13 January 2010

Putting disasters in perspective, or Our crappy economy isn't so bad

[Note: I originally wrote this for my consultancy's blog, where references to localization make more sense. You can read it here if you prefer: http://blog.globalpragmatica.com/?p=62]

Many people are depressed these days, for many valid reasons. The economy is still a disaster. The localization industry is a mess in more ways than I can count. (I don't think I'll get much argument about that, but if anybody questions that, please leave a comment, and I'll elucidate in a future blog post.) Many of us are out of work and have been for a frighteningly long time. Many of us are clinging to scaled-back jobs. Many of us are worried about how long the work we're grateful to have will last.

When even the blue chip companies are slashing workforces and budgets and the banks themselves are declaring bankruptcy, we know our economy is a disaster.

Looking outside the devastated economy of the developed world, let's consider the vastly greater struggles in the two-thirds world.

Terminology break! (T9Y break!) When people say "third world," they mean "undeveloped or developing nations," and these represent over two-thirds of the world's population, so let's stop saying that and say what we really mean: "two-thirds world."

In the news today, hundreds of thousands of Haitians are believed dead after a major 7.0 earthquake hit, its epicenter right in the most populous part of an already fragile island. Most Haitians are black and live on less than US$1 a day. Putting this in perspective, fewer than 3000 people will killed in the horrifying 9/11 attacks. However, I fear that history will show the great failure of our humanity when the global public response to the crisis gets those metrics backwards.

Because I have spent several decades working in statistical software in various roles, I can't help wanting to look at the desperation quantitatively. Here are some graphs that will probably startle most people---and I hope horrify many of you into taking some kind of action, today. Mind you, I'm expecting to startle and horrify even the well-educated, privileged, mostly white people in the developed world who have the means to read my blog.

First, let's compare the death tolls from a handful of disasters that have filled our headlines in recent years. Before you look at the graph, which do you think was worse?
  • 9/11 terrorist attacks
  • Hurricane Katrina
  • Indian Ocean tsunami
  • Haiti earthquake
  • 2008 Earthquakes in the People's Republic of China
And how do you think the economies of these places compare?

First, the scale of the disasters. For my North American readers: remember how devastated you felt watching the TV coverage of 9/11 and of Hurricane Katrina, please.
deathToll

That's right. The devastation of 9/11 and Katrina combined are trivial compared to any of the others.

Now let's consider the economies of these places. Most of us know that USA's wealth dwarfs that of most countries by most measures. A relevant measure for this situation would be the gross domestic product per capita--that is, the total economic output of each state or nation, divided by its number of people.

GDPs

We all know that New York is wealthier than Louisiana, but did you realize that the New York-Louisiana comparison is almost meaningless in the big picture? Even the difference between those two tall bars dwarfs the size of the bars in the two-thirds world nations!

So now let's put those two ideas together: let's look at the wealth in each place lined up with the scale of the disaster in each place, as measured by GDP per capita

abilityRecoverThis composition of the most massive bloodbaths in big red bars lining up directly with the meager economic means of each place in tiny green bars is the most devastating graph of all. The biggest disasters have taken place where people are least prepared to cope with them.

There are many ways to help, and of course there are many craven imbeciles who take this opportunity to scam the people of goodwill with fraudulent donation methods. Here are some ways that have been vetted and determined to be reliable: http://www.google.com/relief/haitiearthquake

Here are some flaws in my analysis that could distract nitpickers from the clarion call to our humanity:
  • My national and state GDP data are from different years and sources, and they're probably inflation-adjusted differently.
  • I'm considering these events to have taken place in New York, Louisiana, Indonesia, China, and Haiti, where the most deaths occurred, although other states and nations were affected.
  • The costs of 9/11 and Katrina were borne nationally, but the victims were (mostly) local, so I considered the state economies instead of the national economy.
  • Estimates of the death tolls in the two-thirds world are always much fuzzier, because the poorer you are, the less likely you are to be accurately counted.
  • Estimates of the death toll in Haiti are wildly premature. Some sources say "hundreds of thousands," and while they might mean "100,000 give or take a few 10,000," a careful speaker would mean the far more frightening "100,000 or 200,000 or 300,000" by that description.
  • It's a little weird to measure ability to recover by comparing the GDP per person to the number of persons dead. The dead people are dead, and no amount of money will help them. But the people left behind are living in economies that are more or less capable of recovering.
  • These data are confounded, if you consider that poorer nations have a lesser ability to build safety into their communities. Wealthier nations have higher survival rates in times of disaster because their buildings are sturdier, more of their citizens live in buildings in the first place, their bridges and roads and so on are more prevalent and higher quality, their emergency responders are more numerous and better-equipped and -funded, and on and on and on. The ways in which wealth mitigates disaster and the lack of wealth compounds disaster are numerous and heartbreaking.
My data sources:
The analysis was my own, and I prepared all the graphs using JMP's Graph Builder.