Skip to content
Publicly Available Published by De Gruyter November 3, 2015

The Effect of Unconditional Cash Transfers on Adult Labour Supply: A Unitary Discrete Choice Model for the Case of Ecuador

  • Andrés Mideros EMAIL logo and Cathal O’Donoghue
From the journal Basic Income Studies


We examine the effect of unconditional cash transfers by a unitary discrete labour supply model. We argue that there is no negative income effect of social transfers in the case of poor adults because leisure could not be assumed to be a normal good under such conditions. Using data from the national employment survey of Ecuador (ENEMDUR) we estimate the effect of the Bono de Desarrollo Humano (BDH). Results show that cash transfers, unconditional in labour, does not produce labour disincentives in the case of household heads, but may be paying for housework and childcare provided by partners and single adults. However, labour market and care work gender inequality must be addressed by complementary policies.

1 Introduction

Income maintenance schemes can be classified by their basis (cash or work) and the method (conditional or unconditional) of provision. Jackson (1999) identifies four (pure) possible schemes. First, unemployment benefits provided in cash but conditioned on employment status. Second, basic income provided in cash to all citizens as an entitlement. [1] Third, guaranteed work provided to anyone on unemployment; in this case the State takes the role of employer of last resort for specially created jobs. Finally, unconditional basic work offering a minimum amount of work as a duty attached to citizenship (1999). All these alternative schemes have as objectives to guarantee a certain income level and to promote labour. [2]

Unconditional cash transfers are a mix mechanism as it is not universal but neither is it conditioned to employment status. In some cases they are seen as a potential first stage for the implementation of basic income in developing countries (see, e.g., LoVuolo, 2012; Standing, 2008; Suplicy, 2003; Van Parijs, 2004). This paper generates a theoretical framework and provides empirical evidence on the effect of unconditional cash transfers on adult labour supply. It is a key discussion regarding the economic effects of cash transfers, because non-negative labour effects are desired in order to foster positive economic returns.

Next to the positive effects of social transfers on social outcomes, [3] they also foster economic performance, at different levels. At the micro level, social transfers alleviate credit constraints by enabling savings, investments and access to credit. They contribute to consumption and assets security, by helping poor households and individuals to access economic opportunities. Finally, social transfers help to cover transaction costs and to protect assets and facilitating their accumulation, enhancing labour supply and fostering local economy effects (Alderman & Yemtsov, 2012; Barrientos, 2012; Mideros, Gassmann, & Mohnen, 2013; Tirivayi, Knowles, & Davis, 2013).

The empirical evidence regarding economic effects of social protection is inconclusive. In the case of labour effects while there is agreement on the reduction of child labour, [4] in the case of adults there still theoretically ambiguity, and then it remains an empirical question (Alzúa, Cruces, & Ripani, 2013). There are at least four channels to explain the effects of social transfers on labour supply. First, a pure income effect due to the increment on non-labour income may reduce labour supply, but the transfer may also help to cover transaction and opportunity costs increasing labour supply. Second, conditionalities may enforce behavioural responses. If children have to go to school it may free up time used for childcare. Third, if child labour is reduced, adults may increase their labour supply in order to compensate for the reduction of income. Finally, spill-over effects may affect non-beneficiary households and the local economy (2013).

Posel et al. (2006) found that a person, between 15 and 50 years old, living in a household with a non-contributory pension recipient has a 3.2 percentage point higher probability of employment in South Africa. They relate this effect to the possibility to cover migration costs and increased support from grandmothers to childcare. In the case of Mexico, Skoufias and Di Maro (2008) exploiting the experimental deign of the cash transfer programme PROGRESA found no significant effect on adult labour force participation and leisure time, but substantial reduction in poverty.

Foguel and de Barros (2010) found positive effects of a conditional cash transfer programme on male labour participation in Brazil. Similar results were found by Barrientos and Villa (2013) using data from Colombia. They found marginal positive effects on the participation of males and single adults with children and on employment of women in formal jobs. In addition, evidence suggesting no disincentives to work has been also found in Argentina, Uruguay and Chile (Maurizio & Vázquez, 2014) and in Ethiopia, Bangladesh (Barrientos & Nino-Zarazua, 2010) and Cambodia (Mideros et al., 2013). Some negative effects have been found by Fernandez and Saldarriaga (2014) in the case of Peru. They found a reduction on working hours in the week following the pay date (short-term effect). However, they did not find any significant long-term effect on labour participation.

Finally, Alzúa et al. (2013) comparing results from three different experimental evaluations done in Mexico, Nicaragua and Honduras found no statistically significant effect of social transfers on labour supply, but a positive labour supply effect and increase in wages in some specific cases. Using the same data Novella et al. (2012) did not find any significant labour effect in the case of Honduras, positive effect on working hours of males but negative for female labour participation in Mexico, and negative effect on working hours of males but no effect on labour participation in Nicaragua.

In this paper, we estimate a unitary discrete choice labour supply model for the case of the Ecuadorian Bono de Desarrollo Humano (BDH). It is an unconditional cash transfer programme introduced in 1998 by the Government of Ecuador. The BDH is not conditioned on labour. However it has conditions on children health care and school attendance, but the accomplishment of those requirements is not supervised. Because of that reason the BDH is considered as a cash transfer with soft conditions, which can be related with a concept of basic income targeted to the poor (Cecchini & Martínez, 2011). Currently, the BDH is targeted using a proxy-means test index and provides a flat transfer (USD 35 per month in 2012). [5] The case is relevant to analyse labour effects of social transfers in developing countries where this kind of instruments are largely being implemented as a poverty reduction strategy.

The rest of the paper is organized as follows: Section 2 discusses the theoretical framework. Section 3 presents the data and the empirical strategy. Section 4 presents the results and final remarks are presented in Section 5.

2 Theoretical framework

Receiving a social transfer increases household disposable income and subsequently affects the labour supply of its members. Following mainstream labour supply theory it can be argued that social transfers discourage labour due to the income effect. Additional non-labour income promotes more leisure (less work) and more consumption. However, this idea assumes that a person can work as much as she wants, and that leisure is a normal good. These are unlikely assumptions in the case of individuals in poor households.

Figure 1 presents the income effect of an increase in non-labour income. [6] Panel A shows the general case where an increase in real income due to a non-contributory social transfer moves the budget constraint from AB to DE (transfer size is AD which is equivalent to BE) allowing an individual to increase both consumption and leisure, and then reducing labour time allocation. [7] This produces an increase in the level of utility moving from U0 (at point C) to U1 (at point F).

Figure 1: Income effect on labour supply.
Figure 1:

Income effect on labour supply.

