Basic income, social democracy and the welfare state

I enjoyed listening to lots of excellent presentations at the Basic Income Lab workshop Basic Income after COVID-19 – Social Security, Work & Wealth on 10 December. It’s terrific to see energy and ideas coming together around a movement making many positive contributions to welfare policy discourse.

It seems like a good time to post about how my social democratic welfare statist position interacts with basic income, something I’d been meaning to do for some time. This will be the first of a couple of posts. In this post, I discuss two powerful critiques advanced by the basic income movement – of mutual obligations and repressive paternalism, and of means testing, particularly family means testing. I agree with both of these, but think they are better advanced, in the present context, by fixing unemployment benefits and other conventional welfare state policy. In a latter post, I will discuss the symbiotic relationship between full employment policy and successful welfare states, and why we should promote universal payments not as a welfare replacement or living allowance but rather as part of a completely separate strategy for socialising ownership of capital and natural resources.

I appreciate the basic income movement for arguing strongly against the onerous and punitive mutual obligation arrangements, as well as repressive paternalism like income management, attached to Australia’s job seeker and disability support payments. Such intensive requirements have no serious economic rationale. Their role is to be a sick punishment on the unemployed and to generate profits for private job search providers. Leftists should push back against the fetishisation of work for work’s sake.

But sometimes it seems to me that the basic income movement skips a couple of steps in arguing in favour of basic income – jumping right over a classic social democratic welfare state approach.

Fixing unemployment benefits is different to establishing a basic income

The straightforward solution to onerous conditions on the unemployment benefit is to… end the onerous conditions on the unemployment benefit. Until the 1980s, there were barely any obligations at all in practice, and even these only became particularly onerous from the 1990s and 2000s. Basic income proponents need to make a much broader and more radical case than the need to wind these back. They need to argue not simply why the unemployed should be treated better, but why everyone should receive guaranteed income, including people who elect to exit from the labour market. Unemployed people only make up a minority of the beneficiaries of a basic income.

A universal basic income would require radical increasing and restructuring of the tax system. Assume, for example, a benefit of $30,000 (just over $80 a day) for 20 million adults and $9,000 for five million children (30% is typically considered the child equivalent of an adult payment). That would cost $645 billion. Deducting existing cash welfare payments (around $130 billion) would bring the cost to around $515 billion. That would involve roughly doubling all revenue, including income tax, GST, company tax and so on. We could lower the payment rate down to pension level, but it wouldn’t fundamentally change the story.

Let’s say we were to implement some means tested model of basic income (or negative income tax). The first point to make is that this wouldn’t really reduce the effective tax impost on labour income, or reduce disincentives to work, it would just apply some of it via means testing taper rates rather than formal taxation. Secondly, it would be absolutely inferior to a UBI on social democratic welfare fundamentals, and I would argue inferior to a strong (but conventional) social democratic welfare state. The problem is that although welfare and tax systems overlap somewhat – both contribute to social equality and both affect incentives – their optimal roles are very different; they are imperfect substitutes. Welfare systems are good at social insurance – providing rapid cash assistance to smooth over fluctuations in living standards when a hardship emerges. Tax is good at taking a longer-term view (typically annual) to reduce inequalities in earnings. Trying to fudge these two approaches leads to inferior outcomes.

For example, if an above average income earner loses their job, they have immediate financial needs. But if they have already exceeded the relevant income threshold for the year in the tax system, presumably no unemployment protection is available to them under a negative income tax model. Whether they can get unemployment benefits depends on what point in the year they lose their job; that is, on whether they’ve had time to trip over the relevant threshold. This is weird and arbitrary. Maybe they’ll get some kind of refund at the end of the year, but that’s not really helpful for their immediate needs. The system probably also requires clunky annual income estimations, which in the existing family tax benefit system results in people accruing large debts that have to be repaid at the end of the year. This is a failure of income smoothing. The welfare system should make your finances easier and more stable, not the opposite.

Such an approach also doubles down on the most obvious large-scale flaw of the Australian welfare system compared to other countries: failing to provide social protections and insurance for families beyond the poor. There is just no good reason for this. Some level of social insurance is an efficient service for the government to provide across all income groups. A failure to do this is both sub-optimal economics and bad politics, a road-block to building a broad political coalition behind the welfare state. It also implies a lot of means testing complexity and mess, which leads to inequitable arbitrariness (people miss out because they made silly mistakes or forgot to fill in a form). My disabled father just lost $20,000 under means testing because he failed to shift his personal finances quickly enough into an aged care capital deposit, despite already being in the aged care facility.

These complexities often trip up people who are really struggling, due to poverty, disability or other factors. Tackling means testing has to be central to any progressive welfare reform agenda.

What to do about NILFs?

