The context: Spain after the outbreak of the economic crisis

In the Spanish political agenda, two factors have boosted the debate on Basic Income (BI) in recent years: on the one hand, the combination of welfare budget cuts and growing poverty rates has spread claims for a radical reform of Spanish welfare policies; on the other hand, the emergence of new political actors such as Podemos or Ciudadanos has generated a vivid discussion on income guarantee proposals, including BI. In this article I will review these political proposals by relating them to the specific problems they would face in the present Spanish welfare state.

Let me start with two spoilers: my conclusion will offer bad and good news for BI supporters. The bad news is that, all things considered, a generous but means-tested Guaranteed Income (GI) program – not a BI – is most likely the best they can get in Spain in the short and medium term. The good news, however, is that a cleverly designed GI might not be so far from BI as many of its defenders might think. I will start by contextualising the present debate on income guarantee policy in Spain.

The effects of the great economic crisis of 2008 arrived in the Spanish economy during that year, but the financial crisis did not really have an impact on the public budget until 2010, when the Socialist PM José Luis Rodríguez Zapatero executed the largest pack of budgetary and social cuts in the history of the democratic period. Those cuts were in overt contradiction with the Socialist Party's electoral program. They triggered the spontaneous and massive popular movement of 15-M, and caused the worst historical defeat of the Socialists in the 2011 national election, when the right-wing Popular Party won an absolute majority in the Spanish Parliament.

What followed was an even more severe set of cuts to social policy spending, a general fall in wages (especially low and middle ones), a huge increase in public debt (which went from around 30 per cent before the crisis to 100 per cent of GDP), and a bank sector bailout of 5 per cent of GDP. Through the period 2008–2014, the unemployment rate rose from 8 per cent to a historical maximum of 25 per cent, and the poverty rate went beyond 22 per cent of the population (32 per cent if calculated with thresholds anchored in 2009 and at constant prices, which equals almost fifteen million people). Poverty rates went up even with lower relative poverty thresholds, which were the effect of the general income loss. Income inequality as reflected by the Gini index rose from an (already high) 0.319 to 0.347.Footnote ¹ Evictions of low-income families were the order of the day. This situation was almost generally acknowledged as a social emergency. Leaving aside the Popular Party in office, a vague consensus emerged that some type of national income guarantee was needed in order to tackle it.

The Spanish welfare state, even without budgetary cuts, was not ready to react properly to that challenge. Spain has traditionally been classified by the social policy literature as a Mediterranean welfare regime, typically joining a low female activity rate, a strong role of private family care, a dual labour market, a wide shadow economy, and weak social assistance benefits. However, during the last three decades the expansion of social expenditure in some key areas reduced the distance from some continental regimes (for example, in contributory pensions and unemployment benefits) and also from social-democratic ones (mainly in a high quality universal health service and a universal public education system).Footnote ²

Precisely, the huge weight of contributory benefits in the social budget reinforced the discriminatory effects of a dual labour market at a time of growing unemployment and income fall. The elderly is virtually the only collective that did not experience a growth in poverty rates or a remarkable income loss, due to a relatively strong and generous pay-as-you-go contributory pension system. On the contrary, the social and labour market position of the younger generations worsened extremely, since virtually no social benefits cover them.

Other traits of the Spanish tax and benefit system also contributed to the perfect storm. A low fiscal capacity combined with strongly regressive fiscal welfare results in a very poor redistributive capacity. Social assistance benefits for the working-age population are limited to an unemployment benefit with a very low coverage and to minimum income programs managed by the regional governments. Spain is a highly decentralised country, with huge inequalities among its seventeen regions, and each of them has its own minimum income scheme, with very different coverage levels, eligibility conditions, amounts of benefits, and associated activation measures. This very fragmented system of minimum income programs is highly sensitive to administrative and political arbitrariness. In short, the effectiveness and equity of the Spanish welfare system suffers from multiple contradictions and institutional design problems, and the crisis did nothing but exacerbate them.

