Would a Universal Basic Income Advance Republican Liberty? Understanding the Impact and Implications of a UBI on the Exit Costs Associated with Unemployment

1 Introduction

Would a universal basic income (UBI) advance republican liberty? The potential threat to non-domination posed by administrative bureaucracies has led some republican theorists to argue in favour of UBI-schemes that would significantly reduce or eliminate people’s dependence on conditional or means-tested government benefits when in need. A sufficiently generous UBI, its proponents argue, would enable people to be financially independent of employers, spouses, family members or others on whose arbitrary wills they might otherwise be dependent, while also ensuring that people are not forced to depend on the discretionary and potentially arbitrary judgements of government bureaucrats when in need financial assistance (Lovett 2010; Pettit 2008; Taylor 2017; White 2020).

The strength of the republican case for UBI has been the subject of a robust debate in the academic literature (Birnbaum and De Wispelaere 2016; Bogg 2017; Casassas 2008; Casassas and De Wispelaere 2016; Dagger 2006; Gourevitch 2016; Gourevitch and Stanczyk 2018; Jubb 2008; Lovett 2010; 2022; Pettit 2008; Taylor 2017; 2018; White 2008; 2020). Though a UBI has appealed to many republican theorists, some critics have questioned whether it would be economically feasible to provide a UBI that was sufficiently generous to give people the kind of “strong exit power” its proponents hope for (see Taylor 2018), while others have argued that a UBI would do little to tackle the impediments to exit power posed by limited opportunities to high quality jobs, or the weak bargaining power workers may enjoy vis-à-vis their employers (see White 2020, for a summary and critical assessment of the central arguments in the debate).

The matter is made more complicated by the fact that the implications of a UBI for republican liberty would depend on whether a UBI replaces or merely complements conditional and targeted benefits–such as an income-based unemployment insurance. This would matter not least when considering the effect of a UBI on the exit costs associated with leaving or being fired from a place of employment. For a UBI to have the effect of freeing people from domination at the hands of government bureaucrats, the UBI would have to be generous enough to at least ensure that people are not (too) financially dependent on being able to access conditional or targeted benefits if they become unemployed or for some other reason are deprived of an outside source income. Yet this need not imply that a more modest UBI could not advance republican liberty in other domains–with respect to workplace domination or domination within the family, for instance. It is thus far from easy to get a clear picture of the implications of UBI for republican liberty, as they will depend not only on the generosity of the UBI but on whether it replaces (wholly or in part) other conditional or targeted benefits. Different policy mixes are also likely to affect different groups in different ways–depending on their employment status, income level, etc.

This paper attempts to advance our understanding of the implications and potential trade-offs of a UBI as it relates to non-domination by developing a more fully theorised account of how a UBI would affect the exit costs associated with unemployment–under different conditions (that is to say, with respect to the policy mix and the broader institutional context) and with respect to different domains of concerns (i.e. domination in the workplace, within the family, and in the relationship to the state).

To develop this account, it begins by asking how replacing conditional and targeted benefits with a UBI would affect the perceived exit costs associated with unemployment. Though many proponents of UBI envision the UBI complementing rather than replacing things like unemployment insurance, asking this question helps clarify the implications for republican liberty of moving from a system designed around income-maintenance towards a baseline-provision model. Indeed, it gives us reason to question what appears to be a common assumption among many republican proponents and critics of UBI, which is that if a UBI could be put in place that was sufficiently generous to enable people to live off it without facing poverty, this would provide people with a measure of financial independence sufficient to substantially reduce workers’ vulnerability to domination at the hands of their employers (even if there are non-financial sources of vulnerability to still worry about). This assumption is based, I suggest, on a faulty picture of how people assess changes in their circumstances, which does not account for the ways in which, and the reasons that, people are loss averse. Asking how replacing conditional and targeted benefits with a UBI would affect the perceived exit costs associated with unemployment also helps us think through how perceived exit costs are assessed under conditions of uncertainty and limited information. This paper shows that once we properly account for how perceived exit costs would be affected by (1) the size of the drop in month-to-month income associated with unemployment under a UBI-only scheme, and (2) the level of certainty/uncertainty regarding access to benefits different systems provide, we can get a better grasp of some of the non-domination related trade-offs that would be at stake in a decision between a more traditional income-maintenance model of welfare provision and a baseline provision model built around a UBI. It identifies and outlines three of these trade-offs, and sketches an approach for how to think about these from a non-domination perspective. Drawing on this approach, it also tries to say something about how to theorise the implications, from a non-domination perspective, of a modest UBI existing alongside a generous system of conditional and targeted benefits.

