The economics of riots

Has anyone linked to the DiPasquale and Glaeser 1996 paper on riots yet?

We examine the causes of rioting using international data, evidence from the race riots in the 1960s in the U.S., and Census data from Los Angeles, 1990.  We find some support for the notions that the opportunity cost of time and the potential costs of punishment influence the incidence and intensity of riots.  Beyond these individual costs and benefits, community structure matters.  In our results, ethnic diversity seems a significant determinant of rioting, while we find little evidence that poverty in the community matters.

Here is a well-known political science paper on economic conditions and riots in India.  Here is an economics paper on riots in India, AER 2008.  Here is Alex’s piece on riots (gated).  In London, the riots are getting closer to the LSE.


Wouldn't poverty correlate highly with opportunity costs?

I was going to ask the same question. If the opportunity cost of losing your job and livelyhood is so low, then you must be poor in the broad meaning of the term. Right?

Tabarrok, A. 1997. A Simple Model of Crime Waves, Riots, and Revolutions. Atlantic Economic Review 25 (3):274-288

Unfortunately, I haven't found an ungated copy.

It reminds me of my experience in public school. The thugs didn't care, or "ain't give no sh!++in' f#@$!" as they might say. Evenhandedness is not impartiality, and I never developed much faith in the legitimacy of the school administration.

Ethnic and racial diversity promotes rioting. Who would have ever guessed any such thing.

Perhaps we should call riots, "diversity celebrations". That will make them PC and fashionable.

London is so vibrant right now

The race element is being ignored in the media I think. Wisconsin and then Philadelphia in the U.S. as well as London and Manchester in the U.K. all show a pretty toxic intersection between racial grievance and social media (i.e., organized flash mobs used to assault and rob by certain demographic groups)...

How homogeneously black were these rioters? Did people see a few non-black faces in these crowds at least?

They are not very homogeneously black, but I think in London they are mostly black (not sure about other towns). I think this just means that white, unemployed people are less concentrated (though more numerous) than black ones.

Last night I saw a black, Labour MP of a mostly black electorate sounding as tough-on-thugs as he possibly could on TV. Apparently he believes that this is what his mostly black voters want to hear.

Fwiw, ESPN the Magazine ran a piece a few weeks back asking why sports fans in America riot when the news is good (e.g., winning a championship), while overseas they riot when the news is bad.

VoxEU posted this paper on austerity measures and riots:

Inequality and Riots – Experimental Evidence

The DiPasquale and Glaeser paper shows that high rates of unemployment, especially male unemployment are correlated with rioting , ethnic heterogenity is correlated with rioting, and dictatorships are correlated with less rioting. Now, I'm sure we'll have a lot of pseudo-scientific armchair-behavioral-geneticist bloggers arguing that this shows ethnic heterogenity is a failure because it contributes to riots, but how many of them do you think will be arguing that it also shows the freedom is a failure because it contributes to riots, and that dictatorships are what we need? I doubt any will, it's the anti-foreign bias at work.

The Abbink, Masclet, and Mirza paper indicates to me what I've suspected for a long time, which is that deregulating the labor market will probably decrease rioting. It's bad enough those regulations, especially in France, make it hard to get jobs, the effect is worsened by the fact that the people who do get jobs get really cushy ones that increase inequality even more than normal.

You must not be familiar with Mencius Moldbug.


The notion that diversity is incompatible freedom is a commonplace observation. Plenty of folks want to restrict immigration because they view restriction as essential to the preservation of freedom. Conversely, other folks advocate restricting freedom to deal with diversity.

See "Turkish workers a mistake, claims Schmidt" (

"Helmut Schmidt, the former German chancellor, has inflamed the country's debate on immigration by saying that multiculturalism can only work under authoritarian regimes, and that bringing millions of Turkish guest workers to Germany was a mistake.

"The concept of multiculturalism is difficult to make fit with a democratic society," he told the Hamburger Abendblatt newspaper.

He added that it had been a mistake that during "the early 1960s we brought guest workers from foreign cultures into the country".

