Income Inequality’s Relationship to Violence

May 5, 2013 10:40 PM

People are more likely to kill their fellow citizens as the gap between rich and poor increases. The same is not true of civil war — although you’d think people would be more likely to turn against the state rather than their neighbor as income inequality increased, this isn’t the case.

posted by eviemath (45 comments total)

I keep hearing that the US murder rate is falling and that the gap between the rich and poor is increasing.
posted by pracowity at 11:02 PM on May 5, 2013 [6 favorites]

A strange pattern recently caught our eye. Numerous studies about murder rates have found that on average countries with high income inequality also have high murder rates.

Correlation doesn't equal causation, etc., or, in other words, whoever wrote this piece did not do a lot of research.
posted by KokuRyu at 11:20 PM on May 5, 2013 [1 favorite]

Well, the study mentioned on the article only looks at Latin American countries. That seems like a rather culturally specific sample. I can't help but wonder if there are other factors driving these numbers besides the income inequality gap.
posted by KingEdRa at 11:23 PM on May 5, 2013

OccupyWallStreet.net? More like PREoccupied...
When all you're looking at is economic injustice, it will appear to cause everything. It could be a FACTOR in a whole lot of things, but well, what KokoRyu said...
posted by oneswellfoop at 11:30 PM on May 5, 2013

It is unfortunate that two of the four linked academic articles in the link are behind a paywall. Does anyone have access to those two particular ones for purposes of academic discussion in this forum? What do folks think about the other academic articles linked? KingEdRa notes that one of them deals only with Latin American countries, for example, which may or may not relate to the context in other countries. What about the others?
posted by eviemath at 11:36 PM on May 5, 2013

Correlation != causation in two comments! New record?
posted by Justinian at 11:45 PM on May 5, 2013 [7 favorites]

ZOMG CORRELATION ≠ CAUSATION GOTCHA

oh wait, the article offers some possible explanations for the link that could be discussed if people weren't so eager to make zomg correlation ≠ causation gotcha comments, never mind, carry on
posted by robcorr at 12:00 AM on May 6, 2013 [17 favorites]

"I keep hearing that the US murder rate is falling and that the gap between the rich and poor is increasing."

Correlation doesn't equal causation, etc., or, in other words, whoever wrote this piece did not do a lot of research."

"Well, the study mentioned on the article only looks at Latin American countries."

Well, since (if you read the fucking article) they posit that what explains this best is unequal access to security (police, mainly), and mention Cape Town (which is not in Latin America), it seems like there may be a fairly reasonable correlation. And while correlation doesn't mean causality, it does mean correlation and correlation generally means that there's an underlying cause of both (in this case, unequal police protection).
posted by klangklangston at 12:01 AM on May 6, 2013

klangklangston: "Well, since (if you read the fucking article) they posit that what explains this best is unequal access to security (police, mainly)"

I was under the impression the drug dealers are better financed than the police, better armed, and that the rich pay can't really count on the police either. It's not so much a domestic politics issue of rich vs poor as it is one of international politics and global trafficking routes.

Which is to say, there is little that can be applied to the American situation. I mean, I guess you could say it is one of rich vs poor, but only if you counted everyone in the Occupy movement among the rich because they live in America.
posted by pwnguin at 12:09 AM on May 6, 2013

One of the authors of the second link in the article (to paywalled study) has a similar paper here (pdf) that might be helpful. The linked paper was from 2002, and this one is from 2004. The author, Anke Hoeffler, is associated with the Centre for the Study of African Economics.

I'll also just note that the blog that the posted article comes from looks interesting: politicalviolenceataglance.org.
posted by taz at 12:15 AM on May 6, 2013 [1 favorite]

ZOMG CORRELATION ≠ CAUSATION GOTCHA

Yes, utterly tiresome noise.
posted by Wolof at 12:55 AM on May 6, 2013

It doesn't make sense, this killing of fellow citizens as the inequality between rich and poor increases. You're not supposed to kill your fellow citizens. You're supposed to kill the rich oppressors.

Goddamn it, people. Haven't you learned anything from history?
posted by twoleftfeet at 1:33 AM on May 6, 2013 [1 favorite]

The theory that the greater the gap between rich and poor in a country, the worse the quality of life (more crime, but also more drug use, more teenage pregnancy, etc.) has been studied at length by Richard Wilkinson and Kate Pickett in their book "The Spirit Level". There is a documentary in the making as well. The book wasn't well received by right wing commentators and politicians.
posted by surrendering monkey at 2:11 AM on May 6, 2013 [3 favorites]

It got plenty bad when The Correlation Boys hooked up with the Causation Mob.