However, it is likely to assume a minimum level of consumption (Cmin) below which there is no normal trade-off between leisure and consumption, but instead a person is willing to give up leisure for some more consumption. It is hard to think on a person who values leisure if she cannot satisfy her basic needs. In this case time allocation will result in a corner solution at the maximum level of possible work-time (minimizing the gap to achieve Cmin, with horizontal indifference curves). It is the case of point B in panel B. Then, if a social transfer is enough to reach Cmin, it is likely that a person will assign time for leisure reducing labour by moving to point H in panel B. On the other hand, if the transfer is not enough to reach Cmin, the effect is higher consumption without any change on labour-leisure time assignation, moving to point E´ as in panel C.

In addition, access to and opportunities on the labour market may be constrained because of labour demand limitations but also due to personal and household conditions (i.e. opportunity and transaction costs). It means a person may not be on paid-labour as much as she wants or needs. It is the case at panel D, where the budget constrain is restricted to the segment AC (it is not possible to allocate more time to labour than that for point C). In this case social transfer does not affect labour at all, if Cmin stills not reachable. The result of introducing a social transfer is a movement from point C to point G.

Finally, if social transfers help individuals to overcome labour constrains by, for example, covering transaction and caring cost, financing labour search or acquiring productive assets, the final effect may be positive on labour supply. [8] While the budget stills restricted without a social transfer (segment AC in panel E), it is not with the transfer (budget constrain becomes DE). Result is a movement from point C to point H in panel E.

Given this theoretical framework, social transfers may negatively affect labour supply in the case of individuals with income/consumption higher than Cmin, while producing positive or zero effects in the case of persons with income/consumption below it. In this sense, our hypothesis is that social transfers produce no-negative labour effects in the case of poor adults.

2.1 The unitary discrete choice model

We rely on an unitary discrete choice model of household labour supply, which has been developed following van Soest (1995) and van Soest et al. (2002), and is widely used in the literature for similar analyses (e.g. Aaberge & Colombino, 2013; Beninger, Laisney, & Beblo, 2007; Bloemen, 2010; Blundell & Shephard, 2012; Breunig & Gong, 2010; Dagsvik, Jia, Kornstad, & Thoresen, 2014; Haan, 2004; Kabátek, van Soest, & Stancanelli, 2014; Kornstad & Thoresen, 2007; Löffler, Peichl, & Siegloch, 2013). It is quite intuitive that labour decisions are constrained on the choices of jobs and working hours, and then we prefer a discrete rather than a continuous labour supply model. Furthermore, as this study aims to analyse the effect of a social transfer delivered to households, we base our model on a household utility function (unitary model) rather than an individual utility function (collective model). [9]

We define household utility as a function of a couple’s time allocation and household income. We assume zero leisure (tils=0) in the case of poor households (those with a level of consumption below Cmin). Further, given that total time endowment (T) is fixed (at a maximum of 24 h per day) we take paid-labour participation as the decision variable (til) for the household’s head (i=h) and her partner (i=p). In this sense, housework (including care-work) time is the complement in the case of poor adults (tihw=Ttil), and a mix of housework and leisure in the case of non-poor adults (tihw+tils=Ttil). Paid-labour participation options are defined by the elements of the choice set (L). In addition, we decompose total household income (yj) into labour income for the household head (whthl) and her partner (wptpl) considering income from wages (wi), social transfers (St), and other non-labour income (Y0).

The budgeting problem is then described by eq. [1]. The utility function Uj. is expected to increase with income; but to decrease with labour if household income is equal to or higher than Cmin, while it is independent of labour otherwise.


3 Data and empirical strategy

We use data from the Urban and Rural National Survey of Employment, Unemployment and Underemployment (ENEMDUR) of the National Institute of Statistics and Censuses (INEC) of Ecuador. [10] The ENEMDUR is a cross section survey conducted quarterly for urban households and twice a year (June and December) for urban and rural households, excluding the Galapagos Islands. The sample framework is representative at the province level in the case of the Coast and the Sierra regions, and at the regional level in the case of the Amazon. The ENEMDUR’s first objective is to collect labour and income data, but it also provides relevant information about individuals and households. Available income information includes, but is not limited to, social transfers. In this paper we use the round of December 2012. The sample includes 73,686 individual observations within 19,840 households; using weights the sample represents the national population, accounting for 14.7 million inhabitants in 3.9 million households.

The BDH is a monthly cash transfer targeted at deprived households by consumption, using a proxy means test index. [11] Targeting was done in 2008-2009 and it has not been updated afterwards. The BDH is part of household disposable income, and thus it affects poverty and inequality measures. The BDH (USD 35 per month, in 2012) reduces the extreme poverty head count by 2.9 percentage points, the poverty head count by 2.7 percentage points and the Gini coefficient by nearly 2.0%. [12] At December 2012, extreme poverty head count was 11.2%, poverty head count 27.3%, and the Gini coefficient 0.477; if the BDH is discounted these figures increases to 14.1%, 30.0% and 0.486, respectively. Poverty and extreme poverty lines are USD 76.35 (2.55) and USD 43.03 (1.43) monthly (daily) per-capita, at December 2012. [13]

As aforementioned the BDH accomplish the basic income’s principles of unconditionality on labour and of being paid in cash on a regular basis. While it is not universal neither individual, it was targeted based on past characteristics but up to the date of our data there were not entry and exit procedures. Because of these reasons our study of the BDH’s effects on adult labour supply generates insights for the analysis of unconditional basic income.

There are some studies evaluating the effects of the BDH on different dimensions of wellbeing (Figure 2). However, the effect on labour supply and economic returns has not been analysed. An exception is the study of Gonzales-Rozada and Llerena (2011), who studied the duration of unemployment for those around the eligibility threshold (exploiting a regression discontinuity setting). They found that the BDH may be financing job search which is related with longer periods of unemployment but with higher future income.

Figure 2: Effects of the Bono de Desarrollo Humano in Ecuador (Literature review).
Figure 2:

Effects of the Bono de Desarrollo Humano in Ecuador (Literature review).

The BDH has been proved to generate positive effects especially for individuals at the bottom of the income distribution (Edmonds & Schady, 2009; Oosterbeek, Ponce, & Schady, 2008; Paxson & Schady, 2010; Ponce, 2008). First, the BDH does increase school enrolment. In this sense it helps to reduce long-term poverty trough human capital accumulation. Second, the BDH does increase food expenditure (Schady & Rosero, 2008). However, León and Younger (2007) found rather small effects on child nutrition, and no difference with other kind of income effect.

Third, the BDH does reduce child labour (Edmonds & Schady, 2009). Finally, it may improve cognitive achievements. Ponce and Bedi (2010) found no statistically significant positive impact on test scores among children close to the program eligibility threshold. [14] Nevertheless, Paxson and Schady (2010) found significant effects on developmental and health outcomes, especially for the poorest children in rural areas. [15] The study found positive effects on long term memory (17% of a standard deviation). [16]

3.1 Descriptive statistics

In order to estimate the effect of the BDH on adult labour supply we use three subsamples, restricting to households with a couple of non-unemployed adults (between 18 and 64 years old). [17] The first subsample (BDH recipients) includes only adults living in a BDH recipient household, accounting for 1,417 households (with a couple of adults). We use this subsample to analyze the effect of the transfer size among those receiving it. The second subsample (all adults) includes also individuals living in a non-recipient household, accounting for 2,853 couples (households). This subsample is used to estimate the effect of receiving the BDH. Finally, the third subsample includes single adults, accounting for 1,086 observations (households). In total we include 3,939 household-level observations, including 1,800 BDH recipient households. [18]

Table 1 presents summary statistics. Household size is 4.4 in the BDH recipients subsample, while it is 3.9 and 2.8 for the all adults and single adults subsamples. Average age is around 40 years old. BDH recipient adults have, on average, three years less of education. Women represent 50% of adults in both subsamples of couples, and 76% in the single adults subsample. Minority groups (indigenous, afroecuadorians and montubios) represent a higher percentage in the BDH recipients subsample, as it is also the case for individuals in rural areas. BDH recipient households have on average a higher number of school-age children (between 6 and 17 years old) and old-age persons, while they have less assets and higher number of unsatisfied basic needs. [19]

Table 1:

Descriptive statistics (ENEMDUR – December 2012).

VariableBDH recipientsAll adultsSingle adults
Household size (number of persons)28344.
Schooling (Years of education)28346.
Sex (Female=1/Male=0)28340.5000.5000.0001.00057060.5000.5000.0001.00010860.7620.4260.0001.000
Member (Head=1/Partner=0)28340.5000.5000.0001.00057060.5000.5000.0001.00010861.0000.0001.0001.000
Married (Yes=1/No=0)28340.9950.0700.0001.00057060.9950.0670.0001.00010860.1360.3430.0001.000
Indigenous (Yes=1/No=0)28340.1050.3070.0001.00057060.0600.2380.0001.00010860.0590.2350.0001.000
Afroecuadorian (Yes=1/No=0)28340.0560.2300.0001.00057060.0460.2090.0001.00010860.0630.2420.0001.000
Montubio (Yes=1/No=0)28340.1290.3350.0001.00057060.0620.2420.0001.00010860.0370.1900.0001.000
White or mestizo (Yes=1/No=0)28340.7090.4540.0001.00057060.8300.3750.0001.00010860.8410.3660.0001.000
Number of children (younger than 3 years old)28340.2080.4420.0002.00057060.1980.4230.0003.00010860.1160.3310.0003.000
Number of children (between 3 and 5 years old)28340.2130.4270.0002.00057060.1850.3990.0002.00010860.1070.3290.0002.000
Number of children (between 6 and 11 years old)28341.0350.9250.0004.00057060.8030.8590.0004.00010860.5660.7390.0004.000
Number of children (between 12 and 17 years old)28340.8700.8930.0005.00057060.6970.8290.0005.00010860.7440.7960.0004.000
Number of young (between 18 and 29 years old)28340.3020.6180.0002.00057060.3570.6680.0002.00010860.1370.3440.0001.000
Number of adults (between 30 and 64 years old)28341.6980.6180.0002.00057061.6430.6680.0002.00010860.8630.3440.0001.000
Number of old-age (older than 64 years old)28340.0530.2490.0002.00057060.0340.1970.0002.00010860.2450.4800.0003.000
Number of unsatisfied basic needs28341.0080.8570.0004.00057060.4990.7590.0004.00010860.4280.6780.0004.000
Number of televisions28340.8850.4710.0003.00057061.2750.8020.0008.00010861.0740.6450.0004.000
Number of telephones28341.4160.9160.00011.00057062.0671.1680.00011.00010861.4441.0750.00013.000
Area (Rural=1/Urban=0)28340.6770.4680.0001.00057060.3640.4810.0001.00010860.2960.4570.0001.000
Labour income per-month2834174.72284.230.006,3595706376.43513.150.0017,0001086286.05331.700.004,194
Labour income per-month per-capita283444.39105.560.002,7405706111.91181.490.008,5001086116.20152.940.002,097
W=Labour income per-hour (if W > 0)17931.691.560.062645662.742.990.041069252.212.000.0321
Household’s social transfer (BDH) per-month283437.198.8535.00105570615.1719.130.00105108611.8218.520.0070
Household’s social transfer (BDH) per-month per-capita28349.474.085.003557063.865.340.003510864.487.410.0035
Household’s other non-labour income per-month28344.5326.310.001,000570626.51139.410.002,5501086104.08190.590.002,000
Household’s other non-labour income per-month per-capita28341.227.660.0033357069.5157.320.001,275108643.2987.600.001,000
Poor (Yes=1/No=0)28340.4640.4990.0001.00057060.2140.4100.0001.00010860.3130.4640.0001.000
Extreme poor (Yes=1/No=0)28340.1360.3430.0001.00057060.0630.2420.0001.00010860.1300.3360.0001.000

Average labour income per month is USD 174.72, USD 376.43 and USD 286.05 for the BDH recipients, all adults and single adults subsamples, respectively, including the inactive with zero income. Labour income per hour is, on average, USD 1.69, USD 2.74 and USD 2.21 for each of the subsamples. 46.4% (13.6%) of adults in the BDH recipients subsample are poor (extreme poor), in comparison with 21.4% (6.3%) in the all adults subsample, and 31.3% (13.0%) in the single adults subsample.

Table 2 shows that unemployment is low in all the subsamples, ranging between 0.4% and 2.8%. The rate is almost zero for household heads in the couple subsamples, while it is higher in the case of poor single adults. [20] On the other hand, underemployment affects 40.9%, 36.2% and 48.1% of adults in the BDH recipients, all adults and single adults subsamples, respectively. Underemployment is defined by labour-income below the minimum wage or working less than 40 h per week. [21] In the case of BDH recipient adults, 60% of underemployed comply with both conditions. This percentage is 59% and 68% in the all adults and single adults subsamples, respectively. Underemployed is higher in the case of poor adults in the couple subsamples, as they may be more willing to take any available job.

Table 2:

Paid-labour condition (ENEMDUR – December 2012).

