It’s a Matter of Degree

Too often, researchers are leaving data on the table when they survey respondents. Typically, when a researcher wants a respondent’s take on an issue, the respondent is unfairly asked to cram their thinking into a few preselected options (do you agree or disagree, do you support or oppose, etc.). The problem with this approach is that opinions are not so simple. Asking respondents to sort themselves into predetermined static choices oversimplifies data before the researcher ever gets their hands on it. Respondents have a whole lot more than just what side they take on an issue to share; they also know the degree to which they believe their stated opinions. Collecting and analyzing that data is interesting, insightful, and increasingly necessary.

This is where Trendency comes in. We are changing the way questions are ask, and how data is analyzed based on the premise that, often times, the most important information is found in shades of grey, not in oversimplified black and white interpretations of the world we live in. Trendency allows respondents to allocate their responses, giving people the ability to share nuance, producing a more accurate picture of where they really stand.

Let’s take the right to self-pardon as an instructive example. On June 4th,  in reference to the Muller investigation, the president tweeted that he has the “absolute right” to pardon himself in the case of indictment. This prompted debates among pundits, legal experts, and the public; does President Donald Trump have the right to pardon himself?

To give context, Article II Section II of the Constitution states that “The President […] shall have the Power to grant Reprieves and Pardons for Offences against the United States, except in Cases of Impeachment” which creates somewhat of a legal gray area, not explicitly outlining whether the President can pardon himself. However, the Department of Justice released a memorandum in 1974, a mere four days before Nixon resigned, arguing, “Under the fundamental rule that no one may be a judge in his own case, the President cannot pardon himself.” So, it’s fair to say there is a legitimate legal debate to be had here.

With impeachment the only mechanism to hold the president to account should he decide to pardon himself, whether or not the president has the right to do so is also an important political question. Given this context, we thought it would be interesting to look at how the American public perceives the right to self-pardon. 


Again, we do not ask our respondents whether they agree or disagree, but to what degree they agree and to what degree they disagree. Above is the average scores that respondents gave for each position. Democrats, Independents and third-party members behave as we might expect: most of those voters disagree with the right to self-pardon.

Republicans, however, are an interesting case: Republicans favor the right to self-pardon, but only ever so slightly. Why is this the case? Could it be that Republicans are generally ambivalent towards the right to self-pardon? Perhaps instead, the party is being torn in two, with one faction of the party enthusiastically agreeing with the right to self-pardon, and another faction behaving a little bit more like Independents and Democrats. Each of these possibilities is, given just the data above, plausible. And with traditional research, that’s about as far as we would get. It leaves gaping ambiguity in place. The void, sadly for most of us, ends up being filled with yammering pundits.

With Trendency data, however, we can dig deeper. Below are the distributions of support within each party. Democrats are unimodally grouped at the far-left end of the distribution, indicating that almost all of them vehemently do not agree that there is a right to self-pardon. The same is generally true of independents and third-party members, but to a lesser degree.


The Republican distribution, on the other hand, is remarkably uniform. There is no cluster of supporters at one end of the spectrum or the other, or even really in the middle. There is a complete lack of consensus within the Republican Party on this issue.  You couldn’t even say there is consensus around uncertainty.

Looking at these distributions is helpful in that they give a quick understanding of the landscape of public opinion within each party. However, there is more we can glean from our data. To put more context on these distributions, let’s turn to another one of Tendency’s tools: Commitment and Rejection Indexes. These indexes categorize respondents based on the certainty and consistency with which they answer. Below are the commitment and rejection indexes by party for the right to self-pardon.


Two things jump out immediately. First, unsurprisingly, 92% of Democratic partisans reject the notion of a self-pardon. That is consistent with what we saw in the violin plot above.  Secondly, there is an odd three-way even split among Republicans. A third heartily support the right to self-pardon, a third reject it, and still a third are ambivalent, falling into the “persuadable” category.

It’s hard to overstate the value of such an insight. Say you hope to persuade the public that the right to self-pardon does not exist. Now you know where to put resources. It’s a waste of time to keep persuading Democrats. Not only are they on your side, but firmly so.  And although there are both persuadable Republicans and persuadable Independent/third-party voters, the percent of Republicans who are uncertain about this issue is substantially higher. That seems to be where your attention should go.

Such a realization bucks our prior expectations and would be impossible to uncover with traditional, binary-style data. By measuring the degree to which a voter holds a belief, we avoid painting over important features of the opinions and values that people hold.   

Brendan Gleason