If you need a definition of the concept, there is no better illustration than the interactive maps below, via the Urban Institute:
There must be a pattern here, somewhere. I’m sure we can find it… oh, wait…
Great gif by Pew:
Via the Pew Research Center, the name of the game is concentration:
It is interesting to note that the only two religions that have spread far wide beyond their place of origin are Christianity and Islam, and this was accomplished through political means, military expansion, colonialism, not because of any voluntary adoption.
I am currently (re)writing our online course on marriage and family (a topic I generally stay away from but them’s the breaks). However, as usual, I decided to integrate a module on data exploration. I stumbled upon this Pew research report on global aging that contained a lot of information and data, so I thought I’d just share some of what I found interesting.
First of all, I find this interactive visual very useful as an introduction the state of the world population by age groups, from 1950 to 2050:
You can either examine data for the US (numbers and percentages) or the world. Focusing on percentages, for the global population, you can clearly see which age group is projected to grow or shrink. So, for instance, the 15 to 64 population is stays pretty much stable from 1950 to 2050 (from roughly 61 to 63%). The under 15 category peaks in 1965 (with 38%, the end of the baby boom) but projected at just over 21% by 2050. However, for the 65 and older age group, the shift is from about 5% in 1950, to 15% projected for 2050. These increases and decreases are clearly visible just by eye-balling the graphic. Switching to the Us, that shift to a “geezerification” of the population is even clearer. as it is for most wealthy countries.
The global overview is nice but only as a starting point. There is some need for some fine-tuning by country and since my main topic here is aging, let’s look at that, for selected countries
Let’s do that below the fold:
Yup, I’ll be watching the World Cup. And a few media outlets have come up with some great dataviz on the subject. My favorite so far comes from the Economist, and you can spend a lot of time playing with this one:
It is fully interactive and it visualizes all the World Cup goals with a lot of neat filters (countries, year, stage). You can see which countries are big scorers and try to derive playing style based on that.
Business Insider looks at the average temperatures at kick-off (it will be interesting to see if that correlates with other things, like goals scored, etc.):
More interesting is this dataviz dashboard that comprises information regarding which league players play in during the regular season, as well as the experience v. age factors:
And here is a bunch of various stats on the World Cup (the whole scrolling thing is a bit tedious… simplicity is often better):
And via Le Monde (in French), here is a dataviz video summarizing roughly the same statistical info:
This is Todd’s turf more than mine but check out this dataviz from the New York Times:
The big question is how and why considering that crime rates are not exactly exploding right now, and it’s not like US law enforcement is fighting the Sinaloa cartel.
Well, first, there is supply to be dumped, according to the article:
“As President Obama ushers in the end of what he called America’s “long season of war,” the former tools of combat — M-16 rifles, grenade launchers, silencers and more — are ending up in local police departments, often with little public notice.
During the Obama administration, according to Pentagon data, police departments have received tens of thousands of machine guns; nearly 200,000 ammunition magazines; thousands of pieces of camouflage and night-vision equipment; and hundreds of silencers, armored cars and aircraft.”
And in a very bureaucratic and Weberian fashion, once the tools are there, they will be used. And sure enough:
“The equipment has been added to the armories of police departments that already look and act like military units. Police SWAT teams are now deployed tens of thousands of times each year, increasingly for routine jobs.Masked, heavily armed police officers in Louisiana raided a nightclub in 2006 as part of a liquor inspection. In Florida in 2010, officers in SWAT gear and with guns drawn carried out raids on barbershops that mostly led only to charges of “barbering without a license.””
And so you end up with a militarized police force even though the crime statistics do not justify it. In addition, the use of militarized gear changes the way police forces approach situations, i.e., they do so more aggressively since the balance of force is more in their favor. And since the equipment is free or would be scrapped if unused, it is easy to see why police chiefs would get stuff that, really, they don’t need. But once they have it, the equipment acquisition has to be rationalized. So, you get jewels like these:
In the Indianapolis suburbs, officers said they needed a mine-resistant vehicle to protect against a possible attack by veterans returning from war.
