Author Archives: davidsligar

There’s not nearly enough middle class welfare

Despite the extraordinary circumstances, the government’s massive support package does give us some lessons for day-to-day policy. The pandemic reminds us of the power of government to protect us against risk and to smooth our incomes through tough times. Because of the government’s power to pool risks and reallocate spending across time periods, it is much more efficient and effective at providing such protection than individuals are themselves.

We are reminded of the importance of welfare right now because so many Australians suddenly need it. In terms of political economy, there is a critical mass of people who are pro-welfare, so the government acts accordingly. When the going gets tough, even the conservatives turn to socialism.

But the basic principles in favour of a stronger welfare system – insurance and income smoothing – hold at other times as well. All of us will face risks and financial challenges, even though usually we won’t have so many people facing them at the same time. All of us will draw down on the welfare state throughout our lives, and we will be grateful for it when we do. However, in normal times, the welfare state can be out of sight, out of mind for many of us – and therefore it is under-appreciated.

Alas well-intentioned people across the spectrum have contributed to this situation by waging a multi-decade war on so-called middle class welfare. The consequence is two-fold. Many middle class Australians have lost protection from risks like losing their job, and as a result welfare is more out of sight and under-appreciated than ever. People are more likely to support policies if they are tangibly helpful and salient in their daily lives, so there’s a viscous circle: welfare cuts foster social attitudes conducive to further cuts.

As a result, Australia is virtually unique among comparable rich countries in that we have no satisfactory system of unemployment insurance for the middle class, due to means testing such as the partner income test. Our already low job seeker payments start to get cut if you have a partner who earns $993.50 a fortnight, about $25,830 annually. You can’t get anything at all once your partner earns $1,845 a fortnight, or about $47,960 annually. This is only half the median pay for a full-time job.

Thanks to our national obsession with means testing, even families well below the middle can’t get unemployment support. You might lose your home – but the government won’t give you a cent. We are more miserly in this regard than even the notoriously ungenerous American welfare state.

What we’ve seen with coronavirus is the government rush to improve this dire situation as more Australians suddenly come in contact with the welfare system. For example the government has roughly doubled unemployment benefits, and upped the partner income test threshold. This is implicit recognition that when large numbers of Australians turn their minds to our existing welfare system, they will find it unsatisfactory.

Unfortunately even with the temporary changes plenty of people still fall between the gaps.

Despite the higher partner income test, you still won’t be eligible for any JobSeeker Benefit if your partner earns $3,086.80 a fortnight ($80,260 a year) – still 10% below the average full time wage.

There is no good reason for this. Governments, with their large budgets and ability to spread resources between different people and across time periods, are inherently better at insuring for risks like unemployment than individuals are.

Partner income testing rewards single income households over dual income families and is grounded in outdated notions of the male bread winner. It is unable to capture the reality and complexity of many modern relationships – such as those that do not necessarily function as a joint economic unit. We don’t assess tax like this, so why we persist with it for welfare?

The government should use the coronavirus crises to experiment with a more comprehensive, universalist approach to welfare. It has temporarily eliminated the assets test for Job Seeker Benefit, so why not eliminate the partner income test? Why should the dole be paid to a let-go finance executive in Point Piper sitting on $20 million in assets, but not a jobless labourer in Campbelltown because the latter has a partner with an ordinary full time job?

Fortunately such anomalies are easy to solve with universalism. People should be entitled to a payment by simple membership of an approved category that is a legitimate barrier to paid work. For example, if you are unemployed, disabled, elderly or a full-time student you should get a payment automatically. Primary carers would all receive payments on behalf of their children below working age. Such a system would be simpler, more elegant and more coherent than the one we have now. It would be financed through higher taxes – which would be the chief political economic hurdle.

But the alternative is a fragmentary, convoluted and incoherent patchwork of means testing that fails to deliver equity and fails to protect the middle class. It imposes extremely high effective tax rates on low to middle income groups as benefits are clawed back through means testing, sometimes double the 45% rate that rich people claim is a work deterrant. It does not address a market failure in social insurance even though the government can and should do so efficiently. And by losing the middle class, the Australian welfare state will never command the mass support needed for a compassionate approach. Work for the dole and robodebt are the consequences of this model.