Let me end this introduction by clarifying some concepts I will use in the rest of this article. I depart from the standard definition of BI as a completely universal, unconditional and individual cash benefit. By contrast, conventional Minimum Income (MI) schemes include some kind of means-test (typically, eligibility depends on falling below some defined income level, or even on having no income at all), some kind of behavioural conditions (such as availability for job offers or activation-training), and are delivered on a household basis.

Two other types of cash benefits typically relax one of the eligibility conditions of MI. Wage Supplements (WS) for low and middle income workers – often, but not always, delivered as tax credits – relax the condition of absence or insufficiency of income, since they are in-work benefits that top-up low and middle wages. GI proposals (which are increasingly present in the Spanish political agenda in recent years) relax the behavioural conditions of MI, by typically suppressing the link between means-tested benefits and availability for work or activation. They usually top-up existing income until reaching the poverty threshold. However, unlike BI, both WS and GI still pay the benefits on a household basis. Finally, a Negative Income Tax (NIT) is a variant of GI paid through the tax system: after a tax declaration or some other type of income check by the tax administration, eligible households/individuals receive a ‘negative tax’ while the rest pay the approved tax rate.

The political prospects: how new and old actors face welfare reform

The question of how to bring about a BI – the question of its political feasibility – has gained momentum amongst BI advocates (De Wispelaere and Noguera, Reference De Wispelaere, Noguera and Caputo2012). Of course, a key aspect is strategic feasibility: how to build a political coalition able to win majority support for BI.

The political context in Spain in the period 2014–2016 was apparently favourable to change. The social unrest generated by the crisis was aggravated even more by constant corruption scandals linked to the conservative Popular Party (in office) and the main institutions of the state. Podemos, a new party born from the popular movement of 15-M, rose from nowhere in the 2014 European election. After some months, it was leading some polls and seriously threatening the ‘1978 regime’. The king resigned and the Socialists elected a new young leader to try to stop their self-destruction. After the 2015 election, in which Podemos reached 21 per cent of the vote, political deadlock led to repeat elections in 2016 (Orriols and Cordero, Reference Orriols and Cordero2016; Rodríguez Teruel et al., Reference Rodríguez Teruel, Barrio and Barberà2016). During all this period, income guarantee proposals gained an important share of the agenda in electoral campaigns and public discussion.

However, the two traditional problems affecting the political feasibility of BI quickly showed up: marginal political support and ‘cheap’ political supportFootnote ³ (De Wispelaere, Reference De Wispelaere2016). In fact, despite widespread discussion during these years, no BI coalition exists today in Spain. There is not a single political party or union nearly unanimous on BI. If we look at the two traditional progressive parties, the Socialist PSOE and the post-communists of the United Left (IU), the first offers only vague promises of ‘studying’ the proposal, while the second is clearly opposed and favours a guaranteed job scheme. Serious attention came only from Podemos. For the first time in contemporary Spain, it seemed like a non-marginal party, with real chances to enter a coalition in a progressive government, was giving non-‘cheap’ support to BI.

Podemos included a universal BI in its program for the 2014 European Parliament election. After reaching a surprising 8 per cent of the vote and being rocketed by subsequent polls, public scrutiny of the party's program by the media quickly identified BI as one of the main targets for attack. BI started to be devastated in prime-time TV talk shows as a utopian, populist, and unaffordable madness in a country that had a narrow escape from bankruptcy only two years before. With a national election coming in months, Podemos leaders asked for expert advice in order to build a new, more thorough and feasible political program. This resulted in a full proposal for a means-tested GI with no activation requirements attached, combined with a WS designed to tackle the poverty trap. In fact, this shift from plain BI to GI as a stepping-stone was apparently reflecting the widespread preferences of the party's members: in an internal open online survey (Barragué and Martínez, Reference Barragué and Martínez2016), a clear majority (62 per cent) chose as preferred option a means-tested GI for poor households over a universal BI (28 per cent).

