Indicators

The answer is 42.
But what does 42 mean, really?

Remember two things: Heisenberg and Bias.

* * *

Dirty old Heisenberg

If you look my way
I might smile or scowl
or curse
depending
on who you might be
friend or foe
or dirty old man.
Even when I seem
to do nothing
know that ignoring you
is a calculated reaction.

* * *

Inherent Bias

We wear yellow glasses. We are looking at a blue world. We see a green world. If we know our glasses are yellow, then hopefully we will remember that the green world we are looking at is actually bluer than that which we see as green.

* * *

The problem of measurement or observation or knowing is not new. Heisenberg was mentioned and there is Schrödinger and his poor cat or the quantum suicide scenario, which are more obtuse discussions about observers and the observed, whose practicality is questionable.  I am a scientist and aid worker, so that’s what’s practical to me.

An interesting and practical discussion about the perspective of being outsider on the Tales From The Hood blog rung true with me in certain situations, but he left something out.  He is right, sometimes an outsider can filter out a local bias, but sometimes an outsider can exaggerate a local bias as well. The important thing is to be aware of one’s effect and one’s biases and the situation and the complex interactions therein. We can never eliminate these things, but we can at least be aware of them.

* * *

Intelligent scientists and aid workers alike know that by measuring something you do indeed affect it. We also know, despite not being quantum physicists, that we will never know the exact value or position or whatever. Whether numbers or simple observations, we are careful to minimize these things by triangulation, awareness and thoughtful measurement methods.

An example: Hydrogeologists poke holes (i.e. dig wells) into aquifers and in doing so disrupt what was once, and in being aware of that, we accept it because it is good enough to understand the system. It is that, in fact, our goal in observing anything: to understand what is going on. 

When we do science in aid work, we are still scientists and indeed report a range of values or other interpretations that reflect uncertainty. The methodology (how these numbers are arrived at) is rigorous and has been peer reviewed, published and duplicated by experts. The documentation of such methodology is available, transparent and clear, even to someone outside the relevant technical field who has basic mathematics background, some curiosity and half a brain.

An example: We measure a certain number of randomly sampled children to represent a greater population which is very specifically defined and say, for example, that for this defined population there is an 80% chance that we are beyond the emergency threshold of global acute malnutrition or that the level of X problem is Y% with a 90% confidence interval of +/-Z.

* * *

On NOT understanding Heisenberg: The danger in science and in aid work is when people do not take the time to understand what a number or observation means: how it was measured, the constraints and limitation inherent in it, some people don’t even know the units associated with that number (e.g. 42). Without some knowledge about the mechanics of the observation, you will never know what anything actually says about the system or population or, really, about anything.

On NOT understanding Bias: This can be your bias (your yellow glasses), as well as the bias of the presentation. The most dangerous and common thing of all is to talk about a number or to compare two numbers that are not related at all and then draw some wide-sweeping conclusions from them. There are many particularly worrying examples of this. Some examples of this are well intentioned but careless, and some straight-up-dirty manipulation of data.

Back to practicality. We make observations to come to conclusions. I hate conclusions. They are usually wrong because they depend on a complexity of things. But alas, conclusions are a necessary evil.

Understanding Heisenberg + Bias = better conclusions. Knowing what the number or observation means helps filter some bias, but can’t get it all. By being aware of bias, and especially the bias of the audience, and by trying to take off or lighten the yellow glasses makes the conclusion stronger, more realistic and inherently better.

* * *

One problem with my argument is that it is too idealistic and simple. Bad science, bad measurements and bad observations are rampent. Which makes pretty much any view of them moot. (But can you expect me to address everything in a single blog post? Really?) And the point is, by using the noggin a little and not just accepting what you read, you can identify this "badness."

Another problem with my argument is that I am guilty too, I sometimes draw conclusions too fast, I have biases, I have a sick sense of humor that gets the best of my biases. But sheesh, I am human. The point is not that I know best, it is that I am thinking about it and you should too.

* * *

Did I bore you? Confuse you?
If so, read it again.
If not, enjoy lighter side of indicators.

* * *

In every niche there are standard, quantifiable indicators. For me professionally, this often boils down to things like the prevalence of WASH related diseases, number of people per latrine, liters of water per person per day, fecal coliforms per sample etc. We also have qualitative indicators which may give another, less number-full view of the situation, but equally valuable.

I have come up with a few indicators of my own, which mean nothing, are not scientific and are most undoubtedly influenced by my sick sense of humor and biases. The ones I will present here relate to the level of development of a place, mostly the economic development of that place.

I hoped I didn’t have to say this, but I guess I need a disclaimer: In all seriousness, these are really complicated subjects. Please realize that by presenting these ridiculous indicators, that I am demonstrating exactly the danger of this entire discussion. My apparent hypocrisy is intentional. I am a sarcastic person and I intend to remain that way.

* * *

Fire: And god said let there be light, and it was good. And man discovered fire and flint. And then matches were available on the local market. And then lighters. And then lighters with flashing lights inside. And then lighters with flashing light and a button to make those flashing disco lights be projected onto a close wall or floor.

Beer for your buck: The bigger and cheaper the beer, the more dire the economic situation of the country.  Can the average cost per swig of beer be positively correlated to the average income of a household or somehow be an inverse proxy indicator for GDP? You find big, cheap beers in Congo and Guinea; smaller, expensive beers exported from Europe. Do note that Budweiser has a pretty cheap, twenty-two ounce beer. America is going to shit.

McDonalds for peace: Some people claim that there are fewer wars and less strife in countries with McDonalds. Clearly a chicken-vs-egg argument should ensue, but in my recent history I have lived in countries with relatively high levels of strife and not one had a McDonalds. (Not my own idea, just one I thought was worth sharing.)

Gambling as an indicator of early recovery after displacement:

I see Haiti eight months after the earthquake. Life is not normal here and won’t be for a long time, but it is transitioning back slowly. With this bias, I look for signs of normalcy. I search them out. And I found one. The lottery is up and running, people are gambling again, so one can conclude that this place is getting back to normal.

The lottery kiosks or shacks or buildings are newly painted. They say “BANK – LOTTO- CHEZ TITI.” Other types of kiosks haven’t been repainted, even if they are starting to reopen. Not the water sellers. Not the pharmacies. But indeed the lottery bosses are back in business. Folks are crowded around the blackboards on the walls and doors which are updated twice a day with the winning numbers. A clear early warning indicator of recovery.

So my boss asks, can you assess Maya Camp, where people are re-settling?  How is it there? I visit. I don’t even have to get out of my air-conditioned, white Land Cruiser to draw my conclusion. I see that a Chez Titi kiosk is up and running within the camp and I know that these people are ready to get back to normal. They are doing fine.

* * *

Perhaps I can solicit Paul Collier and his army of ingenious grad students to run an in depth retro-analysis on existing data to check the validity of my new-fangled-sarcastic indicators.

* * *

On inspiration: This post was gourbled out of my head, somewhat, by checking out the aforementioned blog, and in particular his post about Skynard and simplicity, actually many of his posts. Clearly I dig this dude’s style and outlook on aid work. Take some time to check out his posts.

Editorial note added afterwords: Hey number-geeks and aid-workers, here is some interesting questions posed by statiscians dealing with the MGDS. It's a start. (Thanks Trish, for pointing this out to me.)