They came for Little Tony. "Little Tony," they said, "you've been looking below average height."

"Yeah, so what?" said Little Tony.

"We figure that means you want to take on the Big Boy."

"My being short doesn't mean I want to kill your leader," pleaded Little Tony. "I sent him roses on his anniversary."

"We don't care. You are statistically likely to be a threat. That's why you killed Maria."

"I didn't kill Maria. I was at Coney that night, with Lupe."

"Our models projected that you would be the kind of asshole who would rub out Maria. That's why we're here tonight. You can stick your head in the oven and die of gas fumes, or Vinny over there is going to put a bullet in your brain."

"I'd like to think about that. Have you considered Bayesian probabities?"
posted by twoleftfeet at 3:06 AM on May 6, 2013 [14 favorites]

NYTimes: The Smartphone Have-Nots - "What do we value more: growth or fairness? That’s a value judgment. And for better or worse, it’s up to us."
NYTimes: Our Feelings About Inequality: It’s Complicated - "How much do Americans care about rising income inequality? Surveys send mixed signals."
SciAm: Ask Gini: How To Measure Inequality - "Articles, studies and U.S. Census data focusing on wealth inequality rely on the Gini coefficient. How is it calculated, and what does it tell us?"
posted by the man of twists and turns at 4:04 AM on May 6, 2013 [2 favorites]

Goddamn it, people. Haven't you learned anything from history?

Oppressors learn from history too.
posted by fullerine at 4:34 AM on May 6, 2013 [4 favorites]

2nd Amendment remedies from Occupy Wall Street is one more reason to ignore them. Don't hurt yourselves kids.
posted by three blind mice at 4:36 AM on May 6, 2013

Not sure why this is so hard to believe. For the past decade or so we've been told that it's a zero-sum game, that poor non-white others are taking your stuff and getting Obamaphones and houses they don't deserve. Worse, illegal immigrants are swarming in to finish the job. What can you do about it? Make sure you have plenty of guns to keep the predators away (we'll extend you plenty of high-interest credit to pay for them.) The upper class thrives by setting the middle against the lower (or the lower against the more lower).

This hasn't even been remotely subtle, so I don't know why we're all pish-poshing the very notion.
posted by Legomancer at 5:24 AM on May 6, 2013 [9 favorites]

Goddamn it, people. Haven't you learned anything from history?

As to the younger generations, the only history they think they know was learned from movies and TV. Fictional movies and TV.

But the ban on photos of flag-draped coffins returning from Iraq shows somebody learned from the Vietnam war.
posted by Enron Hubbard at 7:05 AM on May 6, 2013 [1 favorite]

Correlation doesn't equal causation, etc., or, in other words, whoever wrote this piece did not do a lot of research.

That won't stop this article from blowing up and getting circlejerked over all across tumblr, and by half the people on my facsbook feed. It's like some new language getting hyped on hacker news or something.

It says what people want to hear in a certain group, so the fine details of why it isn't good research don't matter as much.

And yet the same people would be picking it apart and destroying it if it was some conservative slanted bad thing they didn't want to hear, like something trying to disprove the gender wage gap or other crap along those lines.

It gets a bit tiresome.
posted by emptythought at 7:48 AM on May 6, 2013 [1 favorite]

I keep hearing that the US murder rate is falling and that the gap between the rich and poor is increasing.

Both these things can be true if another factor (say access to abortion or removal of lead from gas) drives the murder rate down more than the increasing gap drives it up.
posted by Mitheral at 7:52 AM on May 6, 2013 [1 favorite]

Well, since (if you read the fucking article)

I did read the fucking article. The authors extrapolated from a single data point.
posted by KokuRyu at 8:15 AM on May 6, 2013 [2 favorites]

I have a Coleman propane portable, and more than one sharp fork.
posted by Twang at 8:29 AM on May 6, 2013

"2nd Amendment remedies from Occupy Wall Street is one more reason to ignore them. Don't hurt yourselves kids."

Thanks Swedish Neoliberal! Way to hit that straw man!

"I did read the fucking article. The authors extrapolated from a single data point."

Read it again. There are two papers linked in the intro paragraph, one which is about homicide rates in Latin America, the other about civil insurrections in Africa.

