The effects of unconditional basic income on labor supply and the role of intrinsic motivation : theoretical analysis of how the introduction of an unconditional basic income affects the labor supply and an assessment o

(1)

The effects of unconditional basic income on labor

supply and the role of intrinsic motivation

Theoretical analysis of how the introduction of an unconditional basic income affects the labor supply and an assessment of the role of intrinsic motivation

Paco Pos 10545972

BSc Economics and Business

Specialisation: Economics

Date: 2 February 2015

(2)

Statement of Originality

This document is written by Student Paco Pos who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Acknowledgements

Upon completion of this thesis, I would like to thank my thesis supervisor Stephan Jagau PhD for guiding me through this process. We have assessed many possible subjects and eventually settled on a subject that is relevant and interesting. Despite my inability to choose, the subjects were made more explicit and the choice became clearer thanks to the guidance of Mr. Jagau. Besides, he quickly gave critical, useful advice and comments, which have helped me writing an even better thesis.

(4)

The effects of unconditional basic income on the labor

supply and the role of intrinsic motivation

Abstract

In 2016 several unconditional basic income field experiments will be conducted. Proponents of this policy have raised argument for implementation of such as policy with regard to freedom and justice for all citizens, especially those who might lack the means to access their real freedom and develop their human capital. One of the main concerns of critics focuses on labor market participation and the labor intensity. The fear exists that the implementation of an unconditional basic income will be followed by a drop in labor-market participation and/or in average hours worked. Evidence from field experiments suggests that the response pattern might be less straightforward. As a result several authors have set out to explain why

unconditional basic income might actually lead to a net improvement of labor supply on the intensive or extensive margin. While mechanisms such as the poverty and unemployment trap have been useful in understanding part of the regularities observed in empirical basic income studies, there is still a considerable discrepancy between theoretical predictions and empirical findings. Theories of intrinsic motivation might provide crucial insights regarding the

remaining puzzles. Intrinsic motivation can accommodate a rise among employed individuals in labor intensity that cannot be explained by the unemployment trap. Additionally, it

indirectly affects labor in a variety of ways.

Keywords: basic income, social welfare, poverty, income security, intrinsic motivation, labor

(6)

- 2 -

1. Introduction

Unconditional basic income (UBI), also known as basic income guarantee, universal demogrant or citizen’s income, is a form of social security that is regularly, systematically and unconditionally paid out to citizens irrespective of the height of their income. Recently, four Dutch municipalities, Utrecht, Wageningen, Tilburg and Groningen, have stated to start with UBI field experiments soon, with Utrecht being the first municipality to start their experiment in January 2016. Another 28 Dutch municipalities have expressed serious interest in

implementing similar experiments (Bregman, 2015). This is not a phenomenon limited to the Netherlands, but rather the public interest in UBI is growing internationally, with several established academics acting as avid proponents of such a policy instrument. Concurrently, field experiments on the effects of UBI are planned in Finland, France and even a referendum regarding its nation-wide implementation is planned in Switzerland (Streithorst, 2015).This trend is an inviting backdrop for further research and ongoing academic and non-academic discussion on the topic.

Belgian economist and philosopher Phillipe van Parijs (2004a) states that

“unconditional basic income would serve as a powerful instrument of social justice: it would promote real freedom for all by providing the material resources that people need to pursue their aims. At the same time, it would help to solve the policy dilemmas of poverty and unemployment.” The introduction of UBI results in a multitude of social welfare enhancing phenomena. It may redistribute welfare from the better off to the worse off of society and decrease social inequality, as well as enabling every individual with the means to develop and improve her human capital. The steady income stream will secure a level of subsistence for all individuals, irrespective of age, gender, sexual orientation, race, income or employment. For the less advantaged of society this means the opportunity and means to use one’s freedom in order to overcome dependence of social welfare and to alleviate poverty. Poverty alleviation provides a solution to a multitude of problems, as poverty often lies at the heart of

malnutrition, lack of education, crime, drug abuse and poor mental and physical health (Haarmann, 2009; Forget, 2011; SEWA, 2014). Instead of aiming to solve these problems independently of each other, UBI can solve their root cause by providing a certain level of income security to all citizens in a country. The spill-over effects could enhance social welfare on the individual as well as on the aggregate level.

(7)

- 3 -

The most persistent counter arguments of UBI refer to the increased tax burden1 and to labor supply inefficiencies. The focus of this thesis lies on the effects of UBI on the labor supply. Critics argue that a substantial UBI might eliminate the necessity to work and might lead to widespread idleness, as the increased disposable income might make an individual decrease her amount of hours worked (Petersen et al., 2011; Hauser, 2006). An individual might find her new level of income high enough to refrain from working completely or an individual might find her new level of income higher than she needs for her preferred consumption bundle and will therefore substitute leisure for work effort. By contrast,empirical evidence from UBI experiments and similar policies suggests that also a rise in employment might be a likely scenario. In the case of developing countries there is a consensus that a rise in

productivity is caused by the overcoming of the poverty trap, meaning that UBI ensures a regular income stream which enables individual’s access to capital that can be invested in more productive activities that would otherwise have been out of reach of the individual. As a result the individual can maintain a higher and steadier income stream from these new

activities. For developed countries the results of field studies show more ambiguous effects on the labor supply. Negative income tax (NIT2) experiments show a mild drop in labor

participation and intensity, but UBI proponents argue that it is a more efficient policy then NIT, because it is better equipped to overcome the unemployment trap and ensures real freedom and self-determination of individuals. The unemployment trap refers to a situation in which an individual is faced by a perverse incentive not to work, because the rise in income from working will result in an equal or greater loss of entitlement to social welfare benefits, as a result the individual will choose to remain unemployed. Analysis of Alaskan dividend and Cherokee Indian profit sharing suggest that the effects of UBI might indeed be positive along the intensive and extensive margin. However, the unemployment trap is not the only factor that plays a role in the increase of the labor supply under UBI.

1

The proposed financing of a national basic income usually involves a reduction or removal of other benefits into one comprehensible UBI. Still, in predictions regarding national UBI implementations there is a rise in government expenditures (Hauser, 2006). However, it is expected that indirect effects (such as reduced health care costs) and simplification of the social welfare system can compensate this effect (Gilroy et al., 2012; Gamel et al., 2006). Additionally, van der Linden (2002) and Bowles (1992) theoretically show that UBI can be implemented without reducing economy wide profitability and while keeping a balanced government budget. Depending on how the basic income is financed, it can result in an increase in the burden of the currently employed through higher income taxes. As a result, Hauser (2006) argues that there is a disincentive to work and that the gap between the employed and the unemployed will become even greater. The feasibility of UBI remains unclear, as there has not yet been a nation-wide implementation of the policy.

2_{A NIT is a progressive income tax system where individuals earning below a certain amount receive} supplemental pay from the government and individuals earning above that amount pay income taxes to the government.