Paid-labour conditionBDH recipientsAll adultsSingle adults
Inactive (% of persons)30.142.635.911.038.116.8
Paid-labour (% of persons)69.957.464.189.061.983.2
Unemployed (% of persons)
Underemployed (% of persons)35.547.140.931.752.436.2
Employed (% of persons)33.79.922.756.99.246.7
Household head
Inactive (% of persons)
Paid-labour (% of persons)98.697.898.298.197.798.088.176.984.4
Unemployed (% of persons)
Underemployed (% of persons)41.878.258.728.179.739.
Employed (% of persons)56.819.639.669.918.058.748.23.333.5
Inactive (% of persons)58.883.
Paid-labour (% of persons)41.317.030.079.926.168.4
Unemployed (% of persons)
Underemployed (% of persons)
Employed (% of persons)

Finally, inactivity is higher in the case of the BDH recipients subsample and for the poor, but it is concentrated among partners, while it is only 1.8% in the case of household heads. Paid-labour participation rate for all adults is 83%, including unemployed (0.4%), underemployed (36.2%) and employed (46.7%). In the BDH recipient subsample the paid-labour participation rate is 64%, while it is 84% for single adults. Paid-labour participation is higher for non-poor individuals (70%) compared to the poor (57%) among those receiving the BDH. For household heads the participation rate is 98% in both subsamples of couples, while it is substantially lower for partners in BDH recipient households (30%) than in the all adults subsample (68%). This difference is related with access to other income sources and with household’s needs of care work. While 96% of inactive partners in the BDH recipient households do housework, this percentage is 64% in the case of partners in non-recipient households, but landlords and pensioners represent 26%. It means that in the case of BDH recipients households no participation on the labour market does not mean more leisure but more housework, while in the case of non-BDH recipient households it is also due to more leisure and thanks to other non-labour-income sources. In the case of inactive single adults 61% do housework and 20% are pensioners.

Housework depends on household composition. If there are more dependent persons, then more care work will be necessary. Table 3 presents the average dependency ratio by paid-labour condition for different subsamples. In the case of household heads there is a positive correlation between paid-labour participation and the number of dependents. It means that a person is expected to be more active in paid-activities if there are more members of the household to be taken care of, because of more resources are needed to satisfy their needs. However, it also means that the head of the household has the role of generating income. Nevertheless, the behaviour of partners is not clear. Looking at partners at the all adults subsample, it appears that they reduce labour-participation if the number of dependents increases, which may be related with a higher role on work care. [22]

Table 3:

Dependency ratio (ENEMDUR – December 2012).

Paid-labour participationHead of housePartnerTotal
BDH recipients
All adults
Single adults

Based on the theoretical framework presented in section two and these empirical data, we establish as hypotheses that the BDH has non-negative effects on labour participation of household heads, while it may finance care work done by the partners and single adults. It is important to mention that in the all adults subsample 96% of partners are women (97% in the BDH recipients subsample).

3.2 Empirical specification

For the empirical model we divide the utility function (Uj=Vj+εj) of the household in an observable part (Vj) and unobserved characteristics (εj). We assume the latest to be independent and identically distributed (i.i.d.) which leads us to follow distribution assumptions for a conditional logit type probability of household j to choose tjl,k from a discrete and finite set of options: L=tjl,1,,tjl,k,tjl,L. Furthermore we assume the observed part of the utility function (Vj=X`jsβ) to be linear in parameters, with vectors X`js of observable variables and β of parameters. In the same way as described by Haan (2004); Kornstad and Thoresen (2007), Löffer et al. (2013), and Kabátek et al. (2014), the logit choice probability can be then defined as:


We specify 16 alternatives of working hours for the combinations between non-paid labour, part-time, full-time and more than full-time labour for the household head and her partner, and four choices in the case of single adults (Table 4). [23] Choices with inactive partners are the most frequents, while full-time and full-time-plus options are more common in the case of households heads. Based on empirical data we use the median number of working hours per week for each individual choice. It is zero hours for no paid-labour, 28 and 20 h for part-time paid-labour of household heads and partners respectively, 24 h for part-time paid-labour of single adults, 40 h for full-time paid-labour and 50 h for full-time-plus paid-labour.

Table 4:

Distribution of households across labour choices.

BDH recipients
Household head
No paid-labourPart-time paid-labourFull-time paid-labourFull-time-plus paid-labourTotal
PartnerNo paid-labour91853804391,013
Part-time paid-labour5416277185
Full-time paid-labour767037120
Full-time-plus paid-labour713166399
All adults
Household head
No paid-labourPart-time paid-labourFull-time paid-labourFull-time-plus paid-labourTotal
PartnerNo paid-labour211894024701,082
Part-time paid-labour786201219513
Full-time paid-labour1744532230823
Full-time-plus paid-labour1333100289435
Single adults
Household head
No paid-labourPart-time paid-labourFull-time paid-labourFull-time-plus paid-labourTotal

We estimate paid-labour income per hour by a Heckman selection equation (Annex 1). Results are used to define household labour income for each possible choice. In this sense we use paid-labour income and working hours as the choice specific variables, and we interact them with other household specific characteristics. Income variables (including transfers) are presented in USD per week, and we use per capita values in order to take into account household size. Finally, we use population weights for all the estimations.

4 Results

4.1 Utility function coefficients

Table 5 presents results of a standard conditional logit estimation of the above derived labour supply model. As it is expected the average marginal utility of paid-labour income is positive, but decreasing on labour, except for the case of single adults where it is positive only in 23% of the observations. The marginal utility of paid-labour is positive for household heads and single adults (lj), but it decreases with working hours and paid-labour income. It is consistent with our hypothesis that leisure is not a normal good until a certain level of income/consumption is achieved. On the other hand, marginal utility of paid-labour is negative in the case of partners (l_j).

Table 5:

Labour supply effects (reduced utility function coefficients).

VariableBDH recipientsAll adultsSingle adults
CoefficientStandard errorCoefficientStandard errorCoefficientStandard error
y=Paid-labour income per week per capita0.128***(0.004)0.097***(0.001)–0.030***(0.001)
y * lj–0.002***(0.000)–0.001***(0.000)0.001***(0.000)
y * l_j–0.001***(0.000)0.001***(0.000)
lj=Paid-labour working hours per week of the head0.066***(0.002)0.062***(0.001)0.056***(0.001)
l_j=Paid-labour working hours per week of the partner–0.148***(0.001)–0.081***(0.001)
lj * l_j0.001***(0.000)0.001***(0.000)
Number of observations224,045549,19552,784
Pseudo R20.2700.1500.023

Table 6 extend the model by including interactions with different household characteristics. We consider the BDH as exogenous as it is a flat transfer without any condition on household composition or labour. [24] However, this assumption can be relaxed with no relevant effects on the main results. [25] Consistent with our hypothesis the BDH has non-negative effects on the marginal utility of the paid-labour working hours in the case of household heads (ljBDH). The amount of the BDH is not significant in the BDH recipients subsample (BDHsocialtransferBDHperweekpercapita), while receiving the social transfer has positive but decreasing (on transfer size) labour effects in the all adults subsample. It can be seen in the negative coefficient of the square term (BDHsocialtransferBDHperweekpercapita2). It means that negative effects may appear if the transfer is big enough. Using the coefficients of the all adults subsample, we estimate that negative effects occur at a transfer level of USD 4.53 per week per person (around USD 71 per month, for an average household size).