“You have a lot of people who are coming out of the military that have the ability and knowledge to build I.E.D.’s and to defeat law enforcement techniques,” Sgt. Dan Downing of the Morgan County Sheriff’s Department told the local Fox affiliate, referring to improvised explosive devices, or homemade bombs. Sergeant Downing did not return a message seeking comment.
Some officials are reconsidering their eagerness to take the gear. Last year, the sheriff’s office in Oxford County, Maine, told county officials that it wanted a mine-resistant vehicle because Maine’s western foothills “face a previously unimaginable threat from terrorist activities.””
What it does though, is turn police officers into soldiers in occupied territories where all civilians are potential enemies and neighborhoods into potential war zones. And we all know which neighborhoods will face militarized police forces, of course, because we already know who bears the brunt of heavy policing.
The World Cup is about to start, so, unsurprisingly, football is in the news a lot and I have noticed the publication of quite a bit of dataviz relating to sport in general, and football in particular.
First, the Guardian’s great Datablog has a series of quick charts on the statistics of the different teams.
Top and bottom five by average team age (not age by players where the oldest is 43 and the youngest is 18):
The oldest teams are in Latin America and the youngest two are in Africa but the spread is pretty narrow (about three years).
Top and bottom five by caps:
Spain won in 2010, so, it kinda makes sense that they would use experienced players. However, France, for instance (we all remember their great “performance” in 2010) has a comparatively low number of caps for a country that won 16 years ago and was finalist after that. Maybe there’s a generational change at work. I am surprised that Japan would rank so high. Japan is not a big football team on the international scene. Or maybe there is a limited pool of players, so, they get to accumulate more caps.
Top 10 by number of international goals:
The top three are really not surprising at all. In number of goals, Spain and Germany completely outclass every other team. But where are Brazil, Argentina, or Italy?
Clubs with 10 or more players in the World Cup:
No surprise here either. It is the Big teams (those that can afford these kinds of players) that top the list. Note the large representation of the English Premier League. Speaking of which…
Leagues with the most players featured:
The wealthy leagues top the list. Again, no surprise here.
Top and bottom teams with players based in home country:
We know the big leagues extract players out of other countries (especially African countries). That’s another resource flow from the periphery to the core.
But football is a global sport and the migration of players outside their own countries during regular season is also well illustrated by this dataviz (I’ll put the ginormous but interesting version below the fold) by the Pew Research Center:
But sports also involve spectatorship and the Economist has done some data work on that as well:
Even though, the US tops the list with attendance figures for the NFL, that is a relatively small percentage of the population. Surprisingly (at least, for me) is the largest percentage that goes to the Australian football league. Also note that Canada is counted with the US except for football. One could argue that the US has three popular sports (football, baseball, and basketball, then, to a smaller extent, hockey) where other countries tend to have one largely dominant sport (often football).
Pro Publica (a very worthwhile news organization that you should all read) has published a very nice data visualization regarding imports and exports of guns by states (and nationwide) based on tracing done after criminal activity. You can go state by state and look at where guns came from and where they went across state lines.
So, first, the national map:
Ok, I decided to pick Illinois since this is where I am located and in, every discussion on guns, someone will claim to have the decisive argument that Chicago is not murder-free despite a gun ban. So, let’s look at Illinois imports:
About half of the guns traced by the authorities came from out of state. Who would have thought that somehow, things, like guns, would make their way to Chicago and that somehow, Chicago does not live under a dome, Stephen-King style.
For the most part, these imported guns came from Indiana, Wisconsin, and Mississippi. The dataviz also has these imports (and exports) broken down by state.
Anyhoo, Illinois also exports guns to other states:
But check out New York state:
So, close to 70% of guns traced in NY state came from out of state, and you can see that, for the most part, they came from other East Coast states.
But compare that to Texas:
Only 18% of the traced guns came from out of state.
No big surprise here. If a state has strict gun laws, one can expect more imports. But if a state has fewer restrictions on guns, then, by definition, imports will be less necessary.