(8)

- 4 -

According to Self-Determination Theory (SDT3) intrinsic motivation is commonly conceived as doing something because it is inherently interesting, enjoyable or satisfactory, while extrinsic motivation commonly refers to doing something because it leads to a separable outcome (Deci & Ryan, 2000). Additionally, Deci and Ryan (2000) argue that intrinsically motivated tasks are those that provide satisfaction of innate psychological needs, such as the needs for competence, autonomy and relatedness. In the academic debate relating to UBI the role of intrinsic motivation has not received much consideration so far. All the while, the preceding discussion suggests that adding intrinsic motivation to the equation might significantly improve our conceptual understanding of the interplay between UBI and labor supply. This thesis aims to link the literature concerning intrinsic motivation to the literature regarding UBI. The leading question will be whether and if so what intrinsic

motivation can contribute towards explaining the difference between empirical evidence from field experiments in developed countries and standard economic predictions regarding labor supply along the intensive and extensive margin.

The structure of this thesis is as follows; Section 2 introduces the concept of UBI and its merits. Section 3 describes the labor supply critique. Section 4 reviews empirical evidence from field studies. Section 5 assesses two models that incorporate UBI and labor market mechanics. Section 6 investigates the role of intrinsic motivation. Section 7 presents limitations and considerations of the analysis. Lastly, section 8 concludes.

2. Unconditional basic income

The definition of UBI varies among scholars, but it is commonly conceived as “an income paid by a government, at a uniform level and at regular intervals, to each adult

member of society. The grant is paid, and its level is fixed, irrespective of whether the person is rich or poor, lives alone or with others, is willing to work or not.” (Van Parijs, 2004a). The concept of UBI was first introduced in 1848 by the French socialist Joseph Charlier. He advocated the introduction of a territorial dividend that ought to be granted to all citizens in order to redistribute the revenues of the state. A year later, English economist John Stuart Mill was the first to propose an unconditional UBI with the aim of providing everyone with the means of subsistence. Throughout history the idea has gained and lost importance in the academic discourse multiple times. After World War II there was a resurgence of the social

3

(9)

- 5 -

welfare debate in which UBI regained importance, because these policies could mitigate the drop in purchasing power and demand that was caused by the war. UBI has most likely inspired economist Milton Friedman’s idea of a NIT and has led to greater political support for similar policies in the sixties and seventies at the height of the civil rights movement that focused on equality (van Parijs, 2004a). Starting in the 90s, there has been a reemergence of interest in UBI that can be attributed to a failure of current social welfare systems to

overcome persistent poverty, structural unemployment and growing inequality (Clark & Kavanagh, 1996; van der Veen, 1998; van Parijs, 2004a; Offe, 2008). Even though there are social welfare systems in place in many countries, we still observe high unemployment rates. Furthermore, wages for low-paid jobs have stagnated and fully employed individuals in the lower class do not always succeed in earning enough to meet a level of subsistence (van Parijs, 2004b). In the meantime corporate profits are at record heights and educated

individuals are living comfortably. UBI might redistribute the welfare more evenly and ensure the means of subsistence to everyone (Offe, 2008). Field experiments that are currently being conducted or are planned for the near future can be expected to ignite the debate once again.

It is generally accepted that UBI is humane from a normative point of view, however, whether it is also a just policy is a philosophical debate that engages several academics4. According to van Parijs (2004b) UBI has several advantages over common conditional welfare schemes; First of all, the negative social stigma surrounding welfare recipients under conditional welfare schemes is not existent under UBI, because entitlement is based on citizenship rather than income. Secondly, the simplicity of UBI will ensure that every citizen is informed about the entitlements they have a right to. Thirdly, the supplemented income that is inherent to UBI is not lost when an individual accepts a job, while under income related social security benefits the individual has a strong perverse incentive not to take a job as she might lose (part of) her entitlement. This is referred to as the unemployment trap, which causes individuals to remain unemployed under social welfare programs conditional on income, because for individuals that are currently unemployed accepting a low-paid job means that each additional Euro earned is offset by a loss in benefits of one Euro. As a result the individual will not accept many jobs and put no effort in looking for low-paid jobs. In contrast, a UBI will always make the individual better off by accepting any job (as long as the opportunity cost are less than the wage). Similarly, Gamel et al. (2006) and Gilroy et al.

4_{Proponents argue that UBI leads to increased social justice because of decreased social inequality and increased} real freedom for the poor (Goldsmith, 1997; Preston, 1998; van Parijs, 2004a; van Parijs, 2004b). Opponents argue that “the idle should not be rewarded, the prosperous don’t need it and there are so many things waiting to be done in the world” (White, 1997; Offe, 2008).

(10)

- 6 -

(2012) argue that the introduction of UBI to replace income related social security benefits will increase labor participation by overcoming the unemployment trap5.

Thus, the means test6 is abolished under UBI and this may result in a removal of the unemployment trap and therefore increase the acceptance of paid jobs. However, low-paid jobs can be perceived as lousy, degrading, dead-end jobs of which the acceptance should not be promoted, but can also be seen as pleasant, enriching, stepping-stone jobs of which the acceptance should be promoted because of their intrinsic value or future prospects (van Parijs, 2004b). In order to ensure that individuals do not have to accept lousy low-paid jobs, the social welfare benefits have to be unconditional of work tests7, because only an individual who has the ability to discriminate between these two kinds of jobs based on her needs and wants and is not forced to accept or stay in such a job in order to remain eligible for work conditional benefits can reject such a job (van Parijs, 2004b). By taking these two

unconditionalities into consideration UBI can be expected to increase labor participation and labor intensity, as individuals no longer face a loss of social welfare benefits and will accept low-paid jobs in order to increase their disposable income by working or in order to gain experience and without the risk of being trapped in a low-paid job. In this way UBI may simultaneously alleviate poverty and reduce unemployment, while ensuring the individual’s right to self-determination or real freedom.

Van Parijs (2004b) introduces the term real freedom and states that social justice is not only a right to income, but also access to (paid and unpaid) activity. One could ensure

employment in low-paid jobs by introducing, for instance, an earned income tax credit (EITC8), but this may result in providing primarily lousy jobs to the worst off in society and will remove the their bargaining power and ability to discriminate between jobs and result in a loss of freedom. Under UBI there exists “a conception of social justice as the fair distribution of the real freedom to pursue the realisation of one’s conception of the good life, whatever it is.” (van Parijs, 2004b). Because the worth of freedom depends on access to resources to ensure that freedom, in a just society the distribution of opportunity to use one’s freedom must be safeguarded for every individual, especially for individuals that would not have this

5_{Gamel et al. (2006) refer to the unemployment trap by using the term inactivity trap}

6_{A means test determines eligibility for social welfare benefits depending on the level of income of an} individual.

7_{A work test determines eligibility for social welfare benefits based on whether the individual is employed and} the length of the individual’s employment.

8_{A EITC is a refundable tax credit for low income working individuals and the received amount depends on a} recipient’s income.

(11)

- 7 -

opportunity without redistribution. This may decrease social inequality and safeguard the possibility of even the less fortunate to build up their human capital.

Van Parijs’ (2004b) concept of real freedom shows many similarities with Deci and Ryan’s (2000) theory of self-determination. Self-determination affects an individual’s intrinsic motivation and well-being by satisfying several innate psychological needs and greatly influences the individual’s behavior. These universal psychological needs are; competence, autonomy and relatedness. Competence refers to the need for control of the outcome and the mastery of experience, autonomy refers to the need to decide on one’s activity in harmony with one's integrated self and relatedness refers to the need to interact with, be connected to and experience caring for others. Competence and autonomy are

ensured under UBI in the same way as real freedom is ensured and will as a result increase an individual’s satisfaction of doing a task, in other words, the individual is more intrinsically motivated. Under UBI an individual has the right and means to use her real freedom and can discriminate between jobs that yield different levels of intrinsic motivation, which

subsequently may increase one’s well-being. Theories of intrinsic motivation are closely related to implications of UBI and might provide crucial insights regarding the effects of UBI. Apart from normative judgments, some of the policy implications and effects can be assessed positively. There has been a vivid ongoing debate between proponents and opponents

regarding the effects of the policy on the labor supply and section 3 serves as an exposition of this critique.

3. Labor supply critique

Another integral part of the discussion, which also is the main focus of this thesis, concentrates on labor supply. As critics argue, a substantial UBI will eliminate the necessity to work and the increased disposable income might make an individual substitute leisure for work effort with a resulting decrease in overall productivity (Petersen et al., 2011). This, as is further elaborated, might lead to widespread idleness and cause a significant reduction of the labor force with devastating effects on aggregate labor supply in the economy (Hauser, 2006). This prediction is rooted in a classical economic approach to (partial equilibrium) labor market dynamics. In this standard model, an individual chooses an optimal labor effort given her preferences over money income/leisure bundles and her budget constraint. While leisure is seen as a good that the consumer wants to maximize, work effort is conceptualized as a bad that he strives to limit to the level necessary to satisfy his preference for consumption relative

(12)

- 8 -

to leisure. The budget constraint lies in the limited time endowment the consumer has to allocate between the two mutually exclusive activities and his initial money endowment.

In this framework, UBI can be accommodated as an increase in the initial money endowment. As a result of this increase work effort is partly substituted for additional leisure in the post-intervention allocation. This is illustrated in Figure 1, in which an increase in income from to causes an upward shift of the budget constraint of the individual. The new budget hyperplane is tangent to indifference curve ( ) in point . There has thus been a shift from point to point , which features a substitution of work effort ( ) for leisure ( ) caused by the income effect (note that the substitution effect of UBI is equal to zero if we assume that the wage stays constant). Thus the time spent on leisure has increased from to , while the time spent on work decreased by the same amount from to .

Figure 1: Effect of increased income on the division of work effort and leisure

4. Field studies

This section will serve as an overview of empirical evidence based on basic income experiments with varying degrees of unconditionality that have been conducted around the world. The total amount of UBI experiments that have been conducted internationally is a mere four, however, by analyzing effects of similar policies that ensure some form of basic

(13)

- 9 -

income conclusions can be drawn with regard to the possible effectiveness of UBI. There has been a stark contrast between the homogeneity of results obtained in developing countries and the homogeneity of results obtained in developed countries. As mentioned earlier, the

deviation of classical predictions among experiments in developing countries has been mostly attributed to the poverty trap, while the causes of deviation from classical predictions in developed countries have not yet reached a consensus. Therefore, both groups of countries will be assessed separately and the experiments in developed countries will be explained more in-depth for further analysis. Note that the data of the basic income experiments is not openly available, thus all conclusions are drawn from the published research results. Firstly, a short summary of the results of the experiments is presented. Secondly, the results from

experiments in developing countries are discussed. Lastly, an analysis of experimental results from developed countries is provided.

Table 1 is based on several basic income experiments or policies that resemble UBI to a certain degree. Overall, results from developing countries indicate no drop in labor supply along the extensive and/or intensive margin, in contrast to critics’ predictions. Additionally, a growth in earned income and welfare is observed, the latter refers to the quality of life, which incorporates the satisfaction of psychological needs and wants. In addition, a net shift of

employees to jobs with more income security and higher wages is typically observed. In developed countries there has not (yet) been a UBI experiment of which the effects are analyzed, although the Netherlands and Finland will start experiments in 2016 and 2017 respectively. As a result, two NIT programs are analyzed and possible deviations that might occur under UBI are put forward. Additionally, the Alaskan permanent fund dividend and the Cherokee Indians shared profit scheme are assessed because of their resemblance to a partial UBI, meaning that these programs regularly and systematically paid out supplemental income, below the level of subsistence, to all citizens/members. Iran’s redistribution of oil revenues among its citizens is a similar example. Overall, the NIT experiments show a minor drop in labor participation and labor intensity, which was lower than critics anticipated (Munnell, 1987). Analysis of the partial basic income schemes suggests that there is a

predominantly positive effect on labor along the extensive margin and an ambiguous effect in regard to labor along the intensive margin.

Experiments involving income security and poverty add several ethical dimensions. The Namibian UBI experiment chose not to use a control group, because researchers felt that it would be unethical to help some, while standing by as others suffer (Haarman et al., 2009). Of the field experiments and analyzed policies, four were randomized controlled trials

(14)

- 10 -

(RCTs), while another four were controlled trials. The participants in the RCTs were chosen randomly and were located in the same area, e.g. both the treatment and control participants of Mincome lived in Minotoba or Dauphin (Forget, 2011). But the controlled trials studied two separate and comparable regions; in one of the regions all citizens would receive UBI, while the other region would serve as a control group. The researchers tried to eliminate the effect that UBI recipients might have on non-recipients, such as acts of reciprocity towards non-recipients.

(15)

- 11 -

Country Program Conditionality Effect on

job shifting9 Effect on quality of life10 Effect on labor (1) participation and (2) intensity Effect on household income11 Effect on regional economic growth

Namibia BigNam Unconditional positive positive (1) increase (2) increase

positive positive

Kenya Give Directly (RCT)

Only the ultra-poor, one time

positive positive (1) increase (2) increase

positive positive

Uganda Give Directly (RCT)

Only the ultra-poor, one time

positive positive (1) increase (2) increase

positive positive

India MPUCT (controlled)

Unconditional positive positive (1) increase (2) increase

positive positive

India TVUCT (controlled)

Unconditional positive positive (1) increase (2) increase

positive positive

Iran Oil revenue redistributed

Unconditional - positive (1) - (2) -

- positive

Brazil Bolsa Familia (controlled)

Means tested negative positive (1) no change/ decrease12 (2) no change

positive positive

Brazil ReCivitas (controlled)

Unconditional positive positive (1) - (2) increase positive positive Canada Mincome (RCT) Negative income tax - positive (1) decrease (2) decrease no change - United States

Alaskan PFD Only 1 year citizens - positive (1) increase (2) decrease positive positive United States

EBC Indians Only tribe members - positive (1) increase (2) decrease/ increase13 positive - United States 1970’s NIT experiments (RCT) Negative income tax - positive14 (1) decrease (2) decrease negative -

Table 1: Overview of unconditional basic income and similar experiments and their effects

9

When positive, there has been observed job shifts to jobs with higher income or of greater utility.

10_{Quality of life depends on the satisfaction of psychological needs and wants, such as self-esteem, happiness,} safety and access to proper housing, schooling, health care, sufficient food and water.

11_{Household income from work is compared in the situation before and after the introduction of basic income.} 12_{In rural areas there is a significant drop in women’s labor participation, but not for men. In urban areas not} significant drop in labor participation was found.

13_{Households that were persistently poor or never-poor were found to increase the amount of hours worked,} while households that were ex-poor were found to decrease the amount of hours worked.

14_{Overall, schooling, health and other determinants of quality of life were positively affected, but interestingly} enough divorce rates actually significantly increased.

(16)

- 12 -

Developing countries

Iran is the first and so far only country that has a policy that closely resembles a nation-wide (partial) UBI (Tabatabai, 2011). The country distributed part of its oil revenues directly among its citizens in the form of food and fuel subsidies, however, this turned out to be an inefficient and unfair system of price subsidies. As a result, Iran implemented an economic reform plan that substituted the subsidies for direct unconditional cash transfers, in other words, supplying their citizens with a (partial) UBI. Unfortunately, not many of the effects of Iran’s policy have been investigated, which makes it hard to draw any conclusions regarding the net effects of this policy reform.

In contrast, two well-documented controlled experiments from India have more explanatory power. The first experiment was the Madhya Pradesh Unconditional Cash

Transfer (MPUCT) program, which was one of two UBI experiments, that were held between 2011 and 2013 by Unicef and the Self Employed Women’s Association of India (SEWA). Assessments of the experiments were done four times throughout the year in order to gather more insight regarding the temporal effects of a UBI. Out of 20 similar villages, 8 had received UBI and the other 12 served as control groups. The treatment villages significantly improved their basic living conditions, increased the likelihood of having access to sufficient food and water, improved overall health, reduced child labor and increased the number of individuals that received schooling (SEWA, 2014). To be more precise, participants spent more on food and healthcare, children's school performance improved in 68% of families, time spent in school by children nearly tripled and personal savings tripled. With respect to productivity, the individuals that received UBI did spend a little more time on economic activities, however, this result was found not to be significant. Significantly more people in the treatment villages engaged in economic activities and the start-up of new businesses was twice as large as in the control villages, therefore causing a rise in labor participation, especially among women. Recipients of the cash transfers significantly increased income earning work from main and secondary economic activities by 20%. The second controlled UBI experiment was the Tribal Village Unconditional Cash Transfer (TVUCT) program, which compared two tribal villages of which one received UBI and the other served as a control group. The tribals usually live in forest tracts and are considerably poorer than their non-tribal counterparts (SEWA, 2014). Cash grants led to an increase in own-account work, and a relative switch from wage labor to farming and small-scale business. The shift from labor to own-farm work was especially marked in the tribal villages. Treatment households

(17)

- 13 -

were significantly more likely to increase their income from work than control households, in spite of it being a difficult year due to weather conditions in the area. Partly attributed to significantly higher odds of putting in more days at work than households in control villages and the purchase of more livestock to increase production; the treatment village significantly increased its livestock by 70%. Farmers also increased their spending in good quality seeds, fertilizers and pesticides. These two experiments show that there has been a significant shift towards self-employment and investment in agricultural products or manual small scale industry in the treatment groups that would not have been possible without UBI. This is in line with the concept of the poverty trap, meaning that UBI ensures a regular income stream which enables individual’s access to capital that can be invested in more productive activities that would otherwise have been out of reach of the individual. The observation of an

increased earned income after the cash transfers stopped implies that indeed the poverty trap was removed in the treatment group and that the introduction of UBI had long-term poverty alleviating effects. Also, the spill-over effects have enhanced welfare in the region even more.

The Namibian UBI experiment was not a controlled study, however, results indicate reduced child malnutrition increased school attendance, increased the community's income above the actual amount from the grants as the UBI allowed citizens to partake in more productive economic activities (Haarmann et al., 2009). These results are in line with the results from the Indian experiments and suggest the removal of the poverty trap.

According to Standing (2008) one time unconditional cash transfers provide some insight into the dynamics of a UBI and its role in poverty alleviation. He argues that

unconditionally giving cash once is already an effective way to alleviate poverty and reap the benefits of the spill-over effects. This view assumes that the ultra-poor receivers do not resort to idleness and will no longer depend on cash transfers in the long-term. Evidence from several unconditional cash transfer RCTs of the ultra-poor in Kenya and Uganda by the NGO Give Directly has proven that there occur significantly positive changes in assets, earnings, food security, mental health, and domestic violence with no evidence of impacts on alcohol or tobacco use, crime, inflation or drop of labor supply. Also long-term effects of increased steady income have been observed years after the one-time unconditional cash transfer.

The case of Brazil’s income security program Bolsa Família (aimed at increasing human capital among the poor) presents a conditional means tested social welfare program. Still, it provides some insights into poverty reduction and nationally implemented policies in this area, because about 12 million Brazilians receive funds from Bolsa Família, which is about 26% of the total population (IPC-IG, 2013). In line with the previous experiments there

(18)

- 14 -

was an improvement in the job market and real gains on the minimum wage. In rural areas there is a significant drop in women’s labor participation, but not for men. In urban areas no significant drop in labor participation was found. The researchers found individuals that would have been entitled to the Bolsa Família, but were not aware of their eligibility, which is one of the problems of conditional social welfare systems. As a result they showed

similarities to the treatment group and were used as control. Of Bolsa Família recipients, 82.4% reported eating better; additionally, it was reported to increase the incomes of the poorer families by about 25%, but families often chose to be employed in the informal sector, as the income from the earnings from the formal sector were communicated to the

government and one could lose the eligibility for Bolsa Família. This is a problem that UBI, in comparison to income conditional programs, could solve. This would lead to a better allocation of labor into the most suitable job for an individual.