Table 6:

Utility function coefficients.

VariableBDH recipientsAll adultsSingle adults
CoefficientStandard errorCoefficientStandard errorCoefficientStandard error
y=Paid-labour income per week per capita0.063***(0.009)0.035***(0.004)–0.141***(0.007)
y * lj–0.002***(0.000)–0.001***(0.000)0.003***(0.000)
y * l_j–0.002***(0.000)–0.001***(0.000)
lj=Paid-labour working hours per week of the head0.205***(0.007)0.176***(0.004)0.104***(0.005)
l_j=Paid-labour working hours per week of the partner–0.017***(0.004)0.051***(0.002)
lj * l_j0.001***(0.000)0.002***(0.000)
lj * other non-labour-income per week per capita–0.007***(0.000)–0.001***(0.000)–0.002***(0.000)
lj * social transfer (BDH) per week per capita0.000(0.001)0.001(0.001)0.022***(0.001)
lj * (social transfer (BDH) per week per capita)^2–0.002***(0.000)–0.002***(0.000)–0.002***(0.000)
lj * BDH (Yes=1/No=0)0.032***(0.001)–0.056***(0.002)
l_j * other non-labour-income per week per capita0.000(0.000)–0.001***(0.000)
l_j * social transfer (BDH) per week per capita–0.006***(0.001)0.004***(0.001)
l_j * (social transfer (BDH) per week per capita)^20.000***(0.000)–0.001***(0.000)
l_j * BDH (Yes=1/No=0)–0.091***(0.001)
lj * age0.002***(0.000)0.001***(0.000)0.006***(0.000)
lj * age20.000***(0.000)0.000***(0.000)0.000***(0.000)
lj * schooling (years of education)0.000(0.000)–0.004***(0.000)–0.003***(0.000)
lj * sex dummy (Female=1/Male=0)–0.141***(0.002)–0.115***(0.001)–0.079***(0.001)
lj * indigenous dummy (Yes=1/No=0)–0.030***(0.001)–0.024***(0.001)–0.005***(0.001)
lj * afro dummy (Yes=1/No=0)–0.029***(0.001)–0.022***(0.001)0.008***(0.001)
lj * montubio dummy (Yes=1/No=0)–0.023***(0.001)–0.020***(0.001)–0.057***(0.003)
l_j * age0.002***(0.000)0.005***(0.000)
l_j * age20.000***(0.000)0.000***(0.000)
l_j * schooling (years of education)0.000***(0.000)0.001***(0.000)
l_j * sex dummy (Female=1/Male=0)–0.081***(0.002)–0.095***(0.001)
l_j * indigenous dummy (Yes=1/No=0)0.010***(0.001)0.037***(0.001)
l_j * afro dummy (Yes=1/No=0)0.008***(0.001)0.004***(0.001)
l_j * montubio dummy (Yes=1/No=0)–0.011***(0.001)–0.021***(0.001)
y * head’s age0.001***(0.000)0.000***(0.000)0.000(0.000)
y * head’s age20.000***(0.000)0.000***(0.000)0.000**(0.000)
y * head’s schooling (years of education)0.003***(0.000)0.003***(0.000)0.010***(0.000)
y * head’s sex dummy (Female=1/Male=0)0.025***(0.004)0.029***(0.001)0.012***(0.002)
y * head’s indigenous dummy (Yes=1/No=0)0.094***(0.003)0.016***(0.001)–0.013***(0.003)
y * head’s afro dummy (Yes=1/No=0)0.016***(0.002)0.010***(0.001)–0.008***(0.002)
y * head’s montubio dummy (Yes=1/No=0)–0.031***(0.003)–0.020***(0.001)0.069***(0.005)
lj * household dependency ratio–0.029***(0.003)0.035***(0.002)0.009***(0.003)
lj * number of children under 5 attending a public nursery–0.015***(0.001)–0.014***(0.000)–0.019***(0.001)
lj * number of children under 5 not attending a public nursery0.005***(0.001)0.006***(0.000)–0.010***(0.000)
lj * number of old age persons (65+)0.013***(0.001)–0.007***(0.001)0.022***(0.001)
l_j * household dependency ratio–0.095***(0.003)–0.080***(0.001)
l_j * number of children under 5 attending a public nursery–0.001**(0.000)–0.001(0.000)
l_j * number of children under 5 not attending a public nursery–0.012***(0.000)–0.006***(0.000)
l_j * number of old age persons (65 +)0.029***(0.001)0.034***(0.001)
y * number of unsatisfied basic needs–0.029***(0.000)–0.018***(0.000)–0.034***(0.001)
y * number of televisions0.021***(0.001)0.007***(0.000)0.002***(0.000)
y * number of telephones0.015***(0.000)0.007***(0.000)0.008***(0.000)
y * area dummy (Rural=1/Urban=0)–0.005***(0.001)–0.009***(0.000)0.006***(0.001)
y * parish’s poverty by basic needs head count–0.044***(0.003)0.047***(0.001)–0.122***(0.003)
lj * parish’s poverty by basic needs head count–0.036***(0.002)–0.044***(0.001)0.057***(0.002)
l_j * parish’s poverty by basic needs head count–0.009***(0.001)–0.038***(0.001)0.000(0.000)
Number of observations224,045549,19552,784
Pseudo R20.3390.2600.152

However, the BDH generates negative effects on the marginal utility of paid-labour working hours of partners (l_jBDH) and single adults (ljBDH). Nevertheless, partners and single adults paid-labour participation is determinate by other household needs (i.e. care work) which can be paid by the BDH. [26] The marginal utility of paid-labour working hours decreases if the household dependency ratio increases in the case of partners (l_jhouseholddependencyratio). It is also the case if the number of children under 5 years old rises. However, the effect is lower or not significant if children attend to a public nursery (l_jnumberofchildrenunder5attendingapublicnursery) than if they do not (l_jnumberofchildrenunder5notattendingapublicnursery), except for single adults. It means that partners allocate more time to childcare. Finally, there is a positive effect related with the presence of old-age persons in the household (number of old age persons). We interpret these results in the sense that paid-labour participation of partners and single adults may be promoted by increasing access to childcare facilities and by the presence of additional carers (i.e. old-age persons) because the burden of care work is reduced.

Finally, it has to be noted that the marginal utility of paid-labour working hours is lower for women than for men. We relate this effect with childcare role but also due to paid-labour income inequality against women (Annex 1). In this sense both childcare and gender equity policies may be seen as complements of social transfers if paid-labour participation is to be promoted.

