National Salary Insurance – the wrong solution to the right problem

It’s become fashionable among the British left to describe social security as broken – a cornerstone of the original welfare state that not only has lost its political support, but now actively causes worklessness and ‘dependency’.  In a revealing sign of the Labour Party’s direction of travel, the most influential and well-connected think-tank on the left has just come up with the idea of ‘National Salary Insurance’ as a way of renewing social security. Their think-piece is short, punchy, and almost a good idea – but unfortunately, the emphasis is on the almost…

National Salary Insurance: aims and delivery

The think-tank question is the institute for public policy research (ippr), chaired by James Purnell (the controversial former Secretary of State for Work and Pensions); the think-piece itself is written by his former special adviser Graeme Cooke, and Purnell has already pushed the idea on the major TV politics show and in the press.  This is clearly meant as a proposal to be considered by people in high places, and the shadow welfare secretary himself felt moved to provide a supportive comment as it was released.

There are two motivating ideas behind all this.  The first is to regain political support for welfare payments – Cooke argues that there is a toxic combination of no reward for contributions + too few demands of claimants = lack of support.

Secondly, though, Cooke wants to deal with the appalling level of British unemployment benefits, as shown below (taken from the ippr report), and which I’ve blogged about before.  Moreover, much of the British support is means-tested, and the amount available to people with savings or working partners is very low indeed; this is far from providing the ‘real protection to people in times of need’ that should animate the welfare state.

So what would National Salary Insurance (NSI) actually involve? In short, Cooke proposes an additional income-related unemployment benefit for six months, which would provide 70% of previous earnings up to a maximum of £200/pw. The clever bit is that people would then pay this extra money back to the state when they got a job again, the total repayment being a maximum of £3,445 with a zero rate of interest (i.e. being uprated with inflation). Otherwise nothing else would change; NSI wouldn’t be subject to tax or affect other benefit entitlement, and would be subject to exactly the same conditions that unemployment benefits already are. And after six months, NSI would expire and people would be left on the existing system.

False marketing

Cooke makes great claims for NSI – “more than just a ‘safety net’, the impact of NSI would be to stop households falling into poverty when they lost their job, by protecting them from a sharp and dramatic drop in their income.  It would also shield people from having to take on unaffordable debt at high rates of interest… NSI would also act as an effective counter-cyclical stimulus measure” [by paying out more during recessions, and recouping the money afterwards].

I am wholeheartedly sympathetic to Cooke’s argument for better protection against income loss due to unemployment – but I just don’t think this idea floats.

Firstly, it’s a misnomer to call this ‘insurance’ – it’s like if your car is stolen and your insurer tells you that it will give you the money for a new car, but only if you pay the money back to them. Most of us would describe this as a ‘loan’, and feel a bit cheated if it was sold to us as ‘insurance’, however subsidised it is.  [The state subsidy in NSI would be through the zero rate of interest and covering the risk of non-repayment for people who never return to work].  Cooke implicitly recognises this when comparing it to the student loans system, which can hardly be described as the ‘university costs insurance system’.

Secondly, because it isn’t insurance, the system has slightly perverse distributional effects – most of the cost of NSI is borne by the very people who become unemployed! Cooke estimates that the state subsidy (risk sharing paid from general taxation) would only be about 10-20% of the total cost of around £2-3bn, with the rest being income smoothing among people moving from work->nonwork->work. Compared to a system paid from general taxation, this will have negative effects on the people affected, who would have to pay £20-30/pw out of their pockets when they return-to-work – not only creating a new class of the struggling returners, but also sharply reducing the incentives to work. (Cooke unconvincingly tries to argue that this effect will be minimal; p12).

To be fair, the welfare state has long served more to smooth individual incomes over the life cycle than it does to redistribute between rich and poor, as nicely captured in Falkingham, Hills and Lessof’s ‘William Beveridge versus Robin Hood’ (see also their book-length The dyanmic of welfare). Yet there is something troubling about requiring this system to be funded by people who are still at a vulnerable and difficult stage – having lost their job, suffered an income drop, and be starting in re-employment of uncertain sustainability.

Finally, the extent of income protection provided by NSI would be small. Not only does it only last for six months and not cover sickness/disability, but the 70% replacement rate would only apply for people earning up to £286/pw – which is substantially below the median weekly wage of £388/pw. Still, the extra amount could be quite substantial, even for people who were earning larger amounts – £200/pw seems considerable compared to the current Jobseekers Allowance of only £67.50/pw (excluding housing benefit etc).

The future of social security

NSI has all the hallmarks of a good think-tank proposal – it tries to creatively work within the current political parameters to find new ways of tackling social problems. And Cooke is entirely right to want to reclaim public support for the welfare state; he argues that NSI has the strength of (i) only being available to people who were working and therefore actively contributing; (ii) time-limited to six months at a time; and (iii) repaid by the individual, discouraging abuse. While I think his specific idea has many of the problems that led to sustained attacks on the contributory principle, I agree with Alan Deacon and others that increased conditionality is necessary to secure popular legitimacy for any ‘real protection to people in times of need’.

Yet the very trouble with NSI is this over-cleverness. Issues around income protection are long-standing, and it’s difficult to see why the bright young things of the 2010s are more likely to come up with a dazzling new solution than the bright young (or old) things of the 1910s, 1940s or 1960s.

The simpler option would be just to raise taxes (or National Insurance) to cover a better, more generous, earnings-related unemployment insurance along the same lines. Cooke argues that this is ‘unrealistic, given the state of the public finances’. Yet either something is unaffordable for the country as a whole or it isn’t. It just doesn’t seem viable to claim that the only people that could possibly afford to pay for unemployment protection are people who have previously been unemployed.

Put another way, the choice is between putting a tax on everyone, or creating a special new tax on unemployment – and in these terms, I think the NSI proposal comes unstuck. Still, I think this is absolutely the sort of debate that people on the left should be having, for if we are ever to propose a more generous social security system then we need to exhaust all the other potential alternatives to it.

While National Salary Insurance is an affordable way of improving on the current system, my hope is rather that this is a staging post to a better, more generous, more demanding welfare state in a post-crisis world…