ReCivitas was a small Brazilian controlled UBI experiment running from 2008 until 2011 funded by donations from all over the world (Brancaglione & Pereira, 2012). Unlike the Bolsa Família, this experiment showed a significant improvement of the labor intensity. This suggests that UBI might be better equipped than conditional welfare schemes in reducing poverty. However, the effects on the labor participation were not measured in this study. In line with the other UBI experiments in developing countries, this experiment also

demonstrated the ability of UBI to overcome the poverty trap. Thus, there exists a consensus among researchers that the deviation from the classical predictions of the labor supply model is caused by the poverty trap. Especially in developing countries there has been observed a considerably positive effect of basic-income administration on productivity.

Developed countries

The city of Dauphin in Manitoba, Canada, took part in an experimental guaranteed income program ("Mincome") between 1974 and 1979. This was an NIT program financed by the government. Although there was a control group, there has never been a final report published on this experiment after its completion. However Forget (2011) tries to infer as much as possible from what data is available. She finds that retraction in labor participation occurred mostly among new mothers and teenagers. Mothers with newborns stopped working because they wanted to stay at home longer and students felt less pressure to work as the cash transfer from NIT often substituted the support they received from their family and therefore felt less inclined to keep working. This, however, did result in more teenagers graduating.

(19)

- 15 -

Secondary and tertiary wage-earners showed a mild reduction in the amount of hours worked, while primary earners showed little change, implying a slight drop in labor intensity. This resulted in a drop of 1 percent in the amount of hours worked among men and a drop of 3 and 5 percent for married and unmarried women respectively. It was found that in the period that Mincome was administered, hospital visits dropped 8.5 percent, with fewer incidents of work-related injuries, and fewer emergency room visits from car accidents and domestic abuse. Additionally, the period saw a reduction in rates of psychiatric hospitalization, and in the number of mental illness-related consultations with health professionals.

Even though, positive spill-overs in health-care occurred, they are partly offset by a slight drop in work-force participation and intensity. However, labor retraction occurred mostly among students that spend additional time on studies and new mothers that spend additional time on raising their child, these are groups for which it can be socially beneficial to refrain from working to some degree. A normative question resulting from this analysis is whether minor inefficiencies in the labor supply weigh up against the spill-over, real freedom and social justice arguments put forward by proponents. Van Parijs (2004a) argues that UBI in comparison to NIT is better equipped to overcome the unemployment trap. Although NIT phases out benefits more slowly than unemployment benefits, individuals still face a

disincentive to work as additional income leads to a decrease of benefits, which is not the case under UBI. However, Gilroy et al. (2012) expect that in order to finance UBI a higher

marginal income tax rate is needed then when a NIT is financed. This causes a disincentive for working individuals to spend more time on work effort. Future experiments might show which effect is stronger.

Throughout the 1970’s there have been multiple NIT experiments in the US, namely in New Jersey, Pennsylvania, North Carolina, Indiana, Seattle and Denver. Munnell (1987) has analyzed the overall effects of these experiments and has concluded that there is a mild drop in labor intensity among men (on average 7 percent) and a larger loss of labor intensity among women (on average 17 percent). This has resulted in a drop of earned income of households. However, many households underreported their income to avoid higher marginal income tax rates of the phasing out of the NIT. Under UBI reports would be incentive-compatible due to its unconditionality. Munnell (1987) argues that a moderate drop in labor intensity will not jeopardize the national supply of work effort. The effect on labor participation was negative, in the treatment and control group, because of high unemployment levels caused by a lack of labor demand. Again the treatment group had a mildly greater negative effect than the control group.

(20)

- 16 -

In Alaska, the Permanent Fund Dividend (APFD) is a dividend paid to Alaska residents that have lived within the state for a full calendar year and intend to remain an Alaskan resident indefinitely. The amount of each payment is based upon a five-year average of the Permanent Fund's performance and varies widely depending on the stock market and many other factors. The APFD has existed since 1982. Howard and Widerquist (2012) and Goldsmith (2010) assess the suitability of the APFD as an example for basic income, but because the whole state received the fund there is no control group. According to these authors most of the cash from dividends will ultimately find its way back into the Alaskan economy, which has increased employment, population (through migration), and income. A rough estimate of the total (direct and indirect) macroeconomic effects of this increase in purchasing power is 10 thousand additional jobs, 15 to 20 thousand additional residents (drawn to the state because of the jobs), and $1.5 billion in additional personal income. There is thus a deviation from critics’ predictions of basic income with regard to the labor supply along the extensive margin. The external economies of scale and increased expenditures have caused a rise in labor demand. However, there has also been a one percent decrease in labor intensity. The Alaskan dividend was not high enough to ensure means of subsistence and is therefore more comparable to a partial UBI. If a full UBI would be implemented, then there might have been larger inefficiencies in the labor supply. Still, this policy shows that the Alaskan economy benefitted from these dividends, which might also happen in the case of UBI, as is observed in all UBI experiments in developing countries.

Another uncontrolled, originally not basic-income oriented, study has focused on members of the Eastern Band of Cherokee Indians, based in North Carolina. They received payments of several thousand dollars twice a year. These payments are dividends from the profits of the Harrah's Cherokee casino, and have been distributed since 1996. A study by Angold et al. (2003) of the payments' effects on the children of the community found significant declines in poverty, behavioral problems, crime, substance abuse and psychiatric problems, and increases in on-time graduation. The effects were primarily found among those who were youngest when the payments began, and among those who were lifted out of poverty rather than those who were already well-off, which implies that some form of basic income does increase human capital of especially the less fortunate. The effect on the labor intensity differed among types of households. Households that were persistently poor or never-poor were found to increase the amount of hours worked, while households that were ex-poor were found to decrease the amount of hours worked. Furthermore, for households with two parents, the labor participation rates rose. Overall, the net effects were on the labor

(21)

- 17 -

supply lean towards the positive. However, this study on the Eastern Band of Cherokee Indians is not suitable to base generalizations for the United States on, because the group dynamics and role in American society differs from the average American.

To summarize, mild reductions under NIT schemes in labor participation and intensity suggest that this might also hold true for UBI. UBI proponents argue that UBI will have a more positive effect than NIT, especially the safeguarding of real freedom and self-determination as well as the unemployment trap sets these two policies apart. Intrinsic

motivation may thus provide useful in explaining differences between NIT and UBI results in the future. In the case of Alaska and the Cherokee tribe, the results suggest a more positive effect of UBI on the labor supply along the intensive and extensive margin and might indicate that UBI indeed results in less labor supply inefficiencies. However, it remains unclear

whether the introduction of a full UBI leads in significantly different results.

5. Extended classical models

Several authors have proposed models that contribute to accommodate the effects of a UBI on a variety of variables of interest that were not generated by the previous classical model. Firstly, Gilroy et al. (2012) theoretically assess the effects of implementing a UBI on the labor supply side and compare it to a situation under absence of welfare programs and a conditional social welfare system. The analysis assumes that individuals maximize Cobb-Douglas utility function with heterogeneous weights that can be interpreted as individual-specific leisure and consumption affinities. This results in the following formulas if there are no social benefits:

Where utility is determined by consumption and leisure and where represents the nominal net wage level, represents the price level, ̅ represents the time constraint and represent the individuals affinity to consumption and leisure, respectively. Figure 2a shows three individuals (A, B, C) with a different leisure and consumption affinity in the situation of no social security program (NSS), based on a budget constraint with a slope equal to the wage. For a UBI system, the formulas become:

(22)

- 18 -

The individual can now consume based on her income from work and her basic income . Beyond a certain level of basic income she decides to refrain from working altogether. Figure 2b shows three individuals (A, B, C) with a different leisure and consumption affinity in the situation of UBI compared to a NSS. The slope of the budget line becomes flatter, as in this case the employed finance the UBI program either directly through a rise in income tax (resulting in a decrease of their nominal net wage), or indirectly through a rise in value-added tax (resulting in an increase of the price level). At level of consumption the individual has a strong affinity towards leisure and will decide to refrain from working. For a conditional social security system, the formulas slightly change:

If the individual decides to become or remain unemployed (leisure is equal to the time constraint), then she receives benefit , which she then uses for her consumption. If social security is bigger than a certain threshold , then she will decide to refrain from working. Figure 2c shows three individuals (A, B, C) with a different leisure and consumption affinity in the situation of a conditional social security system (CSS) compared to a NSS. At level of consumption the individual has a strong affinity towards leisure and will decide to refrain from working. Figure 2d shows three individuals (A, B, C) with a different leisure and consumption affinity in the situation of a CSS compared to a UBI and for simplicity the UBI is equal to the unemployment benefit. Individuals with a high affinity (A) for consumption will spend a comparable amount on leisure under UBI and under CSS, while individuals with a medium affinity for consumption (B) will choose to work less under UBI than under CSS and individuals with a low affinity for consumption (C) will choose to work more under UBI than under CSS.

(23)

- 19 -

Figure 2: Consumer behavior under absence of welfare programs (a), changes under

unconditional basic income (b), conditional benefits (c) and comparison of consumer demand under basic income and conditional benefits (d) (Gilroy et al., 2012)

The underlying substitution and income effects are summarized in table 2. Gilroy et al. (2012) conclude that the effect of UBI and CSS on the labor supply depends on the

consumption/leisure preference of an individual. They argue that among the individuals with a low consumption affinity, there is an increase in the work effort. More specifically,

individual C with a low consumption affinity will choose to exert no work effort under CSS, while the same individual C will choose to exert a positive level of work effort under UBI. This model provides a theoretical solution of how UBI can overcome the unemployment trap that is inherent to CSS. As a result the net effect of UBI, compared to CSS, on the labor supply might be positive. Although consumption affinities most likely play a role and can be

(24)

- 20 -

attributed to explain part of a net effect on the labor supply, the behavior of individuals also depends on many other factors. The previous sections introduced a solution of the

unemployment trap based on other criteria then consumption affinities that arguably fit the empirical results better, namely the potential loss of eligibility for social welfare benefits. Participants of the NIT experiments in the US and recipients of the Bolsa Familia both

systematically under reported their income and in the case of the Bolsa Famila recipients even rejected better paying, more secure jobs because of a loss of entitlement by accepting them. The addition of intrinsic motivation to the model may provide a useful mechanism to explain labor market dynamics. Firstly, under UBI the problem of the Bolsa Familia is not existent and individuals can choose the job that is most intrinsically satisfying. Secondly, intrinsic motivation can explain how even individuals with a high consumption affinity might not decrease their amount of hours worked under UBI.

Table 2: Effects of different welfare systems on the labor supply (Gilroy et al., 2012)

Secondly, Gamel et al. (2006) also base their analysis on microeconomic theories regardinglabor supply. However, the conclusions they draw based on these theories with respect to the idleness argument do not coincide with the conclusion of the opponents of UBI. In their model they make a distinction between two settings, in the first setting leisure time is a normal good and in the second setting leisure time is an inferior good. Furthermore, the model compares an initial situation with unemployment benefits to a new situation with UBI and increased income is assumed not to alter the price of leisure, which is the opportunity costs associated with the wage rate.

If leisure is perceived as a normal good, then employed individuals will (partly) substitute the amount of hours worked for leisure, because there is a rise in the initial endowment due to the additional UBI which will yield a positive income effect on leisure time consumption at the cost of work effort. Since prices, including the opportunity cost of leisure, remain constant, there is no substitution effect. For unemployed individuals that are

(25)

- 21 -

currently receiving unemployment benefits it is not clear what effect a shift to UBI will have on leisure time consumption, because there is a negative substitution effect on leisure time (the opportunity cost of leisure are now equal to the wage instead of zero15) and a positive income effect on leisure time (as long as the UBI is higher than the unemployment benefit).

If, on the other hand, leisure time is perceived as an inferior good, then employed individuals will decrease their leisure time, because there is again rise in income at constant prices (negative income effect, no substitution effect). Unemployed individuals will

unambiguously have the incentive to work, because the income and substitution effect of a shift from unemployment benefits to UBI on leisure time will both be negative, meaning that the time spent on leisure will decrease. A negative income effect on leisure time as leisure is perceived as an inferior good (as long as the UBI is higher than the unemployment benefit) and a negative substitution effect on leisure time as (the opportunity cost of leisure are now equal to the wage instead of zero). Thus, the effect of UBI depends on how people perceive leisure time and differs between those currently employed and those currently unemployed, which is summarized in table 3.

Employment status Leisure time as normal good Leisure time as inferior good Employed Income effect: positive16

Substitution effect: none Overall effect: positive

Income effect: negative17 Substitution effect: none Overall effect: negative Unemployed Income effect: positive

Substitution effect: negative Overall effect: ambiguous

Income effect: negative Substitution effect: negative Overall effect: negative Table 3: effects of the implementation of UBI with respect to unemployment benefits

If leisure time is perceived as an inferior good by individuals, then the net effect of the introduction of UBI is negative for both initially employed and initially unemployed

individuals when compared to unemployment benefits. This means that unemployed individuals, who would choose not to work under unemployment benefits and who would choose to work under UBI, perceive leisure time as an inferior good, which offers a possible explanation of how UBI can overcome the unemployment trap inherent to unemployment

15_{With unemployment benefits, “the cost of increased leisure time is zero due to the confiscatory marginal tax} rate (100%) for those individuals who are at or below the guaranteed minimum level” (Gamel et al., 2006) 16_{Positive effects represent a rise in the amount of leisure time consumed}

17

(26)

- 22 -

benefits. In order to assess the overall effect of a UBI, which depends on how individuals perceive leisure time, Gamel et al. (2006) used a questionnaire for a sample of young (ages below 25) school drop-outs with a low level of education that often end up in lousy low-paid jobs. There were not enough unemployed respondents, but there were sufficient employed respondents to base an econometric analysis upon. The results of the questionnaire showed that leisure time is not regarded as a normal good for two thirds of the respondents and that the greater the stability of a job, the greater the possibility of not working less. These results suggest thatindividuals’ perception of work and leisure might vary according to their position in society in a way that systematically affects the potential outcomes from introduction of UBI.

