A paragraph from this Politico article about the French president wanna be made me wonder about a way to evaluate a president or more precisely a presidential term:

A President Bigard is a far-fetched idea — even in the age of U.S. President Donald Trump, Ukrainian President Volodymyr Zelenskiy and Beppe Grillo, the political guru of the Italian 5Star Movement.

For me, this slightly suggests that some persons are not exactly fit to be a president or generally some persons are more fit to be presidents than others.

The quality of government has been done in the past. In laymen's terms, I want to know if there is any objective way to compare two presidential terms ("this president was overall better / worse than the other one" or "this term was better / worse than the first one for the same president").

Question: Is there a way to measure the quality of a president / presidential term?

Note: the comparison makes sense only in the same country, of course.

  • 7
    Overall performance is conditional on how you weight performance in specific areas (such as economy, foreign relations, environment, ...). This weighting is the difficult problem for measuring this and results in a subjective measure.
    – Roland
    Commented Jun 10, 2020 at 11:35
  • 1
    @Roland not to mention what is the actual consequence of your actions vs trends inherited from predecessors and effects of global events. Weighting seems to me the easiest problem, because it can be clearly defined and easily changed to show different specific measures with documented reasons for the weightings chosen. Its everything that you measure and the conflating factors that are hard.
    – Jontia
    Commented Jun 11, 2020 at 7:49
  • Do you want to compare the Italian president with the American? There are numbers on everything, a good example is bombs dropped and estimated civilians killed. But some countries in general have much higher numbers than other countries so it can't be compared across countries. Commented Jun 11, 2020 at 11:15
  • 1
    @ThomasKoelle No, it only makes sense for presidential terms in the same country.
    – Alexei
    Commented Jun 11, 2020 at 11:20
  • I would make a distinction between "objective" and "neutral". Yes you can design objective measures (non-subjective, no room for bias in the act of measurement) but they won't be neutral (there is inherent subjectivity in how you go about choosing what to measure and how to weight different values) .
    – Brian Z
    Commented Jun 12, 2020 at 13:12

3 Answers 3


As with anything involving an assessment of Quality the role of the assessor cannot be neglected, and so must be subjective.

We can ask the same question of almost anything: can we objectively measure the quality of a loaf of bread? Sure we can measure the density, protein:carb ratio, the degree of caramelization in the crust and so on. We can devise a protocol for converting these measurements to a single "bread quality index" and this may even be useful! But it doesn't tell us what the quality of the bread is.

Likewise with presidents: you can look at a range of indices: approval ratings, economic growth, social contentment etc. And you could roll these up into one weighted index, which might correlate well with the general perception of "a good president". But this can only be at best a performance indicator. It doesn't tell you how good the performance is.



If there were a way, someone would've done it, but as you can see from all the historical rankings of US presidents page on Wikipedia, they all rely on surveys of historians, the population at large, etc (i.e. are subjective).

  • 1
    To be fair, there is a significant amount of clustering you can observe in those responses (e.g. most historians like Lincoln a lot better than they like Buchanan), so there probably is some objective fact here. Just not quite enough to compose a workable measurement standard.
    – Kevin
    Commented Jun 12, 2020 at 7:38
  • 1
    Re "If there was...someone would have done it,.." resembles pre-1900 arguments against powered flight.
    – agc
    Commented Jun 12, 2020 at 8:29

Millions of ways, but finding a good method would be tricky.

Quality metrics boil down to finding variable measurable properties, good and bad ones, that seem to correlate with traits and public records.

A common linear programming method is to rank a set of such properties -- the ranking might be algorithmic or arbitrary, but it must be made explicit and reproducible. From there a mathematical optimum can be calculated, and the result is a quality metric.

Because there's so many different properties to choose from, and possible sets of them, and so many different ranking methods of those sets, the permutations are endless, and therefore so are the possible quality metrics. Not all of these possible metrics will be useful, probably the vast majority will be distilled but meaningless statistical noise -- like a cake recipe with each ingredient replaced at random with another kitchen ingredient.

One method of checking a metric's utility is to see if the metric coincides with what we know without such a metric.

Example, suppose we pick out a dozen items from a list of hundreds of measurable traits of presidents -- height, weight, lifespan, wealth, number of vetos, voice pitch, number of dogs owned in office, number of affairs, length of various appendages, etc. Then we rank those dozen items, so we might compare whether the number of dogs owned was more important than weight, etc.. Once ranked, we plug in a given POTUS's data to some linear optimization algorithm, and out comes a number. Repeat for every POTUS. Check resulting list against common sense and historical opinion -- if the quality metric closely corresponds to both, it might be an interesting metric.

  • 1
    In 2020 any such attempt would likely be sponsored with partisan intentions, the better to promote an existing partisan candidate...
    – agc
    Commented Jun 12, 2020 at 12:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .