Riots, demonstrations and welfare cuts

Some people just aren’t bothered by welfare cuts. If the cuts don’t affect you personally, and you only discuss them in terms of abstract principles and ‘dependency culture’, then the cuts are distant; the stuff of politicians’ rhetoric and newspaper columnists rather than real life.

What makes people sit up and take notice, though, is if the cuts cause social disorder. As the London riots began last August, former London mayor Ken Livingstone controversially linked the riots to public spending cuts. “These riots were not about poverty,” responded David Cameron, and the latest coverage has been of the millionaire’s daughter who was recently jailed for her part in the disorder. A debate has continued to rage, with the most authoritative study – the ongoing ‘Reading the Riots’ collaboration (first report here) – finding some links to poverty and a sense of injustice, but within a complex set of causes.

The analysis of the 2011 riots will run and run – but is disorder generally more likely when cuts are taking place? An interesting new paper by my Kent colleague Peter Taylor-Gooby looked at precisely this, triggering the Daily Mail headline “Riots that swept the country ‘will be repeated’ warns expect who blames poverty and job insecurity for increased unrest” – and here I explain the (currently unpublished) research beneath this.

Disorder and cuts: an empirical analysis

[The research is reported in two university university press releases and Peter’s blog post – I’ve based this discussion on the full paper which isn’t available online (but you can get a copy if you email Peter Taylor-Gooby)].  

Put simply, the paper sees if changes in public social spending and poverty are associated with disorder. The paper is based on data for 5-year periods from 26 OECD countries for 1980-2005. As well as looking at public spending and poverty, Taylor-Gooby also looks at changes in employment protection (as a measure of the work-centredness of policy) and private social spending (as a measure of privatisation).  The results are shown in the table below.

The results are strongly suggestive of a link between disorder, cuts, and poverty – a 10 percentage point decrease in public spending as a share of GDP is associated with 2 more incidents of disorder. (To put this in perspective, there are 144 total incidents of disorder over these 26 countries over 25 years). This does NOT include the impact of cuts on disorder via their effect on poverty, which is included separately in the model; a 10 percentage point increase in the relative poverty rate is associated with 4 more incidents of disorder. This link between cuts and disorder can also be seen in Ponticelli & Voth 2011 (paper here) over a longer period, and also looking at Latin American countries.

It’s important here to be clear about the meaning of ‘disorder’ here; it includes assassinations, revolutions, purges, political coups, major riots, and – critically – large political demonstrations and strikes (this comes from a database made publicly available Harvard Professor Pippa Norries). Strikes have to be at least 1000 people, by more than 1 employer, and aimed at government; demonstrations have to be at least 100 people, for the main purpose of displaying opposition to government, and excluding ‘anti-foreign’ demonstrations; and these account for the majority of instances of disorder among these countries over this period.

So to a certain extent, these results show that rises in poverty and cuts in public spending are associated with substantial political protest like the anti-cuts marches, and not necessarily with riots like those that broke out last August. There’s also the perennial issue of whether statistical models like this show causal effects (for example, cuts, disorder and poverty may all be caused by economic malaise rather than one another) – although Ponticelli & Voth show that the link is not just caused by economic downturns. These are the sorts of small print which is inevitable in comparative statistical research, even for well-conducted studies like this one.

The welfare state as an instrument of legitimacy

But still, the strong association between cuts and disorder – even after taking into account differences between welfare regimes – should give pause for thought for policymakers across Europe. This post has dwelt on the UK because I’m based here; but the issue of austerity and social stability is particularly alive in countries like Ireland, France, Spain, Italy – and above all Greece. Moreover, the link between economic inequality and social stability is a source of concern elsewhere in the world (e.g. Brazil, China), with ‘income disparity’ even topping the list of world risks in this year’s World Economic Forum.

This takes us back to some classic theories of the welfare state which since went out of fashion. Classic Marxist theories of the welfare state note that the welfare state is valuable to capitalism in two ways: it helps economic growth (providing a skilled, healthy workforce) and it assures the continued existence of the capitalist system (by enhancing stability) – and these two functions have been widely noted by welfare state theorists (e.g. O’Connor, Gough, Offe, Williams). [See Peter Taylor-Gooby’s review in his paper; it’s worth a read when it comes out].

Sometimes people forget these pivotal functions of the welfare state, mistakenly thinking that it is all about redistribution and fairness. Studies such as this are a welcome reminder that austerity may not just lead to poverty and injustice, but may cause wider problems of disorder and weaker economic growth. The precise links are complex, but the existence of a possible connection cannot be ignored.