Despite fierce opposition from some BI radical advocates and even from the Spanish Basic Income Network (official member of the Basic Income Earth Network), the proposal was approved by 80 per cent of the members of Podemos, while BI didn't get enough support to be among the proposals for the vote. This way, Podemos solved a difficult trade-off by coming up with an income guarantee scheme that was hard to attack in the media, but was at the same time ambitious enough to keep support from activists and the indignados 15-M movement. The plan's budget reached 1.5 per cent of GDP and aimed to eradicate monetary poverty in Spain while fighting the poverty trap at the same time.

The discussion on BI had an interesting by-product: it generated growing public interest in income guarantee programs in general, so that other national parties also felt obliged to come up with their proposal. The Socialists and the unions proposed a national MI program (means-tested and with strict activation conditions). The new liberal party Ciudadanos defended a WS administered as a tax credit. Finally, although IU (post-communist) insisted on its guaranteed job scheme, they complemented it with a MI for the unemployed who are not covered by the former. In this scenario, it was quite unlikely to build a coalition for a universal BI. Instead, an agreement between the abovementioned forces for a means-tested GI with weak activation conditions could be worth exploring, provided it were complemented by some WS for low-income workers and by an independent right to social inclusion through training and/or employment. In fact, this is the path taken by many regional governments since progressive political coalitions came to office in several regions in Spain after the 2015 regional election (Noguera, Reference Noguera2018).

The economic prospects: ‘show me the money’

Many BI supporters in Spain often claim that its economic feasibility is out of the question, and that the problem is rather about political feasibility. However, from the very beginning that BI started attracting attention in the public arena, the focus was on its economic feasibility. The ‘show-me-the-money’ mantra was omnipresent in post-crisis Spain's political debates. The few existing studies estimate the net cost of a universal BI at the poverty threshold around 18.7 per cent of GDP, so huge tax raises would be required to fund it even in the presence of powerful economic growth (Arcarons et al., Reference Arcarons, Domènech, Raventós and Torrens2014; Doménech, Reference Doménech2017). On the contrary, the main proposals for income guarantee programs by different political parties and unions moved between 0.6 per cent and 2 per cent of GDP.Footnote ⁴

The huge net cost of BI schemes proposed so far in Spain when compared to other income guarantee proposals is explained by two main factors (Noguera, Reference Noguera2015). The first is the individualisation of the benefit, which would be no longer calculated in terms of equivalence scales for additional household members, but uniform for all adults. The second is the place of the break-even point that separates net beneficiaries from net contributors. With a BI financed through a flat income tax (as in those proposals), this point is usually much higher in the distribution of income than, for example, under current GI proposals in Spain, even those including a WS.Footnote ⁵ Many taxpayers with incomes above the guaranteed threshold would become net beneficiaries, at the price of an increase in the marginal tax rate for taxpayers with gross income above the new point, as Van Parijs and Vanderborght (Reference Van Parijs and Vanderborght2017: 135–7) explain.

For example, in the most important simulation study on how to fund a BI in Spain (Arcarons et al., Reference Arcarons, Domènech, Raventós and Torrens2014), the income flat tax rate required would be 49.5 per cent. According to the results of the simulation, almost half of the taxpayers would be net losers from the reform, and the losers would begin to be the majority in each decile from an annual gross income of 18,750€ (the seventh decile). Average effective tax rates would go up more than three points from 22,500€ upwards, and more than nine from 28,500€. For politicians trying to win an election, this is clearly a non-starter. The only alternative paths would be to lower the amount of BI (but then a means-tested GI could do better for those below the poverty threshold) or to fund BI by raising indirect or other taxes, but this would also face important financial and political risks, so no BI along these lines has been proposed so far in Spain.