Efficiency looms small

My post a couple of days ago highlighted the centrality of high taxation to economic equality.

One well known cost of high taxation is inefficiency. Okun made a famous analogy about this trade off: you take from the rich and give to the poor with a “leaky bucket”.

But how do we strike the right balance on this trade off? There’s obviously no simply answer to this question. Some methods of redistribution have much smaller efficiency costs than others. More fundamentally, there is no consensus on how much society benefits from redistribution. The vast majority of people believe that a dollar means more for the wellbeing of a poor person than for a rich person, what is sometimes called the diminishing marginal utility of income. But there is also some evidence that high levels of inequality can harm economic growth, economic stability, public health, social cohesion and democratic institutions.

Further, some of us believe that inequality can be inherently unjust in principle, independent of the above harms. Non-utilitarian objections to various forms of inequality come from diverse perspectives – from political philosopher John Rawls and Nobel prize winning economist Amartya Sen to the Pope. The various benefits of equality can be combined into a “social welfare function” that ranks different social states, formalising the leaky bucket trade off between equality and efficiency.

But for all the abstract economic, philosophical or religious musings on the right amount of redistribution, in practice it will always be a social choice – conditional on the beliefs, values and preferences of particular communities – and resolved through concrete societal institutions such as democracy.

Nevertheless, it is worthwhile thinking through inequality and redistribution from an economic perspective. The point I want to make is that even extremely large costs to economic efficiency can be outweighed by benefits to economic welfare arising from redistribution.

To illustrate, I will undertake simple, stylised analysis counting only the benefits of equality arising from the diminishing marginal utility of income, setting aside all other economic, social and moral benefits.

Such analysis raises the question of just how much more a poor person benefits from a dollar compared to a wealthy person. Although there is no agreement on this within economics, a popular assumption is that a given percentage increase in income the same improvement wellbeing regardless of the person’s income. For example, $100 means as much to someone on $20,000 as $1,000 does to someone on $200,000. Folks have been assuming this (logarithmic) relationship between income and wellbeing since pioneering work on the “utility function” by Swiss mathematician Daniel Bernoulli in 1738.

In my opinion, it’s probably conservative. Empirical work suggests that it understates the relative benefit of money to poorer people. On the basis of this work, UK Treasury recommends distributional weightings that favour poorer citizens more heavily than a logarithmic shaped utility function would imply. Some survey research finds that many in the community have quite radical beliefs about diminishing marginal utility, implying that matching the benefit of $100 to someone on $20,000 requires someone on $200,000 to receive more than $10,000, rather than the $1,000 we are assuming (even this simplification is significantly more conservative than the actual survey results). The famous 73% optimal top marginal tax rate result from Diamond and Saez disregards the marginal wellbeing of the very wealthy altogether. Nevertheless, the analysis here will proceed with the comparatively conservative logarithmic utility function.

Consider a hypothetical mini-society with only two people, one on $20,000 and one on $150,000. In this example, we will introducing a tax of 0.1% on the citizen on $150,000. Unfortunately the tax is extremely inefficient, with deadweight loss of $1 for every $1 raised in revenue. For comparison, this is about 60% more distorting than the average $1 raised from stamp duty in NSW. (For what it’s worth I don’t buy the stamp duty deadweight loss numbers; the mainstream modelling of stamp duty seems fundamentally muddled as Cameron Murray points out.)

For the hypothetical tax introduced, assume the deadweight loss burden is completely worn by the person paying the tax.


The tax raises $150 in revenue, but reduces the payer’s income by $300, implying a large relative loss to society’s GDP.

The revenue is distributed in equal lump sum payments to each citizen. The poor citizen benefits by $75, but the rich citizen loses three times as much.

The unit for measuring wellbeing here (sometimes called “utils”) has no absolute meaning. For any particular value of utility, all that matters is the sign (positive or negative) and the value relative to other utility values.

Despite the significant loss in economic activity, the benefit to the poorer citizen’s wellbeing is 2.5 times the harm to the wealthier citizen. Overall, a significant economic benefit! In this context, even a tax with huge deadweight loss ($1 for every $1 raised!) is beneficial to society thanks to the power of redistribution.

Let’s consider a less extreme case.