Mr Schmidt, 85, who was the Social Democratic chancellor from 1974 until 1982, said that the problems resulting from the influx of mostly Turkish Gastarbeiter, or guest workers, had been neglected in Germany and the rest of Europe. They could be overcome only by authoritarian governments, he added, naming Singapore as an example."

Diversity works fine with a government like Singapore. In a freer society, not so much.

Yes., Germany was a much better place when one people, one country and one leader was the rule of the land.


If your policy preferences come down to "do everything the opposite of Hitler", they will together be as crazy as Hitler.

Long before Hitler, Germany had a tradition of successful immigration. Poles were brought in as blue collar workers from the early days of Germany's industrial revolution. They kept their religion, but otherwise assimilated into German society. By the time of WWII they had been living and working in Germany for generations. Nazi ideology notwithstanding, they were not persecuted during WWII (unlike the Poles of Poland).

France has similar success stories. France encouraged immigration from Italy and Eastern Europe after WWI to facilitate rebuilding. Sarkozy is descended in part from the immigrants of that period.

That doesn't mean that Schmidt is wrong. Immigration sometimes succeeds and sometimes it fails. Because it is likely to be permanent, immigration is dangerous precisely because it may not work. Turks in Germany are not a success story. Algerians in France are not a success story. Somalis in Sweden (and the U.S.) are not a success story. Sensible countries restrict immigration precisely because it is a threat to the stability of any nation that allows it.

Indians in the US are a success story. Poles, Germans and Irish too. For every combination that failed there are ones that were a spectacular success.

The interesting question is whether the outcomes cold have been predicted beforehand.


"The interesting question is whether the outcomes cold have been predicted beforehand."

A very sane question. A lot depends on the immigrants... And the receiving countries. Thomas Sowell has written extensively on this subject. I also recommend "World on Fire". The author (Amy Chua) is a legend of late, but not for her best work.

But there's no downside to ethnic homogeneity, is there?

Depends. Haiti is pretty homogeneous.

Am I bad person (mebbe prone to rioting) if I don't click through to read the abstract?

Speaking of click-through, by what measure should I choose a blog from which to click-through to Amazon and purchase the Great Stagnation or, almost as good, a Plasma TV? The neediest bloggers? My favorite blog writer? My favorite blog aggregator of links? My favorite community of commenters? The most helpful professionally?

I also noticed also that Tyler's description of the materials behind the link forced you to click through to the document to see what it was all about.

Look at the opening sentence of this post for a clue: "Has anyone linked to the DiPasquale and Glaeser 1996 paper on riots yet?"

Very interesting after we had a discussion the other day about how strong or not some websites were in getting people to go to articles.

You can click here to find the discussion in my paper on this subject for further discussion.

Have you seen my paper yet? Click here. (sorry, no link).

Bottom line, riots lower opportunity costs for thugs acting on their impulses. I could have told him that.

Which Westerners thought it was such a great idea to import millions of low IQ, high-T thugs and give them money to exist and breed? Those Westerners need to be deported to the thug homelands, so they can implement all their bright ideas for multicultural paradise over there.

On the topic at hand, based on my experience at Berkeley and Los Angeles the "the incidence and intensity of riots" depends on the presence or absence of a motivated, properly armed resistance.

After the genial Berkeley cops permitted the the disastrous People's Park ('91) riot, the powers called in the frankly sadistic Oakland PD who quickly quelled the Rosebud riots ('92) by shooting everyone up with wooden shot-gun pellets (never used in the UK). (The mandatory racial sensitivity training riot ('93?) confused most observers; what to do with a bunch of hammered fraternity guys breaking into the Gap, Urban Outfitters, etc. and stealing faux-rapper jeans and shorts? I think they just passed out.)

In L.A., the armed and motivated stopped the '92 riot from incidenting intensely on their territory. Koreans pointed AK-47 from makeshift crenellations in midtown, Beverly Hills & Culver City cops formed a barrier at La Cienega, etc. Unlike them, most LAPD lived in Castaic or other outlying areas and weren't motivated to protect "their" community, leaving it to the marauders.