The next section begins by clarifying the meaning of non-domination before moving on to discuss the republican case for UBI. The subsequent sections develop and defend the central claims of the paper: its account of how reliance on a UBI rather than unemployment insurance would affect the perceived exit costs associated with unemployment, and what this means for how to think about the trade-offs and implications of a UBI from a non-domination perspective.

2 Non-Domination & The Case for UBI

Much work has been done clarifying and explicating the meaning of non-domination (Lovett 2010, 2022; Lovett and Pettit 2009; Pettit 1997, 2001, 2014; Shapiro 1999, 2012; Taylor 2013, 2017). Non-domination captures, or draws on, the insight that freedom requires a kind of relational equality. Lovett defines freedom as non-domination “as a sort of structural independence–as the condition of not being subject to the arbitrary or uncontrolled power of a master” (Lovett 2018). The classic example used to illustrate the meaning of freedom as non-domination is the master-slave relationship, but one can equally well focus on the relationship between the state and its citizens, an employee and an employer, or a husband and a wife. The key insight is that freedom requires not merely the absence of interference by others–for indeed, a permissive slave-master could grant their slave permission to do various things–it requires a kind of relational equality. It is only when people’s well-being is not (too) dependent on the arbitrary good will of another person or organisation that they can be said to enjoy freedom in this non-domination sense.

In broad terms, republican theorists have emphasised the important role social insurance schemes can play in freeing people from domination at the hands of employers, spouses, and parents, as well as the good will of private charity providers, while at the same time worrying about and in various ways criticising the ways in which the administrative bureaucracies that are necessary to distribute targeted, state-provided, welfare benefits can undermine people’s freedom as non-domination. Social insurance schemes help free people from domination by (or to the extent that) they help lower the exit costs people face if and when they leave (or are forced to leave) a place of employment or a relationship with a spouse or family members on whom they were financially dependent. By helping ensure that people can exit these relationships without suffering too much financial hardship, social insurance schemes make people less dependent financially on the other party, removing or reducing a power imbalance that undermines freedom as non-domination. However, while they alleviate one source of domination, they also open up the possibility of another. To receive conditional or means-tested benefits generally requires demonstrating to one or several administrative bureaucrats that one is eligible and meet all the required criteria to receive a particular benefit–that one is unemployed, that one is actively looking for work, etc., depending on the benefit and the rules surrounding it. Though the people that administer these programs are bound by law and procedure, they also have discretionary power and are not always constrained by well-functioning systems of oversight and accountability. The prospect of welfare provision schemes that remove or reduce the threat to non-domination posed by the administrate welfare state have thus appealed to republican theorists.

With a UBI, the state would retain its role as benefit provider, but it would distribute a “basic income” to everyone who meets some basic criteria (like being a resident or a citizen) without any means-testing – everyone eligible would simply get the same size benefit, regardless of their income or wealth, and regardless of whether they are working or looking for work or not. Though UBI proposals differ, Bidadanure (2019) identifies five definitional features that tend to be shared: (1) the benefit is given in cash rather than in-kind, (2) it is given to individuals rather than households, (3) it is given unconditionally (rather than conditionally), (4) it universal rather than means-tested, and (5) it is provided regularly rather than as a one-off payment.^[1] Whether and in what ways a UBI could contribute to furthering non-domination depends on the generosity of the UBI and whether it replaces or merely supplements the traditional, more conditional and targeted forms of welfare provision.^[2]

There are several components to the republican case for replacing conditional and targeted benefits with a UBI. The first, already mentioned, is that it would remove the need for hard-to-determine and discretionary judgements regarding eligibility–like determining whether someone is or has been actively looking for work, or whether someone was fired from or quit voluntarily from their past job–and thus reduce benefit applicants’ and recipients’ dependence on the (potentially arbitrary) good will of public bureaucrats.

A second component of the case for a UBI is that by removing the need to apply for benefits, a source of uncertainty regarding the true exit costs of leaving a relationship–with an employer, for instance–is removed. This is good, from a non-domination standpoint, because it means that people need not factor in that they might not receive unemployment benefits should they lose their job (because they are deemed ineligible for some reason). People can count on the UBI to provide them with money, and thus feel secure that the potential worst-case scenario of being unemployed and unable to receive government assistance will not come to pass.