Now we examine the imposition of a tax that results in an average tax rate of 10% for someone on $110,000 and 2% for someone on $40,000. The deadweight loss will be $0.50 for per tax dollar. In this case, the tax payer bears 75% of the impact of the deadweight loss relating to the tax they pay. The remaining 25% spills over onto the other tax payer.

The revenue is equally distributed between citizens.

This tax-transfer regime reduces income by $9,325 for the citizen on $110,000, while increasing it by $3,425 for the citizen on $40,000.

Again, a negative impact on economic activity and yet – still – a net positive impact on society.

Of course redistribution through cash payments is only one reason we raise revenue. We also do it to spend money on public projects such as infrastructure, education or health. If the project has a benefit-cost ratio above (below) one, this will increase (reduce) the net benefits relative to spending the money on cash transfers.

To illustrate the economics of project spending, we return to our initial mini-society above, which has personal incomes of $150,000 and $20,000. Recall that revenue is collected in an  extremely inefficient way, with a deadweight loss of 100%.


But this time the spending will also be extremely inefficient, going to a project with a benefit-cost ratio of 0.5, prior to distributional weighting. This means that every dollar of spending on the project destroys economic activity by twice as much. Terrible, right?

And yet… it still improves society!

It should be restated that the diminishing marginal utility assumptions here are relatively conservative, and the analysis doesn’t consider other reasons for redistribution.

The implication of these simple results is that we cannot say with confidence whether any particular item of tax or expenditure is “good” or “bad” for social welfare abstracted from its distributional impact. And as we saw yesterday, even “regressive” taxation and spending may still deliver pro-poor redistribution.

Given the tax transfer system is generally highly redistributive,  it is very hard for people who care about equality and social welfare to make confident normative judgements about any given policy based on summary information abstracted out from the role the policy plays in achieving system-level redistribution. In general, analyst priors should be heavily weighted in favour of more spending and revenue initiatives, even in the face of potential high costs to efficiency.

The first principle of social democratic tax design is MOAR

When left-leaning or centrist people talk about taxation they often stress tax progressivity, but a progressive tax schedule is not sufficient (or even necessary) to meaningfully reduce inequality. For that you need to raise a lot of tax.

This is easy to illustrate. Say a person on $300,000 pays a hundred times more tax than a person on $30,000 – a highly progressive tax schedule. But if average tax rates are low – say they only pay $1,500 and $15 respectively – there is never going to be much redistribution.

Let’s imagine a mini-society consisting only of the two people above, with tax raised distributed in equal lump sum payments to each. We see the following:

table 1

Virtually no redistribution.

Now let’s examine a flat tax, but at a much higher rate:

table 2

Clearly society becomes much more equal with the flat but high tax.

Now let’s look at a higher rate again, but with – gasp – regressive tax:

table 3

Even more redistribution!

Now let’s look at an extreme: not only will the tax become even more regressive, spending will be targeted at the wealthy (sometimes called “regressive spending”):

table 4

And… there’s even more redistribution to the poor! Taxation and spending can both be regressive, yet still produce substantial redistribution from rich to poor. I think that simple result would be counter-intuitive to a lot of people in the policy community.

How is it possible? We need to think about redistribution in a relative, not absolute, way. To reduce inequality, government policy does not need to be pro-poor. All it needs to be is less pro-rich than the market. That is not particularly hard, as the market is very, very pro-rich.

While this is a simple – even banal – point, I believe it is counter-intuitive to a lot of thinking in the anglophone policy community. Its recognition should influence how we evaluate and prioritise policy from the perspective of reducing inequality. My sense is that many policy analysts and advocates who want a more equal society overrate the importance of progressivity relative to sheer bigness of redistribution.

Real world successes in tackling inequality seem to support my point. The most redistributive and equal countries aren’t those with particularly progressive taxation or spending. It’s closer to the opposite.

According to OECD analysis in a 2008 Report Growing Unequal, the world’s most progressive taxing nation is the USA (discussed by Peter Whiteford here) – a notoriously unequal country not known for its socialism.

Australia is up near the top of tax progressivity too. In addition, Australia has the most progressively targeted cash welfare spending. And yet it is not near the top of redistribution. Social democratic leaders such as Denmark, Sweden and Norway have close to the least progressive taxation in the OECD. With universalist welfare systems, their spending isn’t particularly targeted at the poor either. What matters, however, is that they do a lot of it.