Ok, I've apparently missed the reports of doctors, lawyers, and CEO rioting and smashing and grabbing and causing general mayhem.

I've also missed the reports of the middle class white and blue collar workers taking time off from work to go looting and stone police.

Granted students who were the children of doctors and lawyers and white and blue collar middle class parents did take time off to march on Washington or engage in sit-ins, but most feared they would be sent to die in some Asian hell hole. Blacks and gays marched, and did get support from fellow idealistic students who took time off from class, but not work, to protest the culture of lynchings and gay bashing, which struck the white middle class with good futures simply for the sin of fraternizing with other races or gays. But those types of criminal activity are the same as done to create the good ol USA in the 18th century.

Do we count the Boston Tea Party and the other American protests and criminal acts before and after in the same class as the Watts riots,and the riots taking place in London and Syria and Bahrain, Egypt, Libya, Tunis,...?

But perhaps the point is that these riots always result from bad government policy, the oppression in Egypt and Syria et al and the government failure to make a better economic future, and also the oppression and government failure to create a better economic future in the UK.

Thus riots are always a sign of the failure of government to serve the people, but instead the government is considered with strong evidence to serve a favored class of elites who are a power wielding minority.

That paper on riots in India...well people with proximity to the mills were targeted because the police couldn't get there

not that they wanted to.

please, this is not a resource war. please. it is matter of identity.

The link about the riots getting closer to LSE is wrong. I study at the LSE and those protests have not happened "one tube stop away". In fact, they have happened "one tube stop away" from the LSE student hall where I live, but not from the school. So today I ventured in the morning to see what had happened in the area during last night (at Walworth road, next to Elephant and Castle) and I could see that there were 30, 40, 50 stores which had been broken into, the glasses smashed, people protecting their stores. Horrific view. This is a 15-minute walk from my residence but about a 50-minute walk or (at least) 3-4 stops away from the school.

Thank you for your fantastic blog.

It might be unfortunate for people whose property is destroyed in riots that things like the minimum wage, and unemployment, do not take "chance to cause rioting" into account in the equations.

But destroyed buildings = job creation, right?

How to avoid any possible rioting as a result of unemployment, either in the UK or the US:

Also will leave this here, since the subject of the post on which comments are closed ( is still relevant:

"The structural theories have their problems, but they can better explain why corporate profits are high and can better explain the distribution of unemployment across income and educational classes."

Basic economic theory (the simplistic explanations used in schools) says that if there is structural unemployment in the sense of not enough people with education, this would cause a rise in wages for those jobs. High corporate profits, from the basic tenants of free market competition, can only arise when only a limited number of companies are able to supply a particular good that is in demand, such as iPods.

This has likely been pointed out by various authors since the start of the 'recovery', but it would seem to be worth repeating in relation to this particular statement about corporate profits.

This would seem to be the kind of thing that is worth pointing out for its 'reality-defying' nature compared to usual assumptions about the problems with the economy. Why are consumers buying goods that aren't necessary for basic living from a highly profitable, noncompetitive market? Because economists tell them to, of course.

"We still don't know what we are doing."

If only people would remind themselves of this more often!.. that blog post, while consistent, was something that most people likely read and then forgot about, while hoping that people only spent more things would, somehow, all work out. Corporations and wealthy people aren't buying enough private jets this year, but just wait until they accumulate another $1.5 trillion of government spending and we'll definitely see the economy start to take off..! Yep.

"We examine the causes of rioting using international data, evidence from the race riots in the 1960's... In our results, ethnic diversity seems a significant determinant of rioting"

They should have used evidence from the Vancouver riots to see if the passion of sports fans are a significant determinant in rioting.

You skip the study that demolishes your case.

"Austerity and Anarchy: Budget Cuts and Social Unrest in Europe, 1919-2009, by Jacopo Ponticelli and Hans-Joachim Voth, Discussion Paper No. 8513, August 2011, Centre for Economic Policy Research: Abstract Does fiscal consolidation lead to social unrest?"

If ethnic diversity causes riots how come there aren't so many in Canada, Australia or New Zealand?