A third component is that the UBI would go to individuals and not households (this is true, at least, for most UBI designs). This is beneficial, from a non-domination standpoint, since it makes the individual members of a family or a household more financially independent of each other.^[3] Though this is not a feature unique to a UBI (it is a feature of many benefit programs used in modern welfare states), one of the potential benefits of a UBI is that it would entrench and make universal this program design. Ideally, at least for some republican theorists (see for instance Lovett 2010), the UBI would also be enshrined in and protected through the constitution. This would help ensure its independence, and would remove (or at least lessen) the threat to people’s financial independence posed by the whims of shifting political majorities.

With respect to the other side of the argument–objections to the republican case for UBI–White (2020) distinguishes between four criticisms that can be found in the literature, which are worth outlining and discussing here: (1) “irrelevance of incomplete exit power”, (2) “even strong exit power will have limited impact on job quality”, (3) “high basic income will not confer strong exit power anyway”, and (4) “the employment contract is inherently dominating” (White 2020, 312–317). The first of these–“irrelevance of incomplete exit power”–boils down to the argument (already mentioned) that in order for basic income to effectively advance republican liberty, it would need to be generous enough to enable people to exit the labour market. A small basic income, the argument goes, would achieve little. Criticisms (2) and (4) – “even strong exit power will have limited impact on job quality” and “the employment contract is inherently dominating”–take aim at the perceived limited effectiveness of even a high basic income when it comes to improving workplace conditions and the structural relationship between workers and employers, at least under certain regulatory regimes. If employees still have absolute discretion over workplace conditions and the terms of employment, workers will continue to be subject to poor workplace conditions and will be dependent on the arbitrary good will of their employers for their well-being in the workplace. A basic income might advance relationships of domination in some respects, in other words, but it is no panacea.

Among the four lines of criticisms White identifies in the literature, it is criticism (3)–“high basic income will not confer strong exit power anyway”–that is most directly aimed at the question that is at the heart of this paper, concerning the effect that a UBI would have on the exit costs associated with unemployment. This argument has been most forcefully developed by Birnbaum and De Wispelaere (2021). The thrust of their case is that people’s concerns related to unemployment are broader than the proponents of UBI seem to assume. People are concerned about losing non-wage related employment benefits–like health insurance, in some countries, or the status a job can afford–and may fear the scarring effects of unemployment or the impact it may have on their future pension. These fears and concerns mean that the exit costs people associate with unemployment will remain high even if they have access to a high basic income.^[4]

This line of criticism captures something important, which is that we need to consider what unemployment would mean from the perspective of the people faced with the prospect of being without a job when we estimate the exit costs associated with unemployment. And they are right, I would also argue, that concerns about losing non-wage related benefits factor into perceived exit costs. However, the example of health insurance illustrates a weakness in this objection when formulated along the lines described above. As White (2020) points out, a proponent of basic income could merely argue that the problem with employer-provided health insurance raising the exit costs associated with unemployment is best addressed by ensuring that people have access to quality health insurance (or easy access to the health care system) independent of their place of employment or employment status. Similarly, rules concerning pension contributions or the levels at which pensions are set could be designed to minimise the impact of (at least brief) periods of unemployment. Of course, not all wage-related benefits associated with employment could be provided through other channels–concerns related to the loss of status, for example, would be hard to do away with. But what White’s argument suggests is that, if we can get the broader social policy context configured in the right way, a high basic income would contribute to strong exit power.

But there are other reasons for thinking that even a high UBI might not confer strong exit power, at least not for all workers, I want to suggest, which the next section will explore.

3 UBI & the Exit Costs Associated with Unemployment

To explore these reasons, it is worth beginning by asking how replacing a traditional UI with a UBI would affect the perceived exit costs associated with unemployment. While many proponents of a UBI envision it complementing rather than replacing UI (Haagh 2008, 2011, 2019; Martinelli and Vanderborght 2022; Van Parijs and Vanderborght 2017), the idea of having a UBI replace targeted and conditional benefits is not without appeal among republican theorists who are (particularly) concerned about the risk of domination posed by the administrative state (see Lovett 2010). And though there are those who are sceptical that a sufficiently generous UBI could be implemented–be it for political or economic reasons (see Taylor 2018)–many critics and proponents of the republican case for UBI seem to share the assumption that if a UBI could be made generous enough to enable people live off it without facing poverty, this would give people a measure of financial independence (“anti-power”) sufficient to remove at least the financial reasons employers are able to wield dominating power over employees. The exit costs associated with unemployment, in other words, are often assumed to be low in a world where people have access to a basic income that protects them from falling into poverty (at least if we ignore the potential costs of losing non-wage related benefits).