To restate my theoretical postulate, the first – and overriding principle – of social democratic tax design is MOAR.

The upshot of MOAR is the critical importance of allocating political and policy-making effort into identifying and promoting relatively tolerated taxes. For a social democrat, it should surpass almost everything else.


1) The argument that the size of the taxation and transfer system are major variables contributing to the redistribution is a simple mathematical truth. I think this is often under-appreciated. Size is crucial for redistribution.

2) The argument that size is virtually more important than anything else is a normative assertion, relating to social democratic strategy, influenced by some empirical political economy. Some support for it comes from the fact that countries with tax-transfer systems that are not particularly progressive – Finland, Belgium, Denmark, France for example – tend to sit up near the top of world for tax-transfer system redistribution, whereas countries like Australia (the most progressive payments) and USA (the most progressive taxation) are below OECD average for redistribution. This should not imply that any given lower taxing nation cannot be more redistributive than a higher taxing nation (providing it raises enough tax).




Big picture thinking needed to tackle poverty

My comments on the Greens’ Jobs and Income Guarantee. I had intended to provide comments for this piece by Asher Moses but couldn’t make the deadline.

The Greens are mostly right in spirit on this one. We need bold big picture thinking to tackle Australia’s entrenched poverty, unemployment and under-performing economy. They should be commended for shifting the needle of public debate.

In the context of pandemic unemployment, getting people back into work is obviously a critical priority. The Greens have been a bit vague on their details, but employment should be supported through a mix of conventional Keynesian pump priming, additional public employment and active labour market programs that support private sector employment through training and wage subsidies. Although conventional public sector employment would be ideal, this could be supplemented by the sort of program that is often described as a “job guarantee” (my variant here).

These types of measures are affordable. Australian governments’ net debt in 2020 is only around 35 per cent of GDP – the G20 average is over 100 per cent. In the long term, such programs can be funded with additional revenue from sensible tax reform. Australia under-performs on taxation, below average for advanced economies, and nearly 40 per cent lower than other rich countries like France, Belgium and Denmark.

Hopefully the Greens’ proposals will instigate a public conversation about poverty and unemployment. Poor and unemployed people suffer a great deal of stigma based on the belief they are slackers. The present crisis is the perfect time to highlight that poverty is driven by structural forces outside of any individual’s control. Even at the best of times, the brutal truth is that the government deliberately enforces a certain level of unemployment via macro-economic policy frameworks, as a safeguard against economic overheating and runaway inflation.

The Greens, to their credit, have turned the tables on accountability for unemployment; rather than blaming individuals for not finding jobs that do not exist, the government is held to account for failing to build an economy that provides sufficient jobs.

The Greens are right to combine a jobs focus with an income focus. There is a commitment to guaranteeing minimum standards of income even to those who do not get a job. It might not be politically correct to say this, but a full-time job will never be the right solution for everyone, including for working age people.

But for this same reason I am a bit concerned about rhetoric that presents jobs as “guaranteed”. It’s important not to oversell. If the public become convinced that a program literally guarantees everyone a job, it will become harder to politically sustain welfare programs for those who for whatever reason are not successfully matched up with a suitable job. Ironically, a so-called job guarantee could intensify the blame heaped on those who remain unemployed, as it implies that any residual unemployment is purely a matter of personal choice.

As someone who grew up in a family and location marked by persistent unemployment, I am sceptical that government will ever completely eradicate it. People have all kinds of complex things going on in their life – including caring obligations, disabilities, illnesses, traumas. The notion that a real-world bureaucracy can somehow find or create a job specially aligned for each individual’s complex needs and distinctive capabilities seems fanciful to me. It is not the way I understand bureaucracies to operate.

The only fail-safe protection against poverty is direct income transfer – and as such leftists and progressives should unapologetically put this front and centre of any anti-poverty strategy. In the long-term, this implies introducing some form of basic income – such as a guaranteed minimum income or participation income in which people are paid for making a societal contribution, whether that is job search, caring or volunteering.

Unfortunately the Greens have made their more comprehensive commitments available only to under 30s as part of the “Next Gen Guarantee”. Commitments to over 30s are vague at best. This is very strange and totally unnecessary. Poverty is poverty, regardless of age.

Unlike the major parties, Greens do not actually have to implement their policies so they do not face real resource constraints. It is therefore likely that the decision to exclude over 30s from the full deal is deliberate and strategic symbolism, rather than a difficult trade off. I believe it’s an attempt to play a kind of generational identity politics.

While this may be appealing to their voter base, it undermines the type of broad social solidarity required to introduce and sustain a generous social democratic welfare state.

Nevertheless, the Jobs and Income Guarantee is to be welcomed for bringing bold initiatives on poverty and inequality to the forefront of public conversation.

Career change and the quest for autonomy

I have decided to change careers – from public policy to psychology. The first step involves dropping my long-suffering Masters studies in economics and public policy (yes two programs, one of which I’m three quarters through), and entering a one-year Bachelor of Psychological Studies. In doing this, I’m moving on from about a decade as a public servant and political staffer.

In studying psychology my aspiration is to go into private practice, as a clinical or general psychologist. The road ahead is long and challenging, and nothing is given. In the case of clinical psychology, it will require at least four years of full-time study, and two years as a registrar. I will need to gain competitive admission to honours and then Masters, which often requires first class grades. Only a minority of students wind up with one of the prized places.

Nevertheless I’m up for the challenge.


Considerations in my decision included:

Favour economics/public policy Favour psychology
Strongly favour:

  • Considerable existing human capital
  • Risk and uncertainty of change (may not succeed in psychology, or may not find it as fulfilling as hoped)


Weakly favour:

  • Passion (but passion can be a curse…)


Strongly favour:

  • the superior autonomy of professionals in private practice
  • motivational benefits of helping people directly
  • a clearly defined career path
  • verbal-centric work (I prefer talking to writing)


Weakly favour:

  • a career directly consistent with my ethical values
  • personal attributes (can be empathetic while staying analytical and somewhat detached)
  • a career with strong demand across Australia and into the future.


In addition to the above, I have a preference for a field adjacent to social policy and a discipline that mixes qualitative and quantitative approaches. These aspects are roughly neutral between psychology and the type of public policy I do.

Public service employment destroys autonomy

The single most important factor in my decision was autonomy. Most people want some level of autonomy, but for some of us it is almost sacred. I think this is a pretty deep issue of personality, and as such is relatively resistant to change.

There are few careers that can deliver the independence afforded to a professional in private practice. A tenured academic is one, but academic employment in Australia has turned into a dystopia of casualisation and labour over-supply. Tenure – the teeth of academic autonomy – has largely been dismantled.

The main career option in public policy is the public service. This is perhaps the worst possible career for autonomy, with legislation uniquely designed to sacrifice public servants’ freedom not just at work but in their private life. In the federal public service, public servants can’t count on their freedom to even like Facebook posts critical of the government. In this respect having a passion in the field (politics, policy) can actually be a negative: it is a bigger loss to lose the ability to speak out about something if you really care about it.

The Australian government’s speech restrictions err on the side of excess risk aversion, curbing public servants’ citizen rights unnecessarily, and hollowing out organisations by pushing away people with valuable divergent thinking skills. Organisations need a diversity of cognitive styles and personalities, including critical thinkers and sceptics.

In addition to loosening restrictions on public servants, the government should pass legislation guaranteeing greater speech rights to private sector employees. The trend towards codes of conduct with vague catch-all restrictions is a serious risk to freedom of expression in practice. The open exchange of ideas, facilitating intellectual and political deliberation and progress, is a public good underprovided by a free market. It should be protected and promoted by government.

But I digress.

On a more mundane level, being a public servant, even in a flagship division of a ‘prestigious’ department, involves doing a lot of bullshit. By that I mean tedious clerical work, destined to be ignored on a staffer’s desk.

(Disclaimers: public servants do important work; many people have a wonderful time in the public service; I found it very rewarding for several years; I have been that staffer.)

Other options

Another public policy option is being a political staffer again, but you can’t do this all your life; most people tire of the hours, stress and insecurity; particularly when they’re in their mid-30s, with kids. Like the public service, this work also requires giving up much of one’s individual autonomy outside work. (Although you get more autonomy on the job, and you can help fight for a cause you deeply believe in.)

Working for a social democratic think tank would be attractive. However, there’s not much going in this space in Sydney, where I’m effectively tied to over the medium to long term for family reasons. I’m not even sure there are enough organisations to allow someone to sustain a career in this field long-term anywhere in Australia.

The corporate world holds little appeal to me, and I don’t think I’d be motivated to do it day-in, day-out.

Journalism seems too risky and structure-less.

I considered ventures like a Nicholas Gruen-style policy consulting firm or even a Matt Bruenig-style crowd sourced think tank. But Gruen is a force of nature, and I doubt I have the intense focus and ongoing dedication to drumming up business required to make a consultancy successful. I will never have Bruenig’s audience gathering ability, and even if I did there would be no way for the model to be downscaled for the Australian market.

Working for an NGO is something I haven’t really tested, but it’s probably not particularly financially lucrative and I suspect I might get bored over the longer term.

Towards psychology

For a few years now I’ve been wandering from work project to work project, uni subject to uni subject, without feeling clarity on what it was all aiming towards. I’ve notched up some reasonable achievements during this time, and I’m proud of them, but I’ve gradually lost motivation not knowing what all the work was adding up to in the long run.

Psychology provides an opportunity for a clearly defined career path. It involves helping people, which is intrinsically motivating, salient for me given family struggles in this area.

In addition, psychology is ethical on a day-to-day level. Public policy work may be ethical on a macro level (obviously public servants do a necessary job), but day-to-day it often involves colluding with and facilitating evil, for example spinning data to help a government disingenuously argue the case for cuts to the most vulnerable. I always found this work extremely dispiriting and demotivating – and even morally dubious, especially when it goes beyond following narrow directions and into territory where you’re pro-actively applying ingenuity to assist the government in its scheming. Watching robo-debt unfold heightened my ethical worries about working in the Australian Public Service.

What’s next

I’ll be living in Adelaide for most of this year. Flinders University has the fastest and cheapest program for my situation. I hope to be able to do honours at a university in Sydney next year.

I’m looking forward to a new phase in my life, a new experience, a new challenge.

Job guarantee errorism

It’s always a pleasure interacting online with the jobs guarantee (JG) and “modern monetary theory” (MMT) crowd. They must be the most passionate, devoted and fervent people on the internet, excluding ISIS.

Due to my concerns that a specific policy – the JG – could not be implemented as they envisage (see here, also here and here), they stopped by my Twitter yesterday and gave some constructive feedback.

Apparently, my worldview is “morally and intellectually impoverished”,  is “fundamentally neo-liberal (and wrong)”, “basically invalidates almost all progressive goals”, “10000% it’s an incredibly uninspiring view and at best a wet liberal, if not genuinely conservative”. I find this kind of deranged hyperbole to be one of the genuinely entertained aspects of Twitter.

Ironically, my views on economic justice largely derive from radical traditions. Inequality is, at root, a construct of legal and institutional fictions known as property rights. We are born into society that determines our command over resources based on arbitrary definitions of ownership. There is nothing natural or just about the prevailing system of property rights. On the contrary, it is fundamentally unjust because it catastrophically fails to allocate sufficiently to those of greatest need. This is not an issue of work, it preceeds work. Property rights must be redefined such that the poor receive a greater share of society’s resources – not because they work for it, but because the principles of economic justice require it.

Nearly 40% of the income in from the Australian economy is capital income, awarded for no work at all. The average adult in the Australian top 1% receives capital income that is more than triple the total average adult income from all sources. Given these facts, how is the central issue of distributive justice in this country insufficient work? If income can be detached from work for the capitalists, it can be for the rest of us. The single most potent tool modern governments have for redefining ownership of resources requires no work at all: it is the tax and welfare system.

Leading JG gurus fetishise work because of a profoundly conservative impulse that it should be a requirement for survival (for the poor). It is not surprising to find they are overtly hostile to the welfare state and that they echo the rhetoric of right-wing welfare reformers. Minsky condemned the “welfare mess” and argued “one must earn one’s keep”. Bill Mitchell would like to “abandon the unemployment benefits scheme”.