All these results are consistent with our hypotheses that the BDH does not generate negative labour effects in the case of household heads. Even more, we find positive effects under some conditions. However, the size of the transfer does matter because if it is big enough paid-labour disincentives may be generated. In the case of partners and single adults we argue that households may be using social transfers to finance childcare.

4.2 Average marginal effects

Relying in a multinomial logit equation we estimate average marginal effects (AME) on the probability of choosing a specific paid-labour option. In this case we assume the decision of one adult as given. It means that we estimate the effects independently for household heads and partners, considering four possible choices. Summary results are presented in Table 7.

Table 7:

Average marginal effects of the BDH on paid-labour.

VariableChoiceBDH recipientsAll adultsSingle adults
Household heads
Social transfer (BDH) per week per capitaPr(no paid-labour)0.011***0.006**0.015
Pr(part-time paid-labour)0.0260.022***0.010
Pr(full-time paid-labour)0.008–0.032–0.038*
Pr(full-time-plus paid-labour)–0.0460.0040.013
BDH (Yes=1/No=0)Pr(no paid-labour)–0.021**0.002
Pr(part-time paid-labour)–0.059*–0.043
Pr(full-time paid-labour)–0.0120.058
Pr(full-time-plus paid-labour)0.092–0.018
Social transfer (BDH) per week per capitaPr(no paid-labour)0.056*0.001
Pr(part-time paid-labour)0.0090.028*
Pr(full-time paid-labour)–0.059**–0.031
Pr(full-time-plus paid-labour)–0.0070.002
BDH (Yes=1/No=0)Pr(no paid-labour)0.346***
Pr(part-time paid-labour)–0.162***
Pr(full-time paid-labour)–0.061
Pr(full-time-plus paid-labour)–0.123**

Among those adults receiving the BDH, an increase of USD 1 per week per capita on the transfer size (it is an increment of 42%, at the mean level of the BDH amount) increases the probability of no paid-labour by 1.1 and 5.6 percentage points in the case of household heads and partners, respectively. However there is not any significant effect on other choices in the case of household heads, but a reduction of 5.9 percentage points on full-time paid-labour in the case of partners. On the other hand, looking at the all adults subsample, receiving the BDH reduces the probability of no paid-labour by 2.1 percentage points for household heads, and increases the probability of full-time-paid labour by 9.2 percentage points (but not significantly). However, as aforementioned the transfer size reduces the positive effect, as an additional USD 1 increases the probability of no paid-labour by 0.6 percentage points. In the case of partners, receiving the BDH increases the probability of no paid-labour by 34.6 percentage points. Finally, we find that receiving the BDH has no significant effects in the case of single adults. [27]

In the case of partners (Table 8), if the dependency ratio of the household increases by 0.01 (between 1.7% and 2.3%, at the mean level) the probability of no paid-labour increases by 0.5 percentage points in the BDH recipients subsamples, while the effect is not significant for the all adults subsample, and it is negative (reduction of 0.2 percentage points) in the case of single adults. On the other hand, the presence of an additional old-age person in the households reduces the probability of no paid-labour by 18.4, 8.3 and 6.5 percentage points in each subsample, respectively, by increasing the probability of full-time paid-labour by 9.3 and 12.2 percentage points in the BDH recipients and the all adults subsamples, respectively, and the probability of full-time-plus paid-labour by 6.6 percentage points in the case of single adults. Finally, the number of children under 5 years old increases the probability of no paid-labour by between 3.4 and 6.4 percentage points in the subsamples of couples, if the children do not attend to a public nursery. However, this effect is not significant if the children attend to a public nursery. In the case of single adults, one additional child under 5, not attending to a public nursery, increases the probability of no paid-labour by 4.6 percentage points; however additional estimations show that this negative labour effect does not happen in the case of men. [28]

Table 8:

Average marginal effects on paid-labour (partners).

VariableChoiceBDH recipientsAll adultsSingle adults
Household dependency ratioPr(no paid-labour)0.513**0.047–0.233**
Pr(part-time paid-labour)–0.057–0.1100.103
Pr(full-time paid-labour)–0.323**0.163*0.085
Pr(full-time-plus paid-labour)–0.133–0.1000.044
Number of children under 5 attending a public nurseryPr(no paid-labour)0.0260.0110.072
Pr(part-time paid-labour)–0.016–0.0170.012
Pr(full-time paid-labour)0.0010.0220.071
Pr(full-time-plus paid-labour)–0.011–0.016–0.155*
Number of children under 5 not attending a public nurseryPr(no paid-labour)0.064**0.034***0.046*
Pr(part-time paid-labour)–0.004–0.0290.006
Pr(full-time paid-labour)–0.037**0.028–0.002
Pr(full-time-plus paid-labour)–0.023–0.033*–0.050
Number of old age persons (65 +)Pr(no paid-labour)–0.184***–0.083***–0.065*
Pr(part-time paid-labour)0.035–0.073–0.040
Pr(full-time paid-labour)0.093***0.122***0.039
Pr(full-time-plus paid-labour)0.0550.0340.066*

All our estimations show that the BDH does not generate negative labour supply effects on household heads, while we find positive effects in some cases. However, the amount of the transfer should be defined at an optimal level. From our theoretical framework we relate this effect with the idea that leisure cannot be considered a normal good in the case of poor individuals, and that a social transfer may help households to solve liquidity constrains and to cover different transaction costs. Nevertheless, we find negative paid-labour effects for partners and in some cases for single adults. The BDH may be paying for childcare, as we relate this effect with lack of access to alternative childcare options, and because of paid-labour income inequality against women. If paid-labour participation is to be promoted social transfers should be complemented by policies addressing gender equity and childcare.

5 Final remarks

Social transfers are largely being implemented as a poverty and inequality reduction strategy. Recent literature provides new analytical frameworks to rely on social transfers as an instrument to generate positive economic returns. However, empirical evidence remains scarce in this field. This study provides a theoretical framework and contributes with empirical evidence on the effects of unconditional cash transfers on adult labour supply, which we believe is a key question to understand the economic effects of social transfers. Moreover, results are relevant to discuss unconditional cash transfers as a first stage for the implementation of universal basic income in developing countries.

Following traditional labour supply theories it can be argued that a social transfer discourage labour due to an income effect, assuming that leisure is a normal good. We argue that it is not the case for poor individuals which cannot cover her basic needs. For example, it is difficult to value leisure without sufficient water, food and clothing. In this sense, social transfers may not generate such kind of income effect in the case of poor households. Furthermore, international evidence suggests non-negative labour effects of social transfers under certain circumstances.