Unemployed people are still only a fraction of beneficiaries even if a basic income is means tested down to the very poor. Unemployed people are only a minority of the unwaged population. A considerably larger number are “not in the labour force” (NILF). The latest conveniently packaged data I can find on the NILF broken down by age (September 2013) puts the NILF population aged 15 to 64 at 3.6 million. There were around 681,000 unemployed people aged 15-64 in 2013, so unemployed people made up only around 16% of the non-employees of working age. But many of the NILF might have a strong claim to welfare support, such as people with disabilities, full time carers and students so lets set these aside and focus on the remainder, who are mostly undertaking home duties and early retirees, but also a small number of vacationers, volunteers and an “other” group. This NILF subset with a less clear claim to welfare accounts for 1.4 million people aged 15-64. This group is over twice the size of the unemployed. It is interesting why the basic income discourse focuses almost entirely on the unemployed, when the policy argument that must be made is why people doing home duties, early retirees, vacationers and “others” should be paid a full living allowance equal to people who lose their job that are looking for work.

Some back on the envelope numbers to illustrate the point. The cost of providing a living allowance at $30,000 only for unemployed people would be $20.4 billion. The cost for the unemployed plus the “weak case” NILF subset above? $61.8 billion. (Note these numbers are total costs, not incremental cost relative to current policy. The numbers cannot be neatly reconciled with existing JobSeeker costs which include some people in part-time work or training, while excluding some unemployed people based on asset and partner income means tests.) So a tripling of the cost is at stake on this conceptual issue alone, setting aside the other NILFs and low income employees. There is an inevitable trade off between adequacy of living standards for the unemployed, and supporting anyone and everyone who consciously chooses to exit the labour market for whatever reason.

I’m not sure where to go with this, but my impulse is the subset of NILFs above should not get a full living allowance (although everybody should receive a smaller payment as a right, reflecting their stake in socialist ownership structures – more on this in the next post). There is nothing morally wrong with early retirement or vacationing, but financing that choice is on them.

Household labour is already implicitly subsidised as it is untaxed, and it is unclear why it should be further subsidised with a living allowance from the state. Moreover, my hunch is that the community in present and foreseeable contexts would not support it either, and that the conflation of unemployed interests with a larger group of people who deliberately choose to exit the labour market for their own or their family’s benefit would harm the cause of the former.

Some activation requirements might be inevitable without a basic income

But an unemployment benefit (as opposed to a basic income) requires a way to fire wall the payment from NILFs. This means people have to establish their eligibility for the unemployment classification, which by definition means actively looking for work. Some type of activation requirements are thus probably necessary, but there is no reason why they have to look anything like the punitive “mutual obligation” model we’ve had since the 1990s. I would consider an appropriate level of activation requirement to be declaring oneself ready and willing to work, and completing something like one or two job search activities (such as submitting an application or attending an interview) in a fortnight. Depending on circumstances (health, location, caring responsibilities and so on), this might be reduced to one per month, or even exempted for a period. It is better for people to focus on a small number of serious applications than being forced to spam employers.

All of the Job Network should be abolished, and replaced with a re-established Commonwealth employment services provided as a public service to assist in job matching.

Abolish family-based means testing

A useful contribution from the universal basic income movement is promoting an individual-centred model of welfare eligibility. Most means testing is problematic, but one type in particular has to go: testing based on the family. This is consistent with classic social democratic approaches to the welfare state, which have long promoted the broader socialisation of what were traditionally considered familial, religious, charitable and local community obligations. There are good arguments for this in terms of economic efficiency, individual freedoms and egalitarianism.

Individual assessment is necessary to properly do the consumption smoothing job of welfare for different sorts of household compositions. If a worker in a two-income household loses their job, they are down an income they rely on. They should be eligible at least for a basic unemployment benefit. There are efficiency gains to this. Risk pooling at the societal level is vastly superior to two-person “risk pooling”.

Moreover, household labour has risks of exploitation, so it is important to ensure people doing home duties have viable options for exit. As soon as someone decides to enter the formal labour market and look for work, they should be eligible for a liveable income, unconstrained by other members of their family. If they establish welfare eligibility for any other reason – being a child or a student, having a disability and so on – they should receive it. This mitigates against dependency and promotes autonomy. It should be considered a basic principle of liberal justice to be treated as an individual before the state, for the same reason people can’t be fined or made bankrupt for actions of their partner. Once you establish personal eligibility in an approved unwaged category, you should get a payment.

Individual based assessment is also required to coherently promote economic equality. If unwaged individuals are forced to rely on waged members of their household, this greatly reduces the per capita income of that household. (This makes sense if the unwaged person has elected to do home duties, as sharing the household income is the implicit payment for domestic production that benefits the household, but not if they are unwaged because they are a child, old, living with a disability, a student or because they can’t find work.)

Variations in household composition thus drive a great deal of inequality. Even in a utopian socialist economy – where workers had captured all capital income and were each paid equally – there would still be considerable inequality, and even significant poverty in households with relatively high numbers of unwaged members. To address household composition inequality, a good welfare state provides a basic living allowance for eligible categories of unwaged individuals.

What we need is a social democratic welfare state that provides cradle to grave support for each individual citizen and resident across the contingencies of life.