It then makes a prediction about the locations of homicide in America. Whether or not that will be true remains to be seen.
posted by klangklangston at 8:49 AM on May 6, 2013 [4 favorites]

Correlation doesn't equal causation, etc.

I'm curious, what is the etc. here? Generally, every time someone posts that correlation does not equal causation, that is the sum total of their point, so if you have more to put after that, please do so.
posted by Steely-eyed Missile Man at 8:50 AM on May 6, 2013 [2 favorites]

I read an article a while back, which I can't find, about how violent cultures like South Africa and Latin America may be linked to their colonialist development. I wonder if there is a strong correlation with violence and recent colonization. This looks interesting:

The Logic of Violence in Africa (word doc) - Sabelo J. Ndlvou-Gatsheni

While in metropolitan societies truce, peace, and friendship applied to social life; within the colonial zones the law of the strongest, violence and plunder reigned supreme. What assumptions, values, and ideas informed coloniality of radical divisions between the metropolitan and colonial zones? The creation of new identities of European, white, coloured, Indian, blacks, natives, Negroes and others was an important foundational component. Linking these new identities that emerged within coloniality was a type of social classification that was vertical rather horizontal, depicting and reflecting superior-inferior assumptions that were developed as Western modernity expanded out of Europe into other parts of the world. Social hierarchization of new identities was not only informed by race but also by degrees of humanity attributed to the constructed identities.
[...]
Why have African nationalist leaders, some of whom were put into power by popular vote, also resorted to violence as a form of governance? Why have different groups of Africans – sometimes self-identifying as races, ethnicities, tribes and sub-groups – so frequently turned to violence? This paper has argued that an understanding of how the imperial period entrenched coloniality and coloniality of being can help us to understand this continuation of violence across colonial and postcolonial epochs.

Also interesting:

Why are Buddhist monks attacking Muslims?

however any religion starts out, sooner or later it enters into a Faustian pact with state power. Buddhist monks looked to kings, the ultimate wielders of violence, for the support, patronage and order that only they could provide. Kings looked to monks to provide the popular legitimacy that only such a high moral vision can confer.

The result can seem ironic. If you have a strong sense of the overriding moral superiority of your worldview, then the need to protect and advance it can seem the most important duty of all.

Christian crusaders, Islamist militants, or the leaders of "freedom-loving nations", all justify what they see as necessary violence in the name of a higher good. Buddhist rulers and monks have been no exception.

posted by Golden Eternity at 8:57 AM on May 6, 2013 [2 favorites]

Both these things can be true if another factor (say access to abortion or removal of lead from gas) drives the murder rate down more than the increasing gap drives it up.

Which is to say that causation also doesn't imply correlation.

That's because science isn't entirely deductive. It's mostly inference.
posted by srboisvert at 12:27 PM on May 6, 2013

In Foreign Affairs, Capitalism And Inequality - "What The Left and The Right Get Wrong" is paywalled, but author Jerry Muller writes again for PBS' Newshour: The Hard Truth About Economic Inequality That Both the Left and Right Ignore

For all these reasons, inequality in advanced capitalist societies seems to be both growing and ineluctable, at least for the time being. That has implications for how those on the right as well as those on left should think about the safety nets provided by the welfare state.

CounterPunch responds: Getting Capitalism And Inequality Dead Wrong

Jerry quotes Marx, but I presume Jerry forgot about Marx discussing the inverted proportion between wages and profits. This is the natural tendency of capitalism, especially as the labor market now exists in a state of oversupply and no policies come down the pipeline to reduce the working week and increase wages in order to: a.) create demand for labor and b.) increase aggregate demand in the economy (Why Keynes of course!). Jerry even supports social welfare programs, but with mantra as caveat about how they need structural reform (nifty word for privatization or reduction).

posted by the man of twists and turns at 7:45 PM on May 6, 2013

Where's the research? The article on Latin America barely qualifies as research, it just has a table comparing homicide rates to levels of poverty, urbanization, and income inequality. There are no controls and no statistical analysis, just an informal discussion.

Let's see, according to the article Venezuela (Gini coefficient 0.44) has 37 murders per 100,000 people, while Uruguay (Gini coefficient 0.45) has 4.4 murders per 100,000 people. According to Wikipedia, the US has a Gini coefficient of 0.45 and 4.8 murders per 100,000 people; Uganda has a Gini coefficient of 0.44 and 36.3 murders per 100,000 people.