But it is worth digging deeper into why this is thought to be the case. There are two mechanisms that seem to be doing important work here, in terms of reducing perceived exit costs. The first is the certainty the basic income provides, in terms of ensuring access to income support. There is no complicated application process, nothing that needs to be reviewed and approved by a civil servant. The basic income is already being paid out, and will continue to provide income no matter what. There is thus no reason to fear that unemployment could mean being left without any income at all. The second is through what I will call the objective state of affairs-mechanism. By providing people with the means to avoid poverty, to afford basic necessities, no matter their employment status, a UBI would ensure that no person is completely dependent on their employer (or anybody else) for their financial well-being. No matter what happens, whether they are fired or chose to quit their job, they will not face poverty. The UBI thus helps bring about a state of affairs where, from an objective standpoint, people are shielded from the worst possible outcome (being faced with poverty) even if they are unemployed.

The first of these mechanisms, I believe it is right to say, captures something important about how guaranteeing and easing access to income support can reduce the exit costs associated with unemployment–and I will have more to say about it in the subsequent section of this paper. With regards to the latter, however, there are reasons to be more sceptical.

4 Implications & Trade-Offs

There are several implications that flow from the above discussion that are worth underlining here, as they point to and help us identify some important trade-offs that would be at stake in the decision between a baseline provision UBI-scheme and a more traditional income maintenance-model of welfare provision. First, replacing a traditional unemployment insurance with a UBI would raise exit costs for people who are (or would have been) eligible to receive the UI, assuming the system of public administration is well functioning enough to provide people with the benefits to which they are entitled within a reasonable time frame. Thus, to the extent that a UBI, under such circumstances, would reduce or eliminate the threat too non-domination posed by the discretionary power of government bureaucrats, this gain would come at the expense of increasing a large share of the working population’s vulnerability to domination at the hands of their employers.

In thinking about the implications of this trade-off for the respective merits of a baseline provision model built around a UBI- and an income-maintenance UI-model from a non-domination perspective, there are several things worth noting. To begin with, it indicates that the strength of the case for baseline UBI-model will depend on to what extent there are reasons to worry about domination at the hands of government bureaucrats and employers in the country in question, which in turn will in part be a function of the broader policy regime and to what extent there are other tools for mitigating the risk to non-domination posed by government bureaucrats and employers. In countries where the institutions of public administration are reasonably well functioning, where norms governing how civil servants ought to make determinations are largely followed–norms regarding objectivity and impartiality, etc.–and administrative courts and ombudsmen are sufficiently independent, the case for making the trade-off a baseline UBI-model would imply would be weaker than it would be in a country where the institutions of public administration are not well functioning and where government bureaucrats have unchecked discretionary power.

Second, there would be distributive trade-offs at stake in the decision between an income-maintenance UI-model and a baseline UBI-scheme when it comes to republican liberty. While a baseline UBI-model would raise the exit costs associated with unemployment for some (perhaps a large) share of workers, a baseline UBI-model would also increase the financial independence, the anti-power, of the segment of workers (or aspiring workers) that the income-maintenance UI-system would exclude or serve poorly, including people that rely on low-paid, part-time or gig-economy-type jobs for their income–whether by deeming them ineligible or as a consequence of administrative hurdles preventing or delaying access. There is no iron law that says how large this group must be, of course. The rules regarding eligibility can be designed to be inclusive and to ease access–by not requiring people to have worked a certain length of time, or a certain number of hours, to qualify, for example, or by simplifying the application process. But regardless of how the rules are set, there is bound to be a set of workers that are excluded, for whom a UBI would lower the exit costs associated with unemployment.

Third, though a more traditional income-maintenance UI-model would do a better job, compared to a baseline UBI-model, of keeping the exit costs associated with a move from employment into unemployment low (for a large segment of workers), it would do a worse job when it comes to enabling people to exit the labour market altogether. This trade-off follows from the insurance-like design of the UI. It is not designed to provide a permanent source of income, but rather to provide people with income support during a period of transition from unemployment back into employment. Though UI-systems can be designed to nudge or encourage people back into employment in a wide variety of ways–some more problematic from a non-domination perspective than others–one of the features of a UI is precisely that it tries to gradually nudge people back into employment. Not too quickly, necessarily–even from a purely economic perspective there are reasons to want people to find work that matches their skills and interests–but in a reasonable amount of time. This means that after a certain point, the UI-benefits will no longer be paid out, forcing recipients to apply for some other form of government assistance if they cannot find work.