Some BI advocates have tried to escape the economic objection with the counter-argument of tax-benefit integration. According to them, under a BI integrated with a flat tax rate and replacing all tax allowances and deductions, the ‘net cost’ would be better calculated by considering only the net income loss for taxpayers in the upper deciles, which goes down to 3.5 per cent of GDP. There are some problems with that. First, this is a method for estimating costs that could in principle be applied (but never is) to every social benefit in cash or in kind, and, in fact, to the whole building of the welfare state: income tax revenue is always partially funding social benefits, so why not calculate the ‘net cost’ of those benefits by just looking to the portion of income tax which ‘changes hands’? It would appear, however, that this confuses the ‘net cost’ with the net redistributive effect. Second, as said above, a substantial net loss would still be inflicted on not-so-high-income taxpayers. Third (and often unnoticed), such a tax-integrated BI would not in fact be a universal ex-ante payment to all citizens irrespective of their income, but instead a NIT which pays tax benefits to net recipients ex-post (see Van Parijs and Vanderborght: Reference Van Parijs and Vanderborght2017, 37–9).Footnote ⁶ In this scenario, one of the main rationales for preferring BI to GI (the absence of means-test) is weakened. A tax-integrated BI would be ‘compensated’ with an equivalent increase in tax at source for most taxpayers (Piketty, Reference Piketty2017), so the difference with a policy pack including a GI plus some WS for low-income workers may be easily minimised. The real differences, as said above, would then just be about how high the break-even point is located in income distribution, about the degree of individualisation, and about the administration and payment mechanism (direct welfare payments in the case of a GI, negative tax payments in the case of tax-integrated proposals).

Of course, by definition it would be always possible to finance a ‘fiscally neutral’ BI by simply reorganising current non-contributory welfare. However, a recent OECD study (OECD, 2017) shows that in this case its amount may be well below the poverty line and would even increase the poverty rate, since the reform would decrease the average protective intensity of the current minimum benefits. Even if we eliminate all income tax allowances and deductions, and distribute them as part of a BI, the results would not be impressive, at least for Spain: 4.3 per cent of GDP distributed among all citizens would just raise the BI amount by 76€/month.

An additional concern refers to opportunity costs. This concern is aggravated by the fact that the Spanish welfare state suffered severe budget cuts in health services, public education, elderly care services and other social investment policies during the crisis. The huge costs associated with BI are viewed by many progressives as an impediment to reversing these cuts and to improving the quality of all these services in the near future. Even if some version of BI were affordable, the question remains as to whether it would be preferable to channel those resources to other social programs where they are urgently needed (and demanded) by the population.

The institutional fit of BI in the Spanish welfare state

Institutional feasibility is a key feature for BI. A BI scheme that is not aimed to simply replace all monetary benefits should be properly articulated with the complex current system of social benefits. The main problems arise from integration with contributory benefits, which are the bulk of monetary benefits in Spain. As I have argued elsewhere (Noguera, Reference Noguera, Widerquist, Noguera, Vanderborght and Wispelaere2013), there is a conceptual tension between a universal BI and contributory benefits: while the former is unconditional, flat-rate, universal, and independent on previous contribution, the second's rationale is income maintenance at the accustomed level. Contributory benefits are not perceived as unconditional aid, but rather as a deferred wage generated by previous contributions and, therefore, as an individual right acquired through a past work record.

It is true that the contributory logic is partially broken by the existence of redistributive aspects that deviate from proportionality between contribution and benefit. However, this does not eliminate the aforementioned tension. In particular, the estimations of the net cost of a BI in Spain assume that contributory benefits equal or higher than BI already ‘count as’ the BI of the corresponding recipients. This is to say that BI will be deducted from current contributory benefits in the same way that it is ‘compensated’ by higher income tax rates for those in the upper deciles.

This method reduces again the net cost that BI would otherwise have, but raises at least three additional concerns. First, it may reduce incentives to pay social security contributions for certain groups of workers (particularly for the self-employed, temporary, and part-time workers), since contributions that entitle them to a pension lower than BI become senseless: in effect, a retired worker whose contributions entitle him only to the minimum contributory pension (which, let us suppose, has a similar amount to BI) would end up having a similar income level than a person who has not paid contributions throughout her working life. It could be easily argued that the first worker has paid twice for BI (through taxes and through contributions). This would raise the need for a substantial reform of the social security system.

Second, we would also face equity problems. For instance, a retired person enjoying a pension higher than BI but lower than (let us say) the minimum wage could rightly ask why his net income remains the same after BI is introduced, while many workers who receive wages higher than the minimum would see their net income increased, since the break-even point is expected to be well above that level. This would pose problems of intergenerational justice that in all likelihood would force a careful, slow, and complex application of the proposal, which would partly distort its theoretical simplicity.

