A means tested basic income solves inequality best, because it focuses on aid with the highest social utility, saves money, and is the most efficient – recent empirical examples prove.

1. Introduction In recent years there is a heated debate on the desirability of Basic Income as a future device for substituting traditional welfare, which continues to be implemented by governments in all countries. In fact, experiments for Basic Income were performed in Canada, Finland, India and Namibia (and proposed also in Switzerland). Recently, Spain has accelerated the implementation of such a system (called "minimum vital income3") in order to cope with the consequences of the Covid 19 crisis, which created in the short run a situation of poverty for many families during the months of economic inactivity, that was imposed in order to avoid contagion. According to the Spanish government's decision, the basic income is transferred to eligible families starting in June 2020. In the background of this proposal, we are witnessing a process of substitution of workers by the use of machines – following the growth process that characterizes modern economies. In a situation where growth enhances the substitution of unskilled workers by machines, it would be desirable to adjust ex-ante the welfare system, in order to gradually cope with increasing unemployment of older workers, as observed in many countries.4 The means tested Basic Income proposal claims for a transfer that will be guaranteed to low-income workers; in such situation this income would allow individuals that confront difficulties for becoming skilled to have decent living, and thus improving social welfare under the new technological environment. Paradoxically, the "machines instead of workers" phenomenon occurs in a period in which more governments are implementing the Earned Income Tax Credit as a policy tool, which actually curtails disincentives for non-participation of the Working Poor and consequently it depresses unskilled wages. Such an expansion in the use of this tool, in a world where machines substitute workers, represents substantial trade-offs in terms of policy, and calls for an analysis that allows assessing the comparative advantage of Basic Income programs against traditional welfare. Since the General Basic Income program has been rejected by most economists (as explained below), it is important to check whether its relevant substitute – a means-tested Basic Income – is the right remedy under the new situation. In particular, policy-makers are interested to know whether under increasing inequality a means-tested Basic Income system produces a policy option with higher social utility compared to traditional welfare. In this paper we deal with this basic question. In order to do so we use a standard framework widely accepted in Public Economics, and we run a substantial number of simulations by mapping results as a function of the most relevant parameters that are traditionally accepted to perform a valuation. In order to evaluate the desirability of a means-tested Basic Income program we distinguish between a Conservative Social Planner, who evaluates the desirability of the proposals by looking only at consumption; and a Liberal Social Planner, who uses a standard framework of utility based on both consumption and leisure. In all scenarios we will concentrate on the short-run implications. 5 2. Literature Survey 2.1 Replacement of workers by machines Acemoglu and Autor (2011) explain the impact of technologically driven growth on labor market developments. The use of machines acts a substitute for routine and non-skilled work, creating the need of re-thinking jobs and doing the needed adjustments for labor supporting government policy. Autor (2014) shows that wages and earnings of lower-skilled workers, particularly men, have stagnated over several decades. While so far technology advance did not cause an apocalyptic increase in unemployment,6 there are other remarkable effects that merit further analysis and adoption of policy measures. One of the effects of "machines replace workers" type of growth is the difficulty for old workers to re-insert themselves into the labor market. Aubert, Caroli and Roger (2006) investigate the relationships between new technologies, innovative workplace practices and the age structure of the workforce in a sample of French firms. They find evidence that the wage bill share of older workers is lower in innovative firms and that the opposite holds for younger workers. This age bias affects both men and women and it is also evidenced within occupational groups. More detailed analysis of employment inflows and outflows shows that new technologies essentially affect older workers through reduced hiring opportunities. This difficulty maybe reflected even in discrimination practices, as studied by Adams (2004). This author shows that this practice is quite common among employers, and it brought about a number of legislation efforts among many US states. The vast literature on the acceleration of the ageing process brought about renewed calls for adopting old proposals for welfare reform, based on the basic income approach. According to this approach individuals should rely on a basic income, which will allow them to cope with this kind of phenomenon from a perspective that assures dealing respectfully with the cost of living. Philosophically inclined advocates of this approach (e.g., Van Parijs & Vanderborght, 2017), emphasize the related value of a Universal Basic Income in increasing human freedom, a difficult concept to express in terms of budget constraints, but nevertheless an important goal. Hoynes and Rothstein (2019) carefully describe the new attempts for renewing the Basic Income initiative, that was accompanied by new pilots implemented at different countries – including Canada, Finland, India and Namibia (and proposed also in Switzerland). Nevertheless, these authors clearly show that the General Basic Income initiative by itself must be rejected, since it is too expensive and implies many inefficiencies. These authors show that there are many issues related to specific groups that may be targeted by welfare and to the incentives associated with the payment schedule, including the phasing out. A clear result that is implied by their analysis is that a relevant alternative shall be based on a restricted version of Basic Income. In this paper we will propose a mean-tested Basic Income approach.