Why, it's almost as if "the relationship between urban population, poverty levels, and the incidence of homicide is still unclear." But let's continue commenting on this remarkable pattern as if it's the result of some really serious research.
posted by leopard at 7:51 PM on May 6, 2013 [1 favorite]

I wonder if there is a strong correlation with violence and recent colonization.

A connection between colonization and violence among the colonized group is definitely something that I've seen discussed before. This is not inconsistent with the main link's thesis of income inequality leading to increased violence among the economically disenfranchised groups, since economic inequality is one of the features of colonialism. But First Nations communities in Canada and Native American communities in the US certainly link the violence (especially the deliberate cultural erasure and physical and sexual abuse in residential schools and similar institutions, but also including violence from police and similar authorities, racist violence from people in neighboring white communities, and outright military violence such as at Wounded Knee and Oka) they've experienced under colonization to violence within their communities.
posted by eviemath at 8:18 PM on May 6, 2013 [1 favorite]

(In other words, unequal police protection - as klangklangston noted the original link posits as the underlying factor connecting economic inequality and increased violence within oppressed communities - is an understatement in many cases.)
posted by eviemath at 8:44 PM on May 6, 2013

The link "What I discovered when I got a gun" (from the recent gun control thread) has some nice links to information about epidemiology of violence, which may also be relevant here.
posted by eviemath at 8:54 PM on May 6, 2013

I live in Oakland, California. It's one of the most murderous cities in America.

Take a look at the income inequality in Oakland at richblockspoorblocks.com

We have a balkanized environment with abject poverty in close proximity to places with ample six-figure incomes. And the poor here turned on each other 20 years ago and haven't stopped.

Anecdotal evidence? Sure. But worth considering nonetheless.
posted by MeanwhileBackAtTheRanch at 10:35 PM on May 6, 2013

Reason: The Haves and the Have-nots - What Happens To Folks In The Lower Half Of The Bell Curve?
posted by the man of twists and turns at 4:17 AM on May 7, 2013

Some commenters above are right, it's so silly to trout out the tiresome CORRELATION IS NOT CAUSATION canard... I mean, normally that should wait until someone bothers to find a correlation first.
posted by leopard at 6:19 AM on May 7, 2013

I just LOVE how everyone here purports to know more about the topic than professional researchers because they too got a C it Stat 101 when they were 19. Jesus christ people, these are professional scholars who do professional scholarly work, and the research they are linking to is in respectable, peer-reviewed journals. That doesn't mean its gospel but for fucks sake, it doesn't deserved to be dismissed.
posted by MisantropicPainforest at 7:55 AM on May 7, 2013

Yes, on every possible topic MetaFilter Knows More Than You Can Possibly Imagine.
posted by Steely-eyed Missile Man at 8:11 AM on May 7, 2013

That doesn't mean its gospel but for fucks sake, it doesn't deserved to be dismissed.

From the respected peer-reviewed journal article on Latin American violence: "the relationship between urban population, poverty levels, and the incidence of homicide is still unclear." This is the same article that doesn't even try to run a regression or calculate a correlation coefficient.

It's amazing what you can do with some preconceptions and a hyperlink.
posted by leopard at 8:37 AM on May 7, 2013

Wait, so the article WASN'T the definitive statement on the relationship between 4 or 5 complex factors? Well then who the hell cares. /s

This is the same article that doesn't even try to run a regression or calculate a correlation coefficient.

Because those methodologies would not help them do what they are trying to do, which is to show that homicide rates are unevenly distributed.
posted by MisantropicPainforest at 8:47 AM on May 7, 2013

Homicide rates are unevenly distributed? You don't need to be an academic to figure that out -- you just need to be able to read the table here. Hong Kong has 0.2 homicides per 100,000 people while Honduras has 91.6. Doesn't seem very even to me. Both of these countries have Gini coefficients in the 0.5-0.6 range by the way (pretty unequal, more so than the US).

I have no beef with the article, which is just an exploration of the various issues related to homicide rates in Latin America. I have a problem with the blog post linked in the OP, which begins: "A strange pattern recently caught our eye. Numerous studies about murder rates have found that on average countries with high income inequality also have high murder rates. People are more likely to kill their fellow citizens as the gap between rich and poor increases." This is much more strongly stated than anything in the academic article.

And if you do want to support the statement "on average countries with high income inequality also have high murder rates," guess what, you might have to do some actual statistics, and maybe it's pretty disingenuous to link to a non-statistical article for support, and to then claim that "numerous studies" have found this pattern.

