The first shows that people on the very lowest incomes no longer have a recognisable safety net, guaranteeing sufficient income to meet even their most basic needs, like an adequate diet and keeping warm in the winter.

The second, based on my research with Eleni Karagiannaki at LSE and published this week, looks at the striking difficulties now being faced by those in the middle of the income distribution. Even on these income levels, about one in five have recently reported struggling to afford food and other essentials, though for those near the bottom of the distribution this rises to nearly two thirds. While they are likely to be in jobs, they face increasing insecurity at work. If they are in mid-life, they are far more likely to face the insecurity and expense of renting their homes privately than a generation ago, when they were more likely to be homeowners. And the rising  cost-of-living means that many face a choice between having an acceptable living standard today and being able to save for an uncertain future.

The instinctive reaction to this is that it’s impossible to help everyone, and the rapid growth of deep poverty means that efforts to relieve hardship need to start at the bottom, with something like an essentials guarantee. And a more adequate safety net will only be affordable if it is quite narrowly targeted, which means that people in the middle of the distribution are likely to see little.

I agree with this perspective when it comes to a general transfer of income. Targeting is needed. Without it, you end up arguing for a universal basic income which, if adequate, requires eye wateringly high tax rates.

But the evidence on the situation of households near the middle of the income distribution suggests that the policies which could help them most are not simply a matter of the state giving them some money (especially if it had to take it back again in higher taxation, which cannot all come from the rich). Rather, policy makers should be looking at how government can get alongside these groups to make it easier to translate a middling income into more stable and adequate living standards. The evidence points to three aspects of this, by no means comprehensive.

First and most importantly, individuals and households need greater stability. One-in-three people in the middle fifth of the income distribution in any one year will be in a lower bracket the following year. This is partly because if you lose your job in the UK, your income goes down much more than in most European countries, where unemployment benefits are initially linked to former wages; new earnings-related protection, proposed for example by the Fabians, could help tackle that. But it’s also about improving stability of employment and worker rights – something that the present government promised but neglected to do, and which should be a priority for the next one. Housing stability can also be improved, and the Renters’ Reform Bill presently going through Parliament starts to do that, after years of tenancies’ continuation being entirely at the discretion of the landlord.

Secondly, households who have escaped dire poverty but still have modest incomes can be helped and encouraged to protect their living standards in the future through saving. This can be most effective in creating resilience against hard times if matched by support from government and from employers, who can benefit by retaining workers who have more stable incomes, as was shown in the furlough scheme. Auto-enrolled pensions have illustrated the feasibility of this partnership; the Resolution Foundation’s latest report on savings shows convincingly how this can be extended to savings for unexpected expenses and temporary losses of earnings.

A third aspect of policies to help people on middle incomes is that while they cannot offer across-the-board financial support to middle-income households, they should be sensitive to the evidence of where difficulties for this group arise. One example is the neglectful treatment of what was once the child benefit ‘affluence test’ – withdrawing this benefit to families earning over £50,000. When introduced in 2013, this affected one in ten families; by next year it will be nearly one in three, simply because the threshold has not changed. In fact, a one-earner couple with two children will now have all child benefit withdrawn at a point where family income is still below the median. These facts about child benefit really struck me in doing this research; I had not realised how much had changed in this regard over the past decade. This shows how important it is to look at evidence afresh and not get stuck with old assumptions.

Another example of evidence related to people on middle incomes which is outside normal assumptions concerns differences in outcomes between singles and couples. We are used to dividing working-age people into those with and without children, but sometimes it’s useful to do so according to the single-couple distinction. Singles can have greater difficulties making ends meet, partly because two paycheques go further in meeting household needs than one, due to economies of scale, particularly when it comes to housing, where these economies are greatest. In addition, the ‘protective effect’ of having a partner in the workforce provides a degree of insurance against job loss. Lone parents have a particularly tough time because they also have caring responsibilities, but even singles without children face different challenges from couples. For example, we found that in the middle fifth of the income distribution, only four-in-ten singles but over seven-in-ten couples have at least some secure employment, and for singles the rate of insecure employment has risen sharply.

My fellow-author, Eleni Karagiannaki found none of this surprising, from her Southern European background; there, it has always been taken for granted that support from other adult family members is central to social protection. A difficulty in the UK is that we are often willing to look at adults as individuals (for example in income tax), but do not make things easy for people living on their own. For example, couples get twice the tax allowance of singles, even though their costs are nowhere near twice as high, and only about half of couples have children. Little has been done to improve the supply or lower the cost of single people’s housing, and policies to change this would be particularly helpful for those struggling with middling incomes but soaring costs.

Some of the above policy approaches would cost governments little or nothing, because they involve new regulations, such as employment protection measures, rather than cash transfers. Others, such as an earnings-based employment insurance scheme, would have substantial costs, but these can be strikingly low considering the benefits. The Fabians’ version of such a scheme would require workers to pay just 7p extra for each £10 earned to guarantee they retain substantial proportions of their earnings during periods out of work. So, policies helping people on middle incomes should be on the next government’s agenda, but in a considered and evidenced based way.

Read more on Donald’s middle incomes work