4 ms·
It does matter. Consider of a hypothetical country consisting of two equally large regions. One region is desert, only 10,000 people live there and unemploymen
by danieldk 8y ago
It does matter. Consider of a hypothetical country consisting of two equally large regions.
One region is desert, only 10,000 people live there and unemployment rate is 20%.
In the other region there are 1M people and the employment rate is 3%.
Now if you look at the map, it seems that the country is doing pretty poorly. Half of the country has a bad employment rate. Whereas in reality it only affect a small portion of the population.
- crispyporkbites 8y agoDoes that apply to the european regions listed here? How significant are the differences in population density that they make a tangible difference?
- Luc 8y agoFor example, the regions of higher unemployment in the south of Belgium make up more than half of the territory, but contain only about a third of the population.
- ptaipale 8y agoArea FI1D, Northern and Eastern Finland, is the largest NUTS2 region by land area you see on this map. It covers about two thirds of Finland (>200 000 km², about 3 times the size of Republic of Ireland), and has a population of 1.29 million people. Unemployment is ~10 %. Population density is a bit over 5/km² This area is not very significant in terms of number of people, unemployed or not, but it is very visible on the map (also due to map projection which enlarges northern areas). (The four NUTS2 regions in continental Finland have roughly the same population, the fifth one Ahvenanmaa/Åland has a population of 30,000 people).
- sammorrowdrums 8y agoLondon has a larger population than the whole of northern Ireland (8,787,892 compared to 1,876,695)
- sammorrowdrums 8y agoAnd to add to that the whole republic of Ireland has only 4,761,865 so certainly it seems hard to comprehend how many people are unemployed in the chart.
- ProblemFactory 8y ago> One region is desert, only 10,000 people live there > In the other region there are 1M people This has already been normalized: these are European NUTS-2 regions, which are drawn to contain roughly equal populations. Each NUTS-2 region is supposed to be 800k to 3 million people per region. It's not perfect, but it's not just geographical borders with wildly different populations. For example, London is split into 5 regions, most big cities are their own region, and small countries are just one region. Going into even smaller regions would be nice (and are defined in the standards as well), but it seems that nobody has put in the effort to collect data at that granularity.
- sammorrowdrums 8y agoThis is important, but I'd still say it's difficult to reason about population size affected in this chart. Its great for showing regional spread and has value, but like most data one visualisation will always prioritise some information over other info. Comparing nations has always been difficult, so I do like that they at least try to show region's proportionate to population.
- tom_mellior 8y ago> Each NUTS-2 region is supposed to be 800k to 3 million people per region. It's not perfect It's far from perfect in some countries. The Community of Madrid with 6.5 million people is its own region (or Catalonia with 7.5 million); that's more than Scotland, which is cut up into five NUTS-2 regions. The Paris region (Île-de-France) has 12 million inhabitants, which is more than, say, Austria or Hungary or the Czech Republic, each cut up according to more or less traditional geographic divisions. You have a point according to how NUTS should work, the grandparent does have a bit of a point according to what it's like in practice.
- ekianjo 8y ago> which are drawn to contain roughly equal populations No, the visualization is completely wrong to begin with, because areas with practically population zero like in the northern parts of Scandinavian countries are still comprised of the same regions as their southern neighbors, which makes absolutely no sense and increases their importance area-and-color-wise on the map.
- chrisseaton 8y agoI don't understand what you mean. The map in your hypothetical situation nor this blog post wouldn't misrepresent anything. It doesn't claim to be a map of 'unemployment rate confounded with population density' in the first place. You've imagined that. It's a map of simply the 'unemployment rate'.