And even if there was a pattern between countries in Latin America, that doesn't mean that the pattern holds outside of Latin America, or that murder rates are caused by inequality rates, or etc etc etc.

On the other hand, I get it. We don't like inequality, here's an article that makes a fairly vague reference to it in connection to murder rates, let's dramatically overstate things and pretend there's a huge body of research about a pattern whose existence isn't even clear, and then when someone calls bullshit let's move the goalposts and cry about "peer-reviewed journals" and about how one article couldn't possibly be the definitive statement about a complex issue. Because it's really *my* fault that we have exactly one link to an article about this topic, shame on me for my weak understanding of social science.
posted by leopard at 10:08 AM on May 7, 2013 [2 favorites]

Ta-Nehisi Coates: Beyond the Code of the Streets

The Code of the Streets, a term popularized by the hip-hop duo Gang Starr and the sociologist Elijah Anderson, is the code of men who have come to feel that they have nothing to lose. Much of the struggle with young black boys and teenagers today lies in getting them to see all that violence endangers. At 13, I could imagine not going to jail, not getting shot, being a responsible father. I could not envision much more. I could name careers and other paths, but I had no real sense that it was possible for me to get there, or how.

posted by Golden Eternity at 12:30 PM on May 7, 2013

The article I linked reads, to me, like the authors found one of the studies that they linked to (the one that's behind a paywall for me) indicating that economically oppressed people are not more likely to engage in violence against their rich oppressors. Many people find this result surprising in addition to the authors of the linked article and our fellow Mefite above. This surprise is kind of the root of comments that I've seen on Metafilter before about "why do poor people vote Republican?", for example.

So the authors took that study indicating a lack of violence against rich people in economically unequal regions, and tried to dig around to figure out what happens instead. They found other studies indicating some correlation between increased economic inequality and increased violence. But if violence against the rich doesn't increase with increased economic inequality, that means that the increase must be in violence against people within the same class. They posited an explanation for this. We've had some nice contributions to this thread linking to additional evidence and studies relating to the possible extent of their posited explanation, other relevant social factors that bear on the situations in the articles they linked to, and the possible extent of the correlation between increased economic inequality and intra-class violence beyond the cases reported in the articles they linked to.

I don't understand the ... vehemence? defensiveness? ... on display in a few comments. In particular, leopard, you're arguing a straw man, and I think missing the basic point of the article, which was, "huh, economic inequality does not lead to the poor rising up against the rich; in fact, if they get violent at all, it's against each other. That's not cool, let's try to understand why that happens."
posted by eviemath at 11:55 PM on May 7, 2013

I'n arguing against a strawman? The blog post says, and I quote, "People are more likely to kill each other as the gap between rich and poor increases." It claims that "numerous studies" have found this effect, but links to a single paper that doesn't make this claim! In fact while the paper links political conditions to murder rates, the relationship it posits is much more complicated, the authors explicitly describe it as "unclear," and in fact they note that there is a very large discrepancy in murder rates between rich countries and poor countries.

I understand that the blog post is really wondering about why inequality doesn't lead to revolution, and that the focus on homicide rates is a means to an end and not an end in itself, but that doesn't mean that I'm arguing against a strawman. I'm sorry if my tone confuses you, I'm just not really fond of bullshit.
posted by leopard at 5:38 AM on May 8, 2013

this is more on income inequality and state violence, but:
Income Inequality And Democratic Militarism

Jonathan Caverly, a former submarine officer and now an assistant professor of political science at Northwestern University, has a fascinating and provocative working paper, which argues that: “[..] democracies will build larger, highly capitalized militaries as inequality in wealth rises.” the traditional view is that wars benefit elites and let the masses bear the brunt of the cost. Caverly argues that as militaries become more highly capitalized, this is reversed. Capital intensive militaries reduce the need for mass armies, limit military casualties, and shift the cost burden to tax payers.

posted by the man of twists and turns at 3:22 PM on May 8, 2013

Striking It Richer: The Evolution Of Top Incomes In The United States(PDF)

The recent dramatic rise in income inequality in the United States is well documented. But we know less about which groups are winners and which are losers, or how this may have changed over time. Is most of the income growth being captured by an extremely small income elite? Or is a broader upper middle class profiting? And are capitalists or salaried managers and professionals the main winners? I explore these questions with a uniquely long-term historical view that allows me to place current developments in deeper context than is typically the case.

posted by the man of twists and turns at 8:25 PM on May 10, 2013