We estimate a unitary discrete labour supply model using data from Ecuador. Results for the utility function and average marginal effects are consistent with our theoretical framework, and prove our hypothesis, as we find non-negative effects of social transfers on household heads labour supply, but limited to a certain transfer level. Moreover, we find positive effects which we relate with the idea of social transfers helping poor households to solve liquidity constrains and to cover different transaction costs. On the other hand, we find negative labour supply effects on partners (who are mainly women) and single adults where a social transfer may be paying for childcare, but also because of idiosyncratic characteristics and labour market inequality against women. We believe that policies addressing gender equity and childcare should complement social transfers if paid-labour participation of partners is a final objective; however, it should be carefully thought regarding child wellbeing and the freedom to choose any kind of work.


The authors are grateful for comments from colleagues of the research group on poverty, public policy and inclusive innovation at the Maastricht Graduate School of Governance/UNU-MERIT, Maastricht University, especially those from Franziska Gassmann, Pierre Mohnen and Nyasha Tirivayi. An earlier version of this paper was presented at the Fifth Bolivian Conference on Development Economics (BCDE2013), Santa Cruz de la Sierra (14–15 November 2013). All remaining errors are ours.


Aaberge, R., & Colombino, U. (2013). Using a microeconometric model of household labour supply to design optimal income taxes. The Scandinavian Journal of Economics, 115(2), 449–475.10.1111/sjoe.12015Search in Google Scholar

Alderman, H., & Yemtsov, R. (2012). Productive role of safety nets. Washington DC: The World Bank.Search in Google Scholar

Alzúa, M. L., Cruces, G., & Ripani, L. (2013). Welfare programs and labor supply in developing countries: Expermitental evidence from Latin America. Journal of Population Economics, 26, 1255–1284.10.1007/s00148-012-0458-0Search in Google Scholar

Arnold, C., Conway, T., & Greenslade, M. (2011). Cash transfers. London: Department for International Development, DFID.Search in Google Scholar

Barrientos, A. (2012). Social transfers and growth: What do we know? What do we need to find out? World Development, 40(1), 11–20.10.1016/j.worlddev.2011.05.012Search in Google Scholar

Barrientos, A., & Nino-Zarazua, M. (2010). Effects of non-contributory social transfers in developing countries: A compendium. Geneva: International Labour Organization.Search in Google Scholar

Barrientos, A., & Scott, J. (2008). Social transfers and growth: A review. BWPI Working Paper.10.2139/ssrn.1538926Search in Google Scholar

Barrientos, A., & Villa, J. M. (2013). Antipoverty transfers and labour force participation effects. BWPI Working Paper, 185.10.2139/ssrn.2275291Search in Google Scholar

Beninger, D., Laisney, F., & Beblo, M. (2007). Welfare analysis of a tax reform for Germany: A comparison of the unitary and collective models of households labour supply. Journal of Population Economics, 20(4), 869–893.10.1007/s00148-006-0077-8Search in Google Scholar

Bloemen, H. (2010). Income taxation in an empirical collective household labour supply model with discrete hours. Tinbergen Institute Discussion Paper (2010-010/3).Search in Google Scholar

Bloemen, H. G. (2009). An empirical model of collective household labor supply with non-participation. The Economic Journal, 120, 183–214.10.1111/j.1468-0297.2009.02292.xSearch in Google Scholar

Blundell, R., Chiappori, P.-A., Magnac, T., & Meghir, C. (2007). Collective labour supply: Heterogeneity and non-participation. The Review of Economics Studies, 74(2), 417–445.10.1111/j.1467-937X.2007.00440.xSearch in Google Scholar

Blundell, R., & Shephard, A. (2012). Employment, hours of work and the optimal taxation of low-income families. Review of Economic Studies, 79, 481–510.10.1093/acprof:oso/9780198749806.003.0014Search in Google Scholar

Bourguignon, F., Ferreira, F. H., & Leite, P. G. (2003). Conditional cash transfers, schooling, and child labour: Micro-simulating Brazil’s Bolsa Escola program. The World Bank Economic Review, 17(2), 229–254.10.1093/wber/lhg018Search in Google Scholar

Breunig, R., & Gong, X. (2010). Child care availability, quality and affordability: Are local problems related to maternal labour supply? Treasury Working Paper (2010–02).Search in Google Scholar

Buser, T., Oosterbeek, H., Plug, E., Ponce, J., & Rosero, J. (2013). Impactos de quitar el BDH en salud infantil. Mimeo.Search in Google Scholar

Cecchini, S., & Martínez, R. (2011). Protección social inclusiva en América Latina. Una mirada integral, un enfoque de derechos. Santiago: Comisión Económica para América Latina y El Caribe (CEPAL).Search in Google Scholar

Colombino, U., Locatelli, M., Narazani, E., & O’Donoghue, C. (2010). Alternative basic income mechanisms: An evaluation exercise with microeconometric model. Basic Income Studies, 5(1). ISSN (Online) 1932-0183, DOI: 10.2202/1932-0183.1162Search in Google Scholar

Dagsvik, J., Jia, Z., Kornstad, T., & Thoresen, T. (2014). Theoretical and practical arguments for modelling labor supply as a choice among latent jobs. Journal of Economic Surveys, 28(1), 134–151.10.1111/joes.12003Search in Google Scholar

Edmonds, E., & Schady, N.. (2009). Poverty alleviation and child labor. NBER Working Paper Series (15345).10.3386/w15345Search in Google Scholar

Fernandez, F., & Saldarriaga, V. (2014). Do benefit recipients change their labor supply after receiving the cash transfer? Evidence from the Peruvian Juntos Program. IZA Journal of Labor & Development, 3(2). Retrieved from in Google Scholar

Foguel, M. N., & de Barros, R. P. (2010). The effects of conditional cash transfers programmes on adult labour supply: An empirical analysis using time-series-cross-section sample of Brazilian municipalities. Est. Econ. Sao Paulo, 40(2), 259/293.10.1590/S0101-41612010000200001Search in Google Scholar

Gonzales-Rozada, M., & Llerena Pinto, F. (20 de April de 2011). The effects of a conditional transfer program on the labor market: The Human Development Bonus in Ecuador. Mimeo.Search in Google Scholar

Haan, P. (2004). Discrete Choice Labor Supply: Conditional Logit vs. Random Coefficient Models. DIW Berlin, Discussion Papers 394.Search in Google Scholar

Harvey, P. (2013). More for less: The job guarantee strategy. Basic Income Studies, 7(2), 3–18.10.1515/bis-2013-0006Search in Google Scholar

INEC. (25 de May de 2013). Instituto Nacional de Estadísticas y Censos – Proyecciones poblacionales. Retrieved from in Google Scholar

Jackson, W. (1999). Basic Income and the right to work: A Keynesian approach. Journal of Post Keynesian Economics, 21(4), 639–662.10.1080/01603477.1999.11490220Search in Google Scholar