Gamel et al. (2006) compare the effects of several income securing policies, namely unemployment benefits (UEB), NIT and UBI. Similarly to the analysis above the substitution effect of all social welfare systems is evaluated, as can be seen in figure 3, but this time the policies are introduced ex nihilo (out of nothing). The substitution effect of UEB and of NIT is both positive. Moreover, of these three policies UBI is the only program in which there is either no (for employed individuals) or a negative (for unemployed individuals) substitution effect. Regarding the income effect, the outcome is based on how the individual sees leisure time. Leisure time is seen as a normal good when the individual perceives work as being a burden, when there is a negative stigmatization of the job, when social integration is mainly focused on non-market networks (friends and family) and when an individual invests full-time in job search and home study. On the other hand, if leisure time is perceived as an inferior good,then the work is seen as gratifying in itself, then others hold the job in high regard, then social integration is mainly focused on market networks (professional relationships at work) and then a job is a way to gain better opportunities through a professional network and on-the-job training. Furthermore, it is assumed that low wage earners with large amounts of leisure will prefer to substitute it for work and increase their purchasing power (negative substitution effect outweighs positive income effect), while beyond an individually specific pay threshold the individual starts to prefer leisure over work (positive income effect outweighs negative substitution effect). Thus, according to this model UBI would result in a negative substitution and income effect and outperform the other policies as long as leisure is an inferior good. The model proposed by Gamel et al. (2006) takes intrinsic motivation into consideration. High intrinsic motivation for doing a job means that the individual will perceive leisure time as an inferior good, which will have positive effects on labor supply. This is in line with the predictions of van Parijs (2004b) in section 2. However, intrinsic motivation directly and

(27)

- 23 -

indirectly affects labor supply in more than one way. Section 6 will elaborate on the role of intrinsic motivation.

Figure 3: Substitution effects of unconditional basic income (a), unemployment benefits (b) and negative income tax (c)

6. Intrinsic motivation

This section aims to clarify the role that intrinsic motivation plays in labor market dynamics. It is common knowledge that when people are intrinsically motivated to do something they will exert more effort and this is certainly expected to apply for aspects of an individual’s job that might be rewarding in themselves. However, classical economic theories only take intrinsic motivation into consideration in the sense that the initial level of effort on a task is high or that cost of effort is low and see external incentives as the way to increase productivity, such as (monetary) rewards and penalties (Pech, 2010). In which only two external stimuli exist, either rewards for displaying a certain effort level (carrots) or penalties for falling short of some effort level (sticks). This view is simplistic and does not cover the complex dynamics of human behavior, insights from the fields of psychology, sociology, business and management are used to create a more accurate description of intrinsic motivation. Pech (2010) takes a rather different approach and disregards classical views of work incentivization. He introduces the motivation crowding-out effect to a basic-income scheme, this effect refers to the undermining of intrinsic motivation by external stimuli. Frey and Jegen (2001) sum up several laboratory experiments that related labor supply with motivation crowding theory. All of which are completely different settings then the model that Pech (2010) proposes. For instance, an experiment in which participants receive different rewards for making a test, it is shown that the height of the reward affects the test score, namely a very low reward causes significantly lower test scores then no reward, which

(28)

- 24 -

implies that small rewards crowd out intrinsic motivation (Frey and Jegen, 2001). Arguably, there is no real evidence on which Pech (2010) bases his labor model upon.

Nevertheless, insights from Pechs (2010) analysis will provide useful for determining effects of intrinsic motivation alone on labor supply. The model differentiates between good and bad jobs where good jobs are those which hold high intrinsic value for the worker and bad jobs are those which hold low intrinsic value for the worker. When there are no or inadequate social security systems a worker might be employed in a bad job in order to achieve income security, while actually preferring another good job that, however, would not yield a sufficiently high income under the current wage schedule. Now, under a UBI system all individuals have income security and no longer face the unemployment trap, which leads to a shift from bad jobs to good jobs and increased labor participation, as was described by van Parijs (2004b). As a result, the wage in good jobs decreases (increased labor supply) and the wage in bad jobs increases (decreased labor supply), as is shown in figure 4.

Figure 4: Labor supply shift in no (A) and high (B) intrinsically motivating jobs (Pech, 2010)

Centers and Bugental (1966) study the role of extrinsic and intrinsic job components for individuals’ job satisfaction and the value they attach to their job by interviewing workers. Intrinsic components, such as opportunity for self-expression, interest-value of work, meaning of work and social interaction with colleagues, were found to be more valued more at higher occupational levels. In contrast, extrinsic components, such as wage and contract length were found to be more valued more at lower occupational levels. As a result an employee in a high level job might be expected to react better to intrinsic rewards, while an employee in a low level job may react better to extrinsic rewards. If this relation is combined with Pech’s (2010) model, then those individuals that are still employed in bad/low-level jobs are now

(29)

- 25 -

compensated more and might be more productive based on external incentives (increased wage), while those that chose to be employed in a good/high-level job might be more productive based on internal motives and are less affected by external disincentives

(decreased wage). As a result, productivity of both kinds of jobs would rise. This may explain positive deviations from critics’ predictions in regard to labor participation and intensity. This suits van Parijs’ (2004b) idea that under UBI an individual has the right and means to use her real freedom and can discriminate between jobs that yield different levels of intrinsic

motivation, which subsequently may increase one’s well-being.

However, Pech (2010) argues in line with the motivation crowding-out effect, the individual that chooses to remain in a bad job is not incentivized by intrinsic motivation and will thus be motived by external motives. The concavity of the slope in figure 5 implies that the higher the wage, the higher the motivation which decreases as wage grows larger.

Depending on the original wage of the individual, she will increase her effort by a little of by a lot when the wage increases. In contrast, the individuals that choose to be employed in good jobs have more intrinsic motivation to do that job, but are also compensated less, as the wage decreases. Additionally, due to motivation crowding-out, effort in high intrinsically

motivating jobs has the shape of a valley parabola, as can be seen in figure 5. For small increases in wage (starting wage at zero) the external rewards crowds out intrinsic motivation and decreases effort. For a large increase, most intrinsic motivation is lost and the effort becomes again determined by the height of the wage. Thus, under motivation crowding, a UBI would increase effort of individuals that remain in bad jobs and would have an ambiguous effect on individuals that are employed in good jobs.