Get check out the whole thing.
Seriously, am I the only one who sees a problem here?
The 37% for biofuels is the only one that looks ok to me. But food crops are listed as 25% in the paragraph but on the chart, it looks more like 12%. Livestock reads as 3% in the paragraph but barely registers on the chart. Non-food crops read as 5% in the paragraph, but at about 1.5 / 2% on the chart.
I re-read this several times, just to be sure, but the paragraph clearly refers to fig 2,1 (which it is, even if it does not show up on the photo).
This is from a serious book, published by a serious academic publisher, and written by one of my favorite (really serious) sociologist, so, this bothers me a lot even though it looks like first world problem.
If you do not subscribe to the US Census Bureau updates, you are missing out on a lot of great information. Yesterday, the update was about poverty, how to measure it, and the latest data on the subject. So, if you teach poverty (and don’t we all), the following are some pretty interesting and discussion-starting visualizations.
First off, of course, we are all familiar with the official poverty line and its shortcomings. So, the USCB uses a supplemental measure designed to overcome these shortcomings. The differences between the two measures are summarized in this nifty infographic (click on all the images for a larger view):
How many people are poor and how long do they stay that way? According to the Census Bureau summary:
“31.6 percent of Americans were in poverty for at least two months from 2009 to 2011, a 4.5 percentage point increase over the prerecession period of 2005 to 2007. Poverty was a temporary state for most people; however, 3.5 percent of Americans were in poverty for the entire three-year period.”
Visually, things look like this (pdf version here):
These data and visualizations are related to a new report on poverty from the Census Bureau. The basic concepts it uses are as follows:
So, using these different measures, here are some of the results.
First off, the overall picture:
No big surprise here. Poverty rates were higher after the recession than before. Note that the use of a three-year panel overestimates episodic poverty compared to calendar year measures, and underestimates chronic poverty compared to calendar year measures. It’s pretty obvious why (people in poverty for more than 2 months but less than three years).
Now, let’s look at chronic and episodic poverty with various demographics:
No surprise here either: blacks and female-householder families lead the pack on chronic poverty,followed by Hispanics and under 18. On the other hand, whites, seniors, and married-couple families have lower rates of chronic poverty. These trends are the same for episodic poverty, but with higher rates.
More specific data show the same trend as well:
Now looking more specifically at poverty entry rates:
The same categories of people as above entered poverty in higher than the overall rate between 2009 and 2011.
Now, poverty exit rates:
Again, this time, one would look at rate below the overall rate to determined which categories of people have a harder time escaping poverty, basically, any rate below 35.4. The pattern we identified above persists.
How long does poverty last? How long is a poverty spell?
This lopsided trend is interesting. Either a spell is short or it is long, but the middle durations show lower percentages. But the vast majority of spells are a year or less.
Looking at median poverty spells:
Here, the only surprise is the rate for 65 and over. Remember, that category had a lower poverty rate but for those who are poor, then, the median spells are longer than the overall median.
This is confirmed by the data on poverty persistence between 2009 and 2011:
This is all pretty interesting stuff. You can read the full report, if you are so inclined.
I should note though, that poverty spells may be short, but insecurity and precarity are not and take a toll.
Take a look at this great visualization of life expectancy around the world. You’ll need to do a lot of zooming in and out, but it’s worth it. And it’s a visually appealing way of presenting it. I like the organization by continent and by color. Don’t forget to scroll all the way down to the bottom for the time series.
Good visualizations on this and related topics can also be found using the Human Development Report visualization tool:
The US Census Bureau has released a series of recent maps showing the wealthiest and poorer counties, nationwide, using data from the Small Area Income and Poverty Estimates program.
First, median incomes (for all, click on the images for larger view):
The Northeast metropolitan corridor is pretty striking: where the power elite is. As the report notes:
“The U.S. Census Bureau reports that five of the counties or county-equivalents nationwide with the highest median household income in 2012 were located in Northern Virginia. Among them were Arlington County, at $99,255, Fairfax County, at $106,690, Falls Church (an independent city), at $121,250, Loudoun County, at $118,934, and Stafford County, at $95,927. Falls Church and Loudoun also had among the lowest poverty rates in the country.”