Kabátek, J., van Soest, A., & Stancanelli, E. (2014). Income taxation, labour supply and housework: A discrete choice model for French couples. Labour Economics, 27, 30–43.10.1016/j.labeco.2014.01.004Search in Google Scholar

Kornstad, T., & Thoresen, T. (2007). A discrete choice model for labor supply and childcare. Journal of Popular Economics 20(4), 781–803.10.1007/s00148-005-0025-zSearch in Google Scholar

León, M., & Younger, S. (2007). Transfer payments, mothers’ income and child health in Ecuador. The Journal of Development Studies, 43(6), 1126–1143.10.1080/00220380701466708Search in Google Scholar

Löffler, M., Peichl, A., & Siegloch, S. (2013). Discrete choice labor supply models and wage exogeneity. Mimeo. Retrieved from in Google Scholar

LoVuolo, R. (2012). Prospects and challenges for the basic income proposal in Latin America. Basic Income Studies, 7(1). DOI: 10.1515/1932-0183.1195Search in Google Scholar

Maurizio, R., & Vázquez, G. (2014). Argentina: Efectos del programa Asignación Universal por Hijo en el comportamiento laboral de los adultos. Revista CEPAL, 113, 121–144.10.18356/11845fc9-esSearch in Google Scholar

Mideros, A. (2012). Ecuador: Defining and measuring multidimensional poverty, 2006–2010. ECLAC Review, 108, 49–67.10.18356/9ed3c0a9-enSearch in Google Scholar

Mideros, A., Gassmann, F., & Mohnen, P. (2013). Estimation of rates of return on social protection: Making the case for non-contributory social transfers in Cambodia. (2013–063).Search in Google Scholar

Novella, R., Ripani, L., Cruces, G., & Alzúa, M. L. (2012). Conditional cash transfers, female bargaining power and paternal labour supply. Washington DC: IDB Working Paper Series IDB-WP–368.Search in Google Scholar

Oosterbeek, H., Ponce, J., & Schady, N. (2008). The impact of cash transfers on school enrollment. Tinbergen Institute Discussion Paper (TI 2008-037/4).Search in Google Scholar

Paxson, C., & Schady, N. (2010). Does money matter? The effects of cash transfers on child development in rural Ecuador. Economic Development and Cultural Change, 59(1), 187–229.10.1086/655458Search in Google Scholar

Ponce, J. (2008). Education policy and performance. Evaluating the impact of targeted education programs in Ecuador. Maastricht: Shaker Publishing.Search in Google Scholar

Ponce, J., & Bedi, A. (2010). The impact of a cash transfer program on cognitive achievements: The Bono de Desarrollo Humano of Ecuador. Economics of Education Review, 29, 116–125.10.1016/j.econedurev.2009.07.005Search in Google Scholar

Posel, D., Fairburn, J. A., & Lund, F. (2006). Labour migration and households: A reconsideration of the effects of the social pension on labour supply in South Africa. Economic Modelling, 23, 836–853.10.1016/j.econmod.2005.10.010Search in Google Scholar

Psacharopoulos, G., & Patrinos, H. (2002). Returns to investment in education: A further update. World Bank Policy Research Working Paper (2881).Search in Google Scholar

Schady, N., & Araujo, M. C. (2008). Cash transfers, conditions, and school enrolment in Ecuador. Economía, 8(2), 43–77.10.1353/eco.0.0004Search in Google Scholar

Schady, N., & Rosero, J. (2008). Are cash transfers made to women spent like other sources of income? Economics Letters, 101, 246–248.10.1037/e602542012-001Search in Google Scholar

SIISE. (2013). Sistema Integrado de Indicadores Sociales del Ecuador. Retrieved from 25 de May de 2013, in Google Scholar

Skoufias, E., & Di Maro, V. (2008). Conditional cash transfers, adult work incentives, and poverty. The Journal of Development Studies, 44(7), 935–960.10.1596/1813-9450-3973Search in Google Scholar

Standing, G. (2008). How cash transfers promote the case for basic income. Basic Income Studies, 3(1), 1–30.10.2202/1932-0183.1106Search in Google Scholar

Standing, G. (2013). Why a basic income is necessary for a right to work. Basic Income Studies, 7(2), 19–40.10.1515/bis-2013-0007Search in Google Scholar

Suplicy, E. M. (2003). Ligitimizing basic income in developing countries: Brazil, or “the answer is blowin’ in the wind. Journal of Post Keynesian Economics, 25(3), 407–424.Search in Google Scholar

Tirivayi, N., Knowles, M., & Davis, B. (2013). The interaction between social protection and agriculture. A review of evidence. Rome: Food and Agriculture Organization of the United Nations (FAO).Search in Google Scholar

UNICEF. (2012). Integrated social protection systems. Enhancing equity for children. New York: United Nations Children’s Fund.Search in Google Scholar

Van Parijs, P. (2004). Basic income: A simple and powerful idea for the twenty-first century. Politics & Society, 32(1), 7–39.10.1177/0032329203261095Search in Google Scholar

van Soest, A. (1995). Structural models of family labor supply. A discrete choice approach. The Journal of Human Resources, 30(1), 63–88.10.2307/146191Search in Google Scholar

van Soest, A., Das, M., & Gong, X. (2002). A structural labour supply model with flexible preferences. Journal of Econometrics, 107, 345–374.10.1016/S0304-4076(01)00128-2Search in Google Scholar

World Bank (7 de July de 2013). The World Bank Indicators. Retrieved from in Google Scholar

Annex 1: Paid-labour income (Heckman selection equation)

Log of labour income per-hourBDH recipientsAll adultsSingle adults
Age squared0.0000.0000.000
Schooling (Years of education completed)0.033***0.078***0.064***
Sex (Female=1/Male=0)–0.232***–0.144***–0.183**
Indigenous (Yes=1/No=0)–0.131*–0.104–0.366***
Afroecuadorian (Yes=1/No=0)0.107–0.122**–0.186**
Montubio (Yes=1/No=0)–0.0760.023–0.001
Number of persons by age groupYesYesYes
Rural-urban dummyYesYesYes
Province dummyYesYesYes
Parish’s rate of poverty by unsatisfied basic needsYesYesYes
Selection equation
Number of unsatisfied basic needs (between 0 and 5)–0.244***–0.384***–0.259***
Number of televisions0.0050.239***–0.237*
Number of telephones0.088*0.165***–0.099
Age and age square variablesYesYesYes
Schooling (Years of education completed)YesYesYes
Sex dummyYesYesYes
Married dummyYesYesYes
Number of observations:2,8345,7061,086
Inverse mills ratio0.6170.3160.275
Published Online: 2015-11-3
Published in Print: 2015-12-1

©2015 by De Gruyter

Downloaded on 2.3.2024 from
Scroll to top button