(30)

- 26 -

Gamel et al. (2006), on the other hand, disagrees with the interpretation of leisure as normal good implicit in the afore-mentioned studies and explains how, if leisure is seen as an inferior good, the implications of UBI changes. According to this model UBI would result in a negative substitution and income effect and outperform NIT and UEB. One of the factors that determine whether leisure is seen as inferior good is intrinsic motivation associated with a job, which will have positive effects on labor supply along both margins. Deci and Ryan (2000) refer to intrinsic motivation as the satisfaction of innate psychological needs, such as the needs for competence and autonomy, which resembles van Parijs’ (2004b) notion of real freedom, which is based on the means and right to make one’s own choices. Competence and autonomy are ensured under UBI in the same way as real freedom is ensured and will as a result increase an individual’s satisfaction of doing a task, in other words, the individual is more intrinsically motivated. Additionally, Deci and Ryan (2000) identified relatedness and social interaction to be a determinant of intrinsic motivation, which was also put forward by Centers and Bugental (1966) and Gamel et al. (2006).

A study, performed after the 9/11 attacks on the World Trade Center occurred in the United States, found that after the attacks many Americans decided to change jobs

(Wrzesniewski, 2002). Such a disaster plausibly makes people reassess some of their live choices, something that does not occur on a daily basis. Some people decided to renew family ties, to reestablish friendships, to more intensively practice their religion, to be thankful for what they had taken for granted and/or to reexamine the meaning of their job for them and their surroundings. The meaning of work is highly related to intrinsic motivation and is less affected by extrinsic values. Wrezinski (2002) shows by analyzing job shift patterns after 9/11 that when people have the possibility to think about the meaning of their job, then job shifts often occur towards more intrinsically motivating jobs. One of the objections to switching jobs is often a lack of income security stemming from the risk of such a change. UBI could eradicate this risk and make sure that individuals are able to think about what kind of jobs hold meaning and satisfaction for them without the need of a disaster.

Similarly, UBI can take away the risk of income volatility of being self-employed. In developing countries many small businesses were successfully set-up after the introduction of UBI. Baumgartner and Caliendo (2008) find that employment programs that focus on self-employment rather than self-employment in an existing firm perform significantly better in regard of the probability of staying employed and the level of personal income based on

administrative data and surveys in Germany. The freedom of self-determination is seen as an intrinsic motivator that may cause an individual to work harder if she is self-employed.

(31)

- 27 -

Furthermore, Burgess and Kingston (1977) argue that the success of reemployment is dependent on the suitability of a job, in other words, the degree to which in individual finds satisfaction from the job. In addition, in their analysis the length and height of unemployment benefits increased the likelihood of finding a suitable job. In relation to intrinsic motivation, suitable jobs are those with high intrinsic motivation. In addition to overcoming the

unemployment trap, UBI can prevent a mismatch of talents in the wrong job, which implies that an individual that has the means to perform a more intensive search for a suitable job will when she finds a job have more reemployment success and thus less likelihood of reentering unemployment.

Thus based upon the literature a job is highly intrinsically motivating when it

incorporates several of the following principles: autonomy, (real) freedom, relatedness, social interaction, meaning, inherent satisfaction, competence, career management,

self-determination and talent. The introduction of UBI satisfies these intrinsic phenomena, which could imply that in the model of Gamel et al. (2006) it is not unlikely that leisure is perceived as an inferior good and that both employed individuals will keep working and unemployed individuals will start working.

7. Limitations and considerations

There are several limitations and considerations relating to the analysis of UBI and the role of intrinsic motivation. First of all, there is a reason that intrinsic motivation is often overlooked in economic literature, it is hard to distinguish which part of motivation is caused by intrinsic or extrinsic components, as a result intrinsic motivation is often ignored due to difficulty of formalization even though it plays an important role in economics and labor (Frey & Jegen, 2001). Although several of the assessed experiments talked about competence, relatedness and autonomy, these are vague hard to measure variables, that cannot be easily measured and compared to other studies.

More importantly, there is a lack of available experiments that specifically involve UBI. Analysis of NIT or partial basic incomes creates a source to base possible policy

implications of basic income upon, but several models show that the effects of these programs differs. Due to a lack of empirical evidence, it cannot yet be tested whether these models are indeed valid. However, in 2016 and 2017 there are experiments starting in the Netherlands, and Finland respectively and a referendum is held for a nationally implemented UBI in

(32)

- 28 -

Switzerland. By the time the Dutch and Finnish experiments are implemented and analyzed, then this offers many possibilities to finally assess empirically many of the debated issues.

In the meantime several studies have been conducted to (theoretically) compare the effectiveness and efficiency of UBI compared to other social welfare programs. Bryan (2005) compares a situation of UBI to a situation of targeted means-tested social welfare benefits with eligibility requirements in the case of the United States. Using a simple simulation, the author shows UBI might be a more efficient way of redistributing wealth, while also ensuring income security. However, he also notes that the tax revenue that needs to be raised in order to finance UBI, in comparison to targeted programs, can put a strain on political support and economic efficiency of the policy.

Tcherneva and Wray (2005) and Cowling et al. (2006) focus on unemployment reduction policies and compare UBI to state-provided Job Guarantee (JG) programs. Both policies are aimed at ensuring income security, but the authors argue that JG programs have a more effective and sustainable way to reach full-employment. They also note that it is

important to analyze the causes of why an individual remains unemployed.

Davutoğlu (2013) compares micro-finance and basic-income schemes as tools of poverty alleviation. Micro-finance has become widely used in the developing world in recent times and plays an important role in the approach of the World Bank towards poverty

reduction. Criticism has focused on a failure of micro-finance in reducing inequality, the fact that micro-finance is market-based and will usually be a profitable arrangement profitable have been cited as positive aspects. Conversely, UBI is held to ensure social justice and combat inequality better, but is harder to finance.

Pech (2010) compares a negative income tax (NIT) to UBI. He finds that a NIT is politically more feasible and requires less financing while having a similar net effect on income. However, van Parijs (2004a) argues that NIT is less equipped to overcome the unemployment trap.

One of the problems that a national UBI solves is the problem that all other small scale experiments are usually short-term. The decision to quit a job is not taken lightly and if there is no certainty there might not be a retraction out of the labor force of an individual that strongly prefers leisure. There is thus a risk that the labor participation will drop more than expected in a fully implemented UBI. Additionally, many experiments were not financed by those receiving the benefits, as a result there has not been an increase in the marginal

(income) tax rate that is often assumed with the implementation of UBI. If this tax increase is integrated, then there might be an additional disincentive to work.

The effects of unconditional basic income on labor supply and the role of intrinsic motivation : theoretical analysis of how the introduction of an unconditional basic income affects the labor supply and an assessment o