Then, poverty rates:
Now, one can see a Southeastern corridor of high poverty, with a few other spots (the tips of Texas, and parts of South Dakota).
Thirdly, child poverty:
The patterns are a bit harder to distinguish (partly because of the color scheme), but you can clearly see that the east coast wealthy corridor is very white and that the same Southeastern corridor is there as well.
Finally, shifts in median income:
Now, one can see a red (as in increase) crossing the central part of the country, from North to South. If I remember correctly, this was also the region least affected by the economic recession and high unemployment (especially the Dakotas). The Southwest is impressive in its decrease (the West overall, but really, the Southwest, especially).
As the report states, again:
“he findings also show that median household income is higher in nearly half of the counties in the Dakotas now than it was before the recession began in 2007. Between 2007 and 2012, 55 of the 119 counties in North and South Dakota experienced a statistically significant increase in median household income. In contrast, of the remaining 3,023 counties or equivalents nationwide, the same was true of only 56 of them. Of all the U.S. counties with a statistically significant change in income relative to 2007, 89 percent experienced a decline.“
[In order to get a better view of larger visualizations, you should click on the “<>” symbols on the upper right corner of the page for flexible page width. I have finally figured out how to embed from Tableau but it makes a mess of the page. As always, click on the images for larger, interactive, views]
In this post, I will wrap up, for now, another set of visualizations on global opinions on homosexuality that can be used as sociological exercises in data analysis. Again, the data come from the Pew Research Center and the visualizations were made in Tableau.
Quite often, one hears the argument that views on homosexuality are generational: younger people are more tolerant than older generations. So let’s explore that hypothesis with global data. For that purpose, I thought it might be useful to divide the set of countries into geographical regions, and then, get the average:
Quite clearly, in all regions except Africa (for which the rates of acceptance of homosexuality are very low across the board), the hypothesis is supported. Older categories seem to be less accepting of homosexuality.
Let us now go region by region and look at selected countries for each.
Right away, you can see that the average for Africa would be even lower if it weren’t for South Africa. For South Africa, the rates much higher than for the rest of the region, but they do fit the pattern of greater acceptance of homosexuality for younger people. Otherwise, it is hard to distinguish a clear pattern for the other countries as the rates are really low. Look at Uganda, for instance. It is the opposite of what one would expect. And even though there is one age category for which data is missing in Kenya, the pattern is reversed. But again, with such low rates, little differences look like larger differences.
Let’s look at the other low average region: the Middle East:
Here, it is Israel that is the big outlier for the whole region and drives up the average, as South Africa did for Africa. And for Israel alone, one can see that the middle age cohort is the one with the highest acceptance rate. Lebanon then follows, with a pattern supporting our original hypothesis. Then Turkey, with rates much lower than Israel and Lebanon, but higher than the rest of the region, and this time, it is the middle age cohort that is the least accepting (but again, the actual percentage point differences are very low). I confess to being surprised by the overall lack of acceptance in Tunisia. I guess secularism does not extend to attitudes regarding homosexuality.
Next up, Asia:
This is one of these cases where the average is actually misleading (see back up) especially when the countries are so divided. On the one hand, you have countries with very high rates of acceptance (Australian, Japan, Philippines, and to some extent, South Korea… look up South Korea in my previous post, it was interesting case there). And one the other hand, countries with very low acceptance rates (China, Indonesia, Malaysia, and Pakistan). But an average smooths these massive differences out. That is why looking country by country is necessary. If I were to hypothesize, I would argue that the high acceptance countries are either Christian or more secular compared to Muslim, more religious countries.
Does our generational pattern hold here? Mostly yes. We lost the patterns only for the countries for extremely low acceptance rates.