Third, there are also financial and accounting complications. To reduce the bulky net cost of BI, the estimations directly deduct the equivalent amount of all contributory benefits, which not only include retirement and unemployment benefits, but also a substantial budget for maternity and paternity leaves, paid sick leaves, survival, disability, and orphanage pensions. In this scenario, social security contributions (whose accounting is today separate from the rest of the Government's budget) would be directly used to fund a BI program that should be expected to depend on the general public budget. But if this ‘simple’ assumption were not made, then the net cost of BI would skyrocket even more (at least an additional 6 per cent of GDPFootnote ⁷).

Last but not least, the Spanish Constitution allows the regional governments to claim total legal and budgetary control over social assistance, so they have generally done so. Therefore, a national-wide income guarantee program could be legally challenged on these grounds. The implementation of a BI or a GI at the state level should be carried out either as a social security benefit (something that could also raise legal problems) or as a NIT. Alternatively, an agreement could be tried with all regional governments for the articulation of an integrated system of top-ups. In any case, however, the risk of establishing perverse incentives for the regional governments to stop allocating the same level of resources to their MI schemes seems evident.

The administration of a BI in Spain: how far is BI from GI?

Means-testing and targeting of the needy have been very controversial issues in the public debate about BI. Perhaps one of the lessons we can learn from the extremely fragmented system of MI programs in Spain is how diverse ‘means-testing’ methods may be, and how the debates on BI often fail to grasp that diversity. Although very different types and degrees of income eligibility conditions and controls exist in different cash benefit schemes, an extended view is that means-testing is essentially humiliating and inefficient, and that only a universal BI paid ex-ante to everyone would grant non-discrimination and equal respect for citizens, not to mention a huge reduction of administration costs.

Contrary to this shared view, it could be argued that ‘nominal’ universality/targeting does not matter as much as the specific details of implementation and the fine-grain design of the program (Schelling, Reference Schelling1981; Gough, Reference Gough1996; De Wispelaere and Stirton, Reference De Wispelaere and Stirton2013, Reference De Wispelaere and Stirton2017). To illustrate this, let me propose a hypothetical example. Imagine we want to implement a universal BI of equal amount to the poverty threshold and integrated with a relatively high flat tax rate (which is, as we saw, the usual proposal discussed in Spain). We then would need to exhaustively identify all citizens who do not declare income tax (who might or not be net recipients with the new tax regulation). We should deploy the administrative means in order to disburse effective payments to all net recipients, while, at the same time, reinforcing hugely the tools for detection of concealed income (for example, in the widespread Spanish shadow economy, or among self-employed workers). We should also completely individualise income tax declarations and start declaring all income from the first euro and no matter the source (something that is not the case now, by far). All this would multiply the number of declarations. Finally, we should monitor monthly (not annually any more) the specific situation of citizens in order to detect whether they shift from being net contributors to net recipients and the other way round. Definitively this scenario does not sound as if it would bring a huge reduction of administration costs compared to the present system, or an eradication of income checks and tests when disbursing net payments to citizens. Of course, a GI would also face many similar problems, except that the number of potential recipients would be much lower.

Now imagine a GI with the same amount implemented like this: if your (household equivalent) income is below the poverty line, you have a legally enforceable right to an income top-up to reach that threshold. No work or behavioural requirements (or very weak ones) are attached. Applications are simple and may even be presented online in order to minimise ‘stigma’ and non-take-up. As for means-testing, the burden of the proof is on the administration side: they have to check (against tax, social security, and other records) whether someone is eligible, so no one has to prove she is poor. In doubt, it is for the administrative bodies to prove she is not. The waiting time in order to process and approve or reject claims is minimised. Finally, the benefits are renewed annually if personal circumstances do not change, or adjusted if they do.

