Good news from Gapminder
I adore Gapminder. It's a vast interactive database of social measures from all over the world for as far back as they've been colllecting them, and they also have a very transparent and open process: I mean there are extensive articles explaining exactly how they gather all the data, from which sources, and how they handle it so that it can all be put into the same database. They even keep you in touch with their tweaks and changes. You can test correlations there, watch social measures change through time, compare countries nobody else has thought to compare. Something I like to do is to grab a bunch of countries that people don't always care to look at and put them into a time-lapse comparison of a couple of social variables and see how they shape up. Then I ask myself if I see a pattern. If I do, I run it through changes -- different countries, different related variables, etc. Then I look to see if the pattern holds. I have been surprised more than once -- some things I thought might correlate well with poverty turned out to correlate better with geography (north/south, or which continent), for example (naturally I can't remember which runs these were on: I'd have to play around to see if I could reproduce the pattern). There are limitations, of course: since the unit of comparison is whole countries, for example, it's not the place to compare how different classes or ethnic groups within one country are doing, or how the same ethnic group is doing in the different countries in which it resides (for example, if you wanted to track education or child/maternal moertality or longevity among the Rom). You can get the Gini index and some other measures of equality, though.
Today Gapminder's news talks about two things. One is about the limits of methodology. For the most recent datapoint, they have to rely on the projections made by statisticians. I expect they use the ones who have the most solid justification for their projections. But still, projections are only guesses, and the world is full of things that change quickly. So when you're looking at the charts, or making a custiomized graph, you should be aware that the latest data in the series is not necessarily correct.
The second thing is about the way that has played out recently. The projections for maternal mortality in Bangladesh were that by now, it would fall to 330 maternal deaths per 100 000 births (that's right: fall to more than .3% of all births ending in the death of the mother. Just visualize that for a moment. Imagine a thousand women going into labor and three of them dying). However, that number has fallen to less than half the projection: now it's 143 mothers dying out of 100 000 births. That's an incredible improvement. It means that in the last decade, Bangladesh has put a tremendous effort into maternal health and safety, and that aid given to Bangladesh has at least in this matter been well spent.
Today Gapminder's news talks about two things. One is about the limits of methodology. For the most recent datapoint, they have to rely on the projections made by statisticians. I expect they use the ones who have the most solid justification for their projections. But still, projections are only guesses, and the world is full of things that change quickly. So when you're looking at the charts, or making a custiomized graph, you should be aware that the latest data in the series is not necessarily correct.
The second thing is about the way that has played out recently. The projections for maternal mortality in Bangladesh were that by now, it would fall to 330 maternal deaths per 100 000 births (that's right: fall to more than .3% of all births ending in the death of the mother. Just visualize that for a moment. Imagine a thousand women going into labor and three of them dying). However, that number has fallen to less than half the projection: now it's 143 mothers dying out of 100 000 births. That's an incredible improvement. It means that in the last decade, Bangladesh has put a tremendous effort into maternal health and safety, and that aid given to Bangladesh has at least in this matter been well spent.