is limited to particular groups that still use land lines, and trying to accost people on the street. Also creeping cynicism means that many groups will not take part in polls, myself included.
No, I see little difference in a poll by PEW, or IPSOS, or a web poll on FaceBook. None of them generate data of any value whatsoever.
Indeed I’ve been watching for the last 20 years how before voting day the news is always full of election polls and then starting the day after to vote the airwaves are full of talking heads trying to explain away why they blew it. It’s time to accept the fact that they always blow it. They really do not generate data of any value whatsoever.


Bryan et al:

This poll, though based on a better formulated questionnaire than that used by Pew, still leaves a bit to be desired and is missing some critically important data to say, with measurable certainty, what fraction of the public side with Apple vs the government.

Indeed, there are three relevant questions that not only bear on this topic, but would provide important insight into how informed the public are and where their sympathies lie, and all three could be identified through a better questionnaire.

First, has the respondent heard about the issue, and if so, have them identify the source(s) using a technique called free listing, which can later be categorised into groups for analysis.

Second, ask the respondent to select from a list what it is that Apple are being asked to do. This requires that the investigators themselves understand what Apple are being asked to do, including 1) unlock a phone, 2) write software that will permit unlimited attempts to guess the password without erasing the phone’s data, along with at least three other choices. This reduces the likelihood of the respondent guessing the correct response by chance alone.

For bonus points, irrespective of the answer, the respondent could be asked if they feel this action could affect other phones.

Third, should Apple comply with this request.

These are not the properly formatted questions themselves, but identifies the information the questions are after. These would provide data on:

What fraction of the public know about the issue;
What fraction of the public actually understand the issue;
Of those who understand it, do they agree with Apple or the Feds

Additionally, it would provide insight into where those who understand the issue got their information.

Finally, such a data set would insight into both the penetration of correct information to public and what the informed public think, in addition to the variance of opinion amongst the uninformed.

Log in to comment (TMO, Twitter or Facebook) or Register for a TMO account