Is this a problem from a non-domination perspective? The answer would seem to partly depend on the nature of the labour market. If the labour market is oligopolistic, the dependence on needing to sell one’s labour can translate into a dependence on the good will of the relatively few employers out there, which would be problematic from a non-domination perspective. However, if the labour market is (truly) competitive, if there are lots of potential employers out there, then the need to sell one’s labour need not (at least necessarily) be a problem from a non-domination perspective.^[6] Not all forms of dependence are problematic from a non-domination perspective, after all–only those forms that place people’s well-being at the mercy of the arbitrary power of particular people or actors.^[7] Of course, creating and sustaining labour market conditions that give workers lots of potential employers is not easy–particularly in all areas or regions of a country, and would require the exit costs associated with moving in search of work to be brought as low as possible. So, there is good reason to worry about the effects on non-domination the need for workers to sell their labour has, even if the need to do so is not inherently problematic from a non-domination perspective.

Thus, whether the case for one side of this trade-off is stronger than the other will depend, in part, on the nature of the labour market and how it is likely to develop as well as the prospects for regulating and managing it in a way that furthers non-domination. Where the labour market conditions are relatively favourable to workers, the case for a UBI would be weaker; where the labour market conditions are less favourable, the case for UBI would be stronger, all other things being equal.^[8]

The answer would also seem to partly depend on the downstream effects a UBI would have on domination within the family. Though there are reasons to think that a UBI could make family members more financially independent of each other (Gonalons-Pons and Calnitsky 2022), at least under certain circumstances, the effect on domination within the family of changes to employment patterns resulting from a UBI would be relevant to consider. In social democratic welfare states like Sweden and Denmark, a key pillar of the policy regimes used to reduce economic dependence (and therefore domination) within families has been the promotion of full employment (Sainsbury 1999). If–and of course this is a big if–the impact of a UBI would be to reduce employment, the effects of this reduction on income disparities and economic dependence within families would be an area of concern for those interested in the question of whether a UBI would advance (on balance) republican liberty (see Gheaus 2008). The evidence from the Finnish basic income experiment suggests a basic income need not have a of negative effect on employment (Kangas et al, 2021), but the question seems far from settled (see, for example, Vivalt et al, 2024), and the impact would likely depend on the broader policy context and the design choices made with respect to the basic income.

Having sketched the implications and trade-offs at stake in the choice between a baseline UBI-model and an income-maintenance UI-model of welfare provision, as they relate to non-domination, it is worth drawing out what this means for how we think about the likely impact on non-domination of a sort of hybrid model–a modest UBI existing alongside a fairly generous system of conditional and targeted benefits (along the lines envisioned by the Finnish basic income experiment).^[9] First, if or to the extent that the (modest) UBI would need to be funded through cuts to the generosity of the UI, the trade-offs and distributive impacts described above would come into play. Second, though this type of model would not eliminate the threat posed to non-domination by the government bureaucrats tasked with distributing and determining eligibility for conditional and targeted benefits (since the benefits would remain and some people would be in need of them), a modest UBI would conceivably remove the need for at least some of these potentially problematic administrative judgements (thus reducing the opportunities for a certain kind of domination). Third, for the groups within society that are poorly served by UI–because of uncertainty regarding eligibility or administrative burdens delaying access–a modest UBI would conceivably lower the exit costs associated with unemployment, since at least some of the uncertainty regarding access to income support would be removed. A modest UBI would also reduce the size of the drop in month-to-month income associated with unemployment for those who are ineligible to receive UI (due to a failure to meet qualifying requirements). There are thus reasons to think that the republican liberty of at least some groups of people could be well served by a hybrid model, combining a modest UBI with a generous system of conditional and targeted benefits, even if there are likely to be trade-offs.

5 Conclusions

This paper has sought to shed light on whether and under what conditions a UBI would advance republican liberty by developing a more fully theorised account of the impact of UBI on the exit costs associated with unemployment. Once we properly account for the fact that people are loss averse and assess exit costs in a context of uncertainty and limited information, it has argued, we are in a better position to gage the non-domination related trade-offs and distributive effects of a UBI under different circumstances and with respect to different domains of concerns. Moves towards a baseline provision-model built around a UBI would raise the exit costs associated with unemployment for workers who qualify for the UI, thus increasing their vulnerability to workplace domination, while advancing the republican liberty of those who are poorly served by an income-maintenance system built around a UI. A UBI would also reduce the need for people to sell their labour, though whether or the extent to which this would advance republican liberty depends on the nature of the labour market (whether it is truly competitive or oligopolistic) as well as the downstream effects of people’s employment choices on domination within the family. Future research on the merits of UBI from a non-domination standpoint would thus do well to focus on the merits of introducing UBI in particular countries and institutional contexts.