How far are these two schemes from each other in practical terms (leaving aside amounts and break-even points)? A seemingly plausible answer is: not so far as currently assumed by theoretical discourses on the difference between universal BI and ‘conditional’ schemes. In fact, only two substantial differences would persist: first, since the GI would top-up income until the poverty threshold, a poverty trap would persist, while BI would eliminate it. Second, the amount of GI would be calculated at the household level according to income equivalent scales, while BI would be completely individualised. These differences, however, might be reduced by combining the GI with a WS for low-income working families, and by splitting the benefit equally between the household's adult members.

Let me be clear about this hypothetical example: my point is not that a tax-integrated BI has a lot of problems while a generous GI would not. My point, on the contrary, is that the many problems faced by these two ambitious income guarantee schemes would be very similar in both cases (perhaps more serious in the first one, since it is targeted to a much larger population). Dismissing the second as an essentially opposed model to the first is probably too simple.

In fact, some features of our hypothetical GI are very similar to what some regional governments (typically supported by Podemos, the Socialists, or regional nationalist parties) are already exploring in several MI programs since the 2015 regional election (Noguera, Reference Noguera2018). The Basque Country GI program has pioneered this path by establishing a generous WS, extending coverage of the GI, and relaxing some of the activation requirements (Sanzo, Reference Sanzo2013; Zalakain, Reference Zalakain2014). Catalonia also passed in 2017 a GI law that introduced a benefit without activation requirements, together with a supplement if participation in insertion plans is accepted (Noguera, Reference Noguera2017). Although these measures are still far from the second scenario in the above example, they are significantly approaching it.

Note that none of the differences between both hypothetical schemes is about making payments ex-ante. Under a tax-integrated scheme, payments would be necessarily disbursed ex-post the income tax declaration, since it has to be determined whether one is a net recipient or not. Checks for every citizen would not be disbursed ex-ante and funded by income afterwards (see Note 6). This is also stressed by Piketty (Reference Piketty2017) when he argued for the particular ‘BI’ scheme proposed by Benoît Hamon in the last French presidential campaign. Some Spanish BI advocates (Arcarons et al., Reference Arcarons, Domènech, Raventós and Torrens2014; Gimeno Ullastres, Reference Gimeno Ullastres2014) present a very similar scheme as if it were a pure ex-ante BI, but it would be better described as a NIT. The truth is that under both schemes, income controls would end up being quite similar, and net recipients would continue claiming benefits either through conventional applications or through income tax declarations. One may wonder whether it still makes sense to contrast both scenarios by dismissing one as ‘conditional’ and ‘means-tested’ while presenting the other as ‘universal’ and ‘unconditional’.

The trilemma of BI in Spain (and elsewhere?)

When examining the political and economic feasibility of BI in Spain, the most plausible conclusion seems to be that a GI scheme (probably combined with a WS for low-income workers) would be a much more feasible option in the short term. Unlike BI, it would be economically affordable. It would be also able to raise support from a wide range of political parties and social organisations (note that the recent Catalonian GI law was passed with the unanimous vote of all seven parties in the Catalan Parliament). Finally, it would fit better the present institutions of the Spanish welfare system, particularly contributory benefits.

This might be good or bad news depending on the reference point. For those whose main concerns are about tackling poverty and social exclusion by maximising transfers to the worst-off in the short term, GI may well become an end in itself. For those whose dream is a BI society governed by principles such as real freedom or non-domination, GI may be seen as an insufficient progress, or perhaps even as an unacceptable compromise endangering the final success of their fight. However, all the evidence we have suggests that BI advocates face a political trilemma. They cannot have simultaneously purity (that is, an ex-ante universal and unconditional BI), urgency (a direct implementation of such BI in the short term), and consistency between the usual definition of BI and the real policy options they are proposing. The main lesson from the Spanish experience is therefore that BI supporters must be ready to compromise and accept generous GI and/or WS proposals as stepping-stones if they want to see some progress in the right direction in the future. The prospects for a fully universal BI may always improve in the medium-long term, and new opportunity windows may open. If that were the case, they would be worth exploring. However, the needs of many families hit by poverty and exclusion cannot wait until a full and ‘true’ BI is feasible.