All recent riots in Australia had ethnic diversity as their public, direct, cause.

perhaps ethnic diversity isn't causal, only places that are diverse are more likely to experience rioting, and like I said above, identity (either perceived genuinely or whipped up by politicians) has a real role to play. Problems of identity are more likely in places which are diverse.

So how does Seoul, riot capital of the world and one of the most racially/culturally homogenous city fit into this?

You can view footage of such riots on YouTube: color coded bamboo staves, police advancing in tight phalanxes, doing fight chants, no less. They are strikingly different from the race war which the British government is frantically quashing.

I took a very quick look at the DiPasquale and Galeser paper - it is not very good. They have a small sample size, which is particularly bad since they are doing tobits, they don't have an identification strategy or deal with endogeneity even cursorily and they pretty much jump from their 60s evidence to '92 LA without justifying external validity. Also, I once saw a riot in a rural are - three dudes fought with sticks.

And I really can't make out if they are reporting coefficients or marginal effects.

In the US riots data, given that it is from the 60s and that most of the riots were from the civil movement era, it is really not surprising that the only significant variables are the race ones - I am pretty sure if I put in a dummy for black communities everything else would become insignificant.

Plus, I am not very happy with their use of the word "rioutous" to indicate riot-prone countries. :)

How much will they hate it? Unrest and budget cuts over the long run (

"In London, the riots are getting closer to the LSE." There's aways a silver lining.

Any data on social gap issues? You know steal a million get a knighthood, steal a pound go to jail.


I think those who are reacting to the reference to ethnic diversity in the quoted section of the abstract really should click through and read the paper. The ethnic diversity effect is not nearly so well-supported as some of the commenters here seem (want?) to think.

In our results, ethnic diversity seems a significant determinant of rioting

So an area with 1/3 Swedes, 1/3 Japanese and 1/3 Jews would be likely to have riots?

Matt Sinclair has mentioned this:

Rob a bank and you get 10 years The Bank robs you and they get a seven figure pension.

Factors that contribute to rioting are population size, the breakdown of respect for social order, poverty, the lack of opportunities for personal advancement and Debt.

Today the people that led the world into debt by their corrupt practises within the Banking, Insurance, and Financial sector have been reappointed by President Obama to head his financial team. No one has been brought to justice for the debt all the world is now paying and rioting about.

All the banks had AAA status just before they collapsed from Standard and Poor the same people who have just downgraded USA economy?

University Professors who advised the governments were working without declaring their paid interest for these corrupt banks etc. just before the crash.

If you check the facts you will find a transfer of wealth from the poorest to the top one percent and these crooks are still running all our economies.
This was achieved by getting companies like standard and poor to give false assessments of bad debt which was then sold to our pension funds as triple A.
Whilst ever these crooks go unpunished and are rewarded by huge golden handshakes and top government jobs, riots will get worse.

We all have a remedy it is our vote and a free press.
Make sure you give your vote wisely and all the politicians who supported crooked bankers, insurance companies, are swept from power.
To the press it is time to expose these crooks. Name and Shame

How to solve or Worlds Economic woes.

The problem with all our world economic problems was the collapse of the housing market brought on by corrupt banks swindling their customers.
This can never be solved by governments pouring Taxes to prop up these Banks.

What we need is a stimulus to the housing market and I would do this in two ways.
First give tax incentives to builders and make land available from government land for redevelopment.

Secondly; we need mortgages which guarantee the buyer that in the event of a collapse in his house price, the difference will be paid upon sale of that property to the buyer from an insurance company. A new type of insurance of protected mortgage also the availability of work protected mortgages.

This can be government protected like the Banks are government protected and I would make these mortgages available throughout all the banks.

Any costs involved could be paid for by selling off all social housing.

This scheme will within five years have the economies of the entire world growing again.
It will also bring confidence back to the financial markets.

It is possible to build eco friendly houses for around 50,000 and if government land is freed up for this, we can look forward to prosperity for everyone.
Because every ones homes will start inflating and not just bank profits

Peter Baxter

Comments for this post are closed