McNamara "fallacy"

It is stated that focusing purely on quantitative data in a complex debate is often misleading as it overlooks factors best measured by qualitative data, which may also be essential to forming a judgement.

However, don't many quantitative factors also have a qualitative element?

To use an example given:

The numbers on gun violence speak for themselves. We should ban guns in the country!

This is referred to as a fallacious use of logic, because the person is focusing on quantitative data (stats) and ignoring the qualitative 'benefits' of gun ownership. The implicit argument can be constructed as follows:

P1) We can measure the impact of guns using statistics.

P2) *insert gun violence stats here*, this is too high

Implicit P) The qualitative impact of guns is irrelevant

P3) We need to reduce gun violence (and can do so by banning guns)

C) We need to ban guns.

Obviously the implicit premise is false (mcnamara fallacy), and the conclusion is also arguably a standard non sequitur since 'reducing' gun violence does not imply banning all guns (that said, if the qualitative impact is in fact irrelevant, and banning guns would save more lives than keeping some gun ownership legal, this could also be said to be a valid syllogism).

However, doesn't P2) refer to something qualitative too? Gun violence stats reflect real people's lives. If I lose a brother to a shooting, that's 1 extra person counted in the spreadsheet that the CDC will produce at the end of the year. I don't need to spell out the obvious impacts of bereavement on mental health, wellbeing, etc.

What do you guys think?

(It's been 6 months since I asked a question too...feared I might turn into Clarence Thomas!)

To agree on an answer to the question, "don't many quantitative factors also have a qualitative element?", we'll need to agree in what "qualitative" and "quantitative" mean.  Without such an agreement, there's a strong chance we'd be talking at cross purposes.

In its simplest form, quantitative information relates to what can be counted or measured ... in the gun example, information such as the number of guns in a particular area, the number of crimes and other bad incidents involving guns, the number of incidents (positive and negative) involving guns, how much $$ guns and related materials and services add to the economy, etc.  These data are generally represented by numbers that result from counting or describing factors on some sort of numeric scale.

In its simplest form, qualitative information relates to qualities and characteristics ... in the gun example, information that describes benefits that guns provide, problems related to guns, the impacts (positive and negative) that guns provide to individuals and society, etc.  These data tend to be more difficult to measure and analyze and are less likely to be presented numerically and more likely to be categorized to identify themes or general patterns.

So, a really abbreviated study of guns in a community might involve counting the number of guns in a particular area, counting how many guns each individual owns, determining how often gun owners use their guns, or counting the number of (harmful or positive) gun incidents occur over a particular period of time.  These data would provide a quantitative measure for some aspects of guns in a community.  

The same sort of abbreviated study of guns in a community might involve interviewing community members asking about: their feelings toward having a gun in the house, their feelings about having a neighbour who keeps a gun in the house, the level to which having guns around make them feel more (or less) safe, which types of guns cause more (or less) of a concern, their acceptance of current rules about guns, and their openness to various sorts of changes to gun rules.  These data would provide a qualitative measure for some aspects of guns in a community.

So, back to "do quantitative factors have qualitative elements?" ... I think a better question would be, "Do issues have both quantitative and qualitative elements?"  For me, the answer to this latter question is "Absolutely, yes!"  I believe the presence of both elements for most situations is the basis for the mcnamara fallacy that reminds us the basing a conclusion solely on quantitative information while ignoring qualitative information can lead to a faulty conclusion. This point is what I understand from Rationalissimo's questions about P2 having a qualitative element ("too high").  

One element missing from the premises given (assuming we're looking for an unbiased argument) is perhaps a P1 1/2) *insert gun benefit stats here*, this benefit is [greater than/less than] the down side of gun violence.

While quantitative and qualitative data are two distinct forms of information, they can certainly be blended together ... for example, statements like "based on interviews of 87% of the shoppers [quantitative] in a particular store, 73% of them [quantitative] reported feeling more relaxed [qualitative] while shopping on the north side of the store (where classical music was playing and the lights were pale blue fluorescent [qualitative]) and 68% of the shoppers [quantitative] reported feeling more rushed [qualitative] when shopping on the south side of the store (where rock music was playing and the lights were "bright white" incandescent [qualitative])".

By calling the mcnamara fallacy a "fallacy", we seem to accept that both quantitative and qualitative data are important and that ignoring one while basing decisions solely on the other isn't likely to lead to a good conclusion.

By the way, is there a "converse" to McNamara ... a fallacy about basing a conclusion totally on qualitative data while ignoring quantitative?

As an aside ... regarding "Implicit P) The qualitative impact of guns is irrelevant." ... is the implication that qualitative impact is irrelevant or is the implication that qualitative impact cannot be measured -- or at least, not be measured easily?

I think you make a good point. There is a difference, however, between qualitative data and a narrative (implied or explicit) attached to a quantitative factor . Perhaps the form of the argument does a better job explaining this fallacy than the description:

Measure whatever can be easily measured.

Since we can easily measure number of deaths we do it. We can spin a narrative from that data (the qualitative part that is a result of the quantitative).

Disregard that which cannot be measured easily.

Since something like "the feeling of protection and safety" is not (as) easily measured, it is ignored. Therefore, we don't even have a narrative for this because it is ignored .

So the fallacy can be more accurately be stated as ignoring the qualitative factors that do not have a quantitative component. It gets even trickier actually, because the line between quantitative measurement and qualitative is often obscured. For example, we can create a self-assessment tool asking people about their level of safety, say before and after gun ownership, and quantify that data on a Likert scale . However, this data is not typically seen as "hard" data and not seen as the same quality of data like the number of deaths due to the number of possible confounding factors in the collection of the data for the "softer" data.

In summary, all quantitative data comes with a narrative (qualitative) but if there is no easily quantifiable data then the qualitative data (narrative) is ignored, and this is fallacious.

I think that the mcnamara fallacy is you can ignore the consequences of quantitative data is you value something strongly enough, it basically makes you seem like a fanatic but it is a valid move. 

You can say that people being able to wield guns is so good a thing that it worth a few extra deaths to have it.  Trouble is that as nothing bad seems to happen if civilians do not have guns, so the goodness of civilian gun ownership is an unsupported premise.