Moving on, Central / South America:
Obviously, the rates are high and our age pattern holds solidly for every country in our sample. El Salvador and Bolivia seem to be trailing behind a bit. That is usually an indication that some more digging is required, especially some correlation work. On the other hand, Argentina, Chile, and Brazil have very high rates. Venezuela and Mexico play middle of the pack. Those high rates are interesting in a region marked by strong Catholicism, but also Pentecostalism.
Let’s move North and look at North America:
Depending on how you look at it, either Canada is driving up the regional average, or the US is driving it down. I blame evangelicalism, puritanism and conservatism. The US rates are actually comparable to the middle of the pack South American countries and other countries on other regions score higher. This validates the idea that economically, the US is a highly developed, core country, but on social issues and indicators, it scores in a fashion resembling more semi-peripheral countries. Our age hypothesis, though, holds for both countries.
And last but not least, Europe:
Obviously, for Western and Northern Europe, the rates are incredibly high. However, no one following the news should be surprised by the low rates in Russia, Poland, and Greece.
For instance, this:
The lower house of Russia’s parliament unanimously passed the Kremlin-backed bill on 11 June and the upper house approved it last week.
The Kremlin announced on Sunday that Putin had signed the legislation into law.
The ban on “propaganda of nontraditional sexual relations” is part of an effort to promote traditional Russian values over western liberalism, which the Kremlin and the Russian orthodox church see as corrupting Russian youth and contributing to the protests against Putin’s rule.
Hefty fines can now be imposed on those who provide information about the lesbian, gay, bisexual and transgender community to minors or hold gay pride rallies.”
So, no surprise there. Things are chaotic in Greece with the rise of neo-fascists (who are usually not friendly to gay even though these movements drip homo-eroticism).
The age pattern, though is much more irregular, but within the context of high rates across the board for the other countries.
Finally, and just for fun, I tried my hand at a heat map on this. The colors correspond to the regions (and the countries are grouped that way). The size of the square is a function of the %, by age categories.
That is it on this topic. As you can see, there is a lot of exploration to be done and puzzles to be teased out on this.
[I should have noted this before, but, in order to get a better view of larger visualizations, you should click on the “<>” symbols on the upper right corner of the page for flexible page width. I have finally figured out how to embed from Tableau but it makes a mess of the page.]
Here is just quick data snippet from the Pew Global Attitudes Project that measured changes on acceptance of homosexuality (link to Tableau) for selected countries, from 2007 to 2013. Here is a static dual point plot, the link in the previous sentence will take you to the interactive version:
First of all, out of 26 countries, 17 had a better acceptance score from 2007 to 2013, 2 had no change, and 7 had a worse score in 2013 than in 2007. But just to look at these raw numbers does not tell us much. Not every country was at the same level of acceptance in 2007.
Look at the 2007 static bar chart (interactive one here):
So what changes in 2013? (interactive chart here)
France takes a tumble, for sure, but is still in the top 5. I would explain this with the fact that France legalized gay marriage this year, and the run-up to the final passing of the law revealed a rather dark and ugly underbelly of homophobia that might have been latent when there was no law in perspective, but that reared its hideous head when legislative action started. The top of the list is still occupied by the same countries with a bit of shuffling but still high acceptance scores. Interestingly enough, the US plays middle of the pack in both years, probably due to its high level of right-wing religiosity and puritanism compared to European countries. The bottom of the list stays roughly the same. The religiosity hypothesis seems confirmed by this scatterplot:
We see a very negative correlation: as religiosity increases, acceptance scores decline. Finally, let’s look at the differences between years (interactive chart here):
Looking at the differences between 2007 and 2013, South Korea and France are the two shocking stories. The acceptance score for South Korea jumps 21 percentage points (even though the majority still finds homosexuality unacceptable), and France takes a 6 percentage points drop (Which I explained above). Frankly, I have no idea what the heck is happening in South Korea. Most Western countries see improvements in their already high acceptance scores. Another noticeable improvement is Kenya, with a jump of five percentage points, which gets it out of the bottom three. The bottom of the list is now occupied by quite a few Western countries. One could argue rising activism in the Global South, and backlash in the Global North.
Looking a bit more globally, the North / South differences are still striking: