Patricia Pino has written a thoughtful and detailed response to my recent posts on modern monetary theory (MMT) and the Job Guarantee (JG), which I shared this morning. I welcome that. These are issues that deserve serious engagement, not sloganising. And if there is one thing that matters in political economy right now, it is that we are willing to interrogate our own assumptions as rigorously as we challenge those of others.

What follows is my response.

Patricia's core claim is straightforward. She has suggested that the Job Guarantee is not an optional add-on to Modern Monetary Theory, but a logical necessity arising from it. The argument is that if the state creates the obligation to pay tax in its own currency, it must also guarantee access to the means of earning that currency. Otherwise, it is said, the state is imposing an obligation without providing the means to meet it.

That is an elegant argument. It is also, in my opinion, fundamentally flawed.

First, it assumes that paid employment is the only legitimate means for people to access the currency needed to settle tax obligations. That is not true. There is an inherent problem in this claim: it implies people must work, which is clearly not always possible, and the assumption has inherent within it an implication of compulsion which those promoting it deny, but which I think would be embraced by anyone who might propose it politically. There is also a conflict in here. The implicit assumption is that money is created by work, which must be created to let taxes be paid. MMT makes it quite clear that this is wrong: money is created by government spending, and taxes are only required to recover it. Patricia's argument for the JG appears to contradict this core MMT understanding. 

Second, the state already provides income through multiple channels other than work, including pensions, social security, and various kinds of transfers, including interest payments. There is no inherent requirement within a fiat monetary system that income must be mediated through a job. To suggest otherwise is to conflate monetary design with labour market structure. 

Third, the JG presumes that unemployment is a deliberate policy tool embedded within macroeconomic management. That may be true in an economy embracing NAIRU (Non-Accelerating Inflation Rate of Unemployment) orthodoxy. It is not a necessary feature of a well-managed economy. The objective of any proper fiscal policy (whether embracing MMT or not) should be to eliminate involuntary unemployment through appropriate fiscal design. That is not the same thing as creating a permanent institutional mechanism to absorb it, when a proper fiscal policy should ensure it does not exist, which a proper understanding of MMT should facilitate.

Fourth, it assumes that the only credible way to secure full employment is via a standing labour buffer administered by the state, which is what the JG would create. That is precisely the point at issue. The JG cannot be assumed as a premise in order to prove itself as a conclusion.

So I do not accept that the Job Guarantee is intrinsic to MMT. It is a policy preference. It may be a sincerely held one. But it is not logically unavoidable.

Let me also make this key point. In my opinion, and as a matter of fact, a government that did not subscribe to the MMT view of how money works within an economy could, nonetheless, decide to offer a job guarantee as part of its social security programme, as a consequence seeking to maximise employment, reduce benefits paid, and increase its tax revenues, whilst maintaining a workforce that does not suffer the damaging psychological consequences of unemployment. It would, in fact, be an entirely reasonable thing for any government that cares about the population that it serves to do this if it thought that its own fiscal policies could not deliver full employment, although I would suggest that it should revise those fiscal policies instead. I think that the fact that this possibility exists is unarguable. As a result, it necessarily follows that there is no intrinsic link between MMT and the job guarantee; no other conclusion is possible. 

I would also stress that a job guarantee delivered in this way by a government that did not subscribe to the ideas implicit in MMT on fiscal policy management could be accommodated within a neoliberal framework that embraced fiscal rules, existing monetary policy management systems, and similar approaches. Again, this has to be true, because those frameworks can at present accommodate payments using a social security system to maintain those who are made unemployed through no fault of their own within the existing economy and, as the proponents of the job guarantee within MMT make clear, inclusion in that programme would always be voluntary, would be an alternative to the payment of benefits, and would therefore necessarily not displace the existing social security system, which is going to be retained as a result within any economic system.

My point, then, is simple: MMT does what it says on the tin. It describes how money works in a modern economy; that is it. Let us not pretend otherwise. Winning the battle to achieve credibility for the facts around that system is a big enough task for us to take on. Latching onto this claim, something which is a policy proposal that it makes possible, but which, as I have demonstrated, is not in any way integral to it, and which could be adopted and implemented by a government, makes the likelihood of MMT adoption very much harder, to the point of impossibility.

Let me then turn to the specific claims made in its defence.

On administration, it is argued that complexity is not a decisive objection. We run the NHS. We run education. Therefore, we can run a Job Guarantee. That analogy is superficially appealing. It is also misleading. The NHS delivers defined services within professional standards, with clear objectives and established institutional structures. The Job Guarantee, by contrast, requires the continuous creation of new work, necessarily tailored to individuals  (or they would not undertake it unless compelled to do so) whose circumstances are constantly changing, in ways that must avoid displacement of existing employment, while also being meaningful, productive, and socially valuable.

That is not the same type of problem. It is a qualitatively different one. The issue is not simply scale. It is the nature of the task. Designing and sustaining an ever-evolving set of roles for potentially millions of people is not comparable to delivering healthcare or education. It is an open-ended, indeterminate administrative challenge, and that matters.

Next, regarding inflation, it is suggested that the Job Guarantee provides a nominal anchor through a fixed wage floor, which is already provided by the minimum wage. Again, that is an elegant idea. But it assumes that labour markets behave in a way that allows such an anchor to operate effectively across a highly differentiated economy when the modern economy is not a single labour market. It is a complex set of interlocking markets, with different skill levels, bargaining structures, sectoral dynamics, and constraints. A single wage floor set by a JG could not, in practice, anchor this system as some MMT theory suggests. And, as Patricia acknowledges, much inflation (such as that now coming our way) arises from factors entirely unrelated to wages, such as energy price and supply shocks, market power, and geopolitical disruption. A job guarantee does not address those. Nor does it provide a credible mechanism for managing them.

On stigma and labour market position, it is argued that being in work, any work, is better than being unemployed. That might be true. But it is not the relevant comparison. The question is not whether a job guarantee is better than unemployment. It is whether it is the best way to eliminate unemployment in the first place. If we accept a Job Guarantee as a permanent feature of the economy, we are, in effect, institutionalising a secondary labour market. We are saying that there will always be a group of people whose employment is not secured through the primary dynamics of the economy, but through a fallback scheme. That has consequences. It risks entrenching a hierarchy of work, with all the social and economic implications that follow, which I have previously discussed.

On fiscal policy, Patricia suggests that investment and industrial strategy cannot guarantee that every individual will be matched to employment. That is correct. But it, again, misses the point. The objective of fiscal policy is not to match every individual to a specific job at every moment in time. It is to create the conditions in which sufficient demand exists to make full employment possible. That requires sustained investment, active industrial policy, regional development, skills provision, and social support. It requires the state to take responsibility for the level and composition of economic activity. What it does not require is the creation of a permanent, parallel employment system to mop up the consequences of failure elsewhere.

And that brings me to what I think is the real point of disagreement. It is suggested that my approach is “project-led” whilst the Job Guarantee is “person-led”. I do not accept that characterisation.

My approach begins with people. It begins with the recognition that everyone who wants to work should be able to do so in a job that is properly paid, secure, and socially valued. It then asks what economic structures are required to make that possible.

The Job Guarantee begins with a different premise. It assumes that the economy will not, and perhaps cannot, deliver that outcome. It therefore creates a residual mechanism to ensure that no one is left entirely excluded. That is a fundamentally different ambition.

Finally, there is the question of meaning, and here my correspondent invokes Viktor Frankl, as I had done previously. This matters because work is not just about income. It is about identity, recognition, purpose, and, most of all, meaning.

The claim is that the Job Guarantee can confer legitimacy on socially useful activity that is currently unrecognised. That may be true in some cases. But we should be cautious. Meaning in work is not created simply by institutional recognition. It is shaped by social context, by relationships, by the perception of value, and by the integration of that work into the wider economy. There is a risk that work created within a job guarantee framework might be seen, by participants and others, as something different. As something less. That is not inevitable. But nor can it be dismissed as I think the probability is high.

So where does this leave us? Let me suggest three conclusions.

First, the Job Guarantee is not a necessary implication of MMT. It is one possible policy response to unemployment, but it is not the only one.

Second, the case for it rests on assumptions about administration, inflation, and labour market dynamics that are, at best, unproven at scale in an economy like the UK's and which are, at worst, problematic.

Third, and most importantly, it risks lowering our ambition. Instead of asking how we create an economy that delivers full, meaningful employment for all, it asks how we manage the failure to do so. I think we should aim higher than that. The real task is not to design a better buffer stock of labour. It is to eliminate the need for one. That means using the full capacity of the state, across fiscal and monetary policy and institutional power, to create an economy that works for everyone. It means investment on a scale we have not yet contemplated. It means rethinking the state's role in shaping markets, and not merely correcting them.

And it means being clear about this: unemployment is not an inevitable feature of a modern economy. It is a policy choice. We should choose full employment.

NOTE: This article was modified after publication, with sentences being added to the third paragraph of the non-italicised section

PDF of article