If this is the language of the JG’s foremost advocates, is it any wonder why some of us worry that it would turn into glorified work for the dole?

In my model of social democracy, social security and safety net concepts are cleanly separated from labour requirements. This protects beneficiaries from coercion and exploitation. Unlike the JG advocates, I do not believe the centrepiece policy for economic justice should depend on work.

What would society look like under this evil conservative vision?

  • generous unemployment benefits with lenient conditions
  • active labour market programs
  • no work for the dole
  • mass unionisation and sectoral level bargaining
  • generous and universal family benefits
  • a universal social dividend to all citizens funded by a social wealth fund.

As I have argued, there is room to trial a transitional and optional job of last resort (JLR), which unlike the standard JG proposals would be sensibly modest and designed to mitigate against risks. It would not pretend to be a universal “guarantee”. If it is wildly successful, by all means, expand it. But don’t be so dogmatic and hubristic to assume a highly complex, bureaucratically intensive theory that screams risk will work just because you will it too, not when the most vulnerable in the community are at stake. As Daniel Patrick Moynihan rightly cautioned in his Senate speech against Clinton’s 1996 welfare reforms:

But first, do no harm. Do not pretend that you know what you do not know. Look at the beginnings of research and evaluation that say, “Very hard, not clear.” Do not hurt children on the basis of an unproven theory and untested hypothesis.

The JG folk don’t seem to comprehend the danger of committing to something – a “guarantee” – that they may not be able to deliver on, of the way a phony guarantee would stigmatise the welfare recipients who are unable to be matched with a suitable job. They don’t seem to understand how the concept of a guaranteed job could, in the actually existing Australian society, turn into a social expectation that the recipient will take a “guaranteed” job, something entirely unreasonable given the health and welfare profile of so many of the unemployed.

They don’t seem to understand how maybe, just maybe, allocating a complex range of individuals, each with highly specialised needs and capabilities, to an extremely narrow range of jobs possibilities, which must satisfy restrictive criteria such as not displacing other workers, could be beyond the power of actual human bureaucracy and that people might get hurt in this process. They don’t seem to appreciate the risk that these pseudo-jobs would in fact displace real jobs, despite best efforts, trapping people in dead end drudgery with limited opportunity for promotion.

Much of my concern about bureaucracy relates to my personal experience as a kid in a single parent family, with a mum who was often sick. Knowing the stigma attached to poor single mothers, I would have zero confidence in fairness from local committees either.

If concern for the most vulnerable in our community is “morally and intellectually impoverished”, “fundamentally neo-liberal” and “genuinely conservative”, I plead guilty.


Local bureaucracy is still bureaucracy

Last week I proposed a “job of last resort” (JLR), a pared-down variant on the so-called job guarantee (JG). In this post, I want to clarify some of the problems with a comprehensive JG in response to online discussion.

Under the JLR, eligible unemployed workers are able to apply for a special government job at minimum wage – similar to the JG in this respect. But the change in terminology – dropping the word “guarantee” – is deliberate and reflects the reality of the program. The JLR does not guarantee a suitable, or even any, job for all unemployed people.

I do not believe guaranteeing over 700,000 unemployed people a suitable job is something the bureaucracy can deliver on. I do not believe in promising something that cannot be delivered. A fake guarantee would only add to the stigma of unemployment. If the public is led to believe jobs are “guaranteed”, it implies that an unemployed person who does not succeed in obtaining a suitable JG job is consciously shirking.

Under the JLR, the pool of eligible workers is narrow (only the very long-term unemployed or underutilised), and the job availability is also narrow: city and regional beautification (gardening, landscaping, etc) or, for the less fit, low skill tasks embedded in government departments that would otherwise not be performed (door greaters, so-called “tea ladies”, etc). It may be possible to expand this modestly from government employment to community employment, but the principle remains: if you don’t fit into one of the available categories, you’re not in luck. Also you can be dismissed like any regular employee. No job is “guaranteed”.

This pared-down model of reserve public employment is necessary to ensure the program is actually deliverable.

A comprehensive JG would need the capacity to evaluate over 700,000 unemployed (well over a million in recessions) and somehow find or create them a job, tailored for their individual circumstances. Unemployed people disproportionately face challenging life circumstances. They are substantially over-represented with physical and mental health conditions and disability, life traumas and complex family situations. Some have conditions that make them a risk to themselves or to others. Some have messed up circadian rhythms or neurological conditions that cause them to sleep throughout the day. Some have run away from violence; some have just left prison after perpetrating violence; some have done both. Somehow, a program will take this complex diversity and neatly slot them all into a job, and things will turn out fine. I just don’t buy it.

JG advocates try to handwave away the bureaucratic problem by declaring the program will be “federally funded, but locally administered”. But this just kicks the can down the road. Local bureaucracy is still bureaucracy. It still requires resources to evaluate workers, find or create suitable jobs, and monitor the employment relationship. Indeed local-level administration would be more resource intensive due to diseconomies of scale – the local level would lack the (relative) production line efficiency of a mass scale processing – and because Commonwealth level administration would be required in any case to oversee the funding agreements with, and performance of, local entities.

What would the local-level administration look like? JG advocates sometimes say the job allocation will be determined by a mix of local councils, democratically elected committees, local community groups and local businesses. To me this sounds like nightmare rule-by-committee, and no panacea for the JG’s inherent administration problems. Big government bureaucracy gets a bad rap, but local government, small business and local communities can be just as problematic, at times even more so. Local councils lack the media accountability of the Commonwealth government, and as such are not infrequently corrupt and incompetent.

When we romanticise the local, we downplay the profoundly repressive force of what Noah Smith calls “local bullies“:

…freedom of the individual can be curtailed not only by the government, but by a large variety of intermediate powers like work bosses, neighborhood associations, self-organized ethnic movements, organized religions, tough violent men, or social conventions. In a society such as ours, where the government maintains a nominal monopoly on the use of physical violence, there is plenty of room for people to be oppressed by such intermediate powers, whom I call “local bullies.”

Twitter personality James A. Robichaux poses the local bullies problem (without calling it such) in the context of a JG:

Job Guarantee advocates talk about how a local, democratically-elected council will determine which jobs are needed, including jobs suggested by citizens who want to do them.

I can imagine how that will go.

I think that it’s no wonder that all major JG proponents are white.

The cool kids and their sycophants will form a majority and vote themselves the coolest jobs.

The minority of uncool kids will have the bad jobs and won’t get their own job ideas approved.

[Edit: A commenter (reasonably) points out not all major JG proponents are white. To be fair to the original author, it was an offhand comment on Twitter. This specific empirical claim was not the purpose of using the quote, and I think the broader argument stands up.]

To the extent local-level decision makers have discretion, the more appealing jobs may be allocated to people who model stereotypes of the “deserving” poor or other prejudices prevalent in a particular community. It is not hard to imagine conservative communities administering welfare very differently to left-leaning communities. It is not hard to imagine a conservative local-level bureaucrat thinking, for example, that a trans person is partly responsible for their unemployment by “choosing” a “lifestyle” outside the mainstream. It’s not hard to imagine a local official turning a blind eye to job displacement issues to provide free labour to an organisation they have associations with.

Local biases are a real problem in the context of a national welfare state that should be administered according to the rule of law, ensuring fair and equitable treating for citizens regardless of where they live. Even when not overtly pernicious, life changing decisions being made on ad hoc local whims are hard to justify when they are funded by national taxpayers.

Remember the Commonwealth Government’s Building the Education Revolution program. Some pricey shadecloth contracts administered mainly by the NSW Government embroiled the Commonwealth Government in years of damaging controversy. Yet we think hundreds of local bodies can make hundreds of thousands of life changing decisions, using Commonwealth funds, and everything’s going to be fine?

Spending decisions on a nationally funded program must have accountability to Australian citizens. The Commonwealth would therefore need to negotiate detailed inter-governmental agreements with 547 local councils (or authorities at a similar level) stipulating rules, processes and principles for the equitable administration of the funding, including in evaluating job seekers and avoiding job displacement. It would need to monitor these agreements. In addition, it should establish a national appeals process.

Under a “federally funded, locally administered model”, all of the complexities associated with evaluating workers and placing them in jobs would remain, but the intergovernmental component would create an additional layer of bureaucracy. All of this is so unwieldy it is unlikely to achieve any of its goals particularly effectively.