1 Introduction

Wellbeing, or sukha-pawa in Thai, is a broad concept embracing feelings of joy or fulfilment, as well as satisfaction of a given list of capabilities, functionings or needs. Approaches to wellbeing tend to be either objective, for example, focusing on whether someone can satisfy their needs, or subjective, focusing on how they feel about their lives. Scholars using both approaches have observed that linking objective measures such as income to individual perceptions is extremely difficult (e.g. Diener and Lucas 2000; Gasper 2004). This problem has been extensively reported in happiness studies since the work of Easterlin (1974) and has not been resolved by incorporating a wider range of objective measures than wealth alone.

Although income or wealth have been the objective indicators most commonly used by economists, broader measures such as basic needs or capabilities are increasingly employed to assess societies’ wellbeing; for example, the United Nations’ Human Development Index. These are implicitly founded on normative theories of the ‘good’Footnote 1 such as the Capabilities approach by Sen (1985) and Nussbaum (2000) and the Theory of Human Need (hereafter THN) by Doyal and Gough (1991). Normative theories offer richer accounts of wellbeing than approaches focused only on income or wealth. They acknowledge its multidimensionality and identify ‘a flourishing life’ or a ‘good life’ as the ultimate societal goal. For example, the THN maintains that physical health and autonomy are necessary to participate in one’s chosen form of life and avoid serious harm.

Few studies have tested the relationship between the satisfaction of basic needs and subjective wellbeing (hereafter SWB) (Diener et al. 1995; Lelkes 2005; Guillen-Royo and Velazco 2006; Tay and Diener 2011), as opposed to the relationship between wealth and SWB. The reason for this omission is the common assumption that at lower economic levels income is used primarily to satisfy basic needs and so the two indicators can be used interchangeably (Veenhoven 2007). The present paper questions this assumption by investigating whether satisfaction of basic needs and level of wealth can be used interchangeably in relation to subjective wellbeing in Thailand. This is a country where there are striking contrasts between the wealth of modern elite centres and the majority of rural communities (especially in the Northeast) and struggles over issues of identity and particularly what constitute ‘good’ Thai values. Due to the growth of European hypermarkets in urban Thailand consumers can enact middle class identities in settings such as ‘Talad Tesco’ and use their consumption choices to indicate their preference for ‘elite’ (modern, clean, private, quick) rather than ‘traditional’ ways of life (old, dirty, public, slow) (Isaacs 2009). These spatial divides are compounded by generational ones where young people are exposed to modern and globalised aspirations, including growing consumerism, while their parents struggle to adjust (e.g. see Mills 1997 in relation to the aspirations of female garment workers from Northeast Thailand where our fieldwork took place). The 2007 National Human Development Report tackled these themes directly by attempting to operationalise the ‘sufficiency economy’ proposed by King Bhumipol, which would ensure that the “quality of the packaging”, i.e. globalisation and economic development, is not “all too easily mistaken for the quality of the product” (UNDP 2007: 72).

Following previous research on the relationship between income, basic needs and happiness (Diener et al. 1995; Guillen-Royo 2007; Guillen-Royo and Velazco 2006; Camfield and Guillen-Royo 2010) our hypothesis is that measures of wealth and the satisfaction of basic needs capture different aspects of people’s objective wellbeing which have separate effects on happiness. The paper reports an analysis done at two levels: first we carry out a descriptive analysis of the relationship between the three basic indicators of wellbeing, namely material wealth, basic needs and happiness, highlighting how it differs between communities. Second, we analyse the relevance of wealth and basic needs in determining happiness using an Ordered Probit model following the tradition of happiness studies in economics (Van Praag and Ferrer-i-Carbonell 2004; Graham 2005; Rojas 2007; Frey 2008).

Thus, the structure of the paper is as follows: we start by discussing the role of wealth, satisfaction of basic needs and happiness as indicators of objective and subjective wellbeing. Then, we present the characteristics of the Thai communities surveyed in the study and the methodology. Next, we report the level of basic needs and wealth of the communities and discuss the extent to which they are inter-related. Finally, we present the empirical model used to analyse the effect on happiness of material wealth and satisfaction of basic needs and discuss the results.

2 Wealth, Basic Needs and Happiness in the Objective and Subjective Wellbeing Traditions

Income, wealth or expenditure have often been used interchangeably as indicators of objective wellbeing (OWB); i.e. wellbeing assessed through variables not based on perceptions or feelings (Gasper 2005). Following Sen’s capabilities approach (Sen 1985), other multidimensional measures of OWB have been popularised, such as the Human Development Index or the targets included in the UN Millennium Development Goals. These measures have extended traditional concerns about people’s available resources to acquire material goods to include their capacity to participate in the political, social, religious and cultural life in their society. Thus, they have usually included indicators of health status, literacy, housing conditions, employment status and/or social integration.

The Theory of Human Need (THN) by Doyal and Gough (1991) relates to Sen’s capabilities approach in that it takes a broader approach to OWB by defining a list of requisites for wellbeing. The THN differentiates between ‘universal goals’, ‘basic needs’ and ‘intermediate needs’ or needs satisfiers. As universal goals they identify avoidance of serious harm, social participation and critical participation. Physical health and critical autonomy are considered to be basic needs.

Following Doyal and Gough (1991: 2002) physical health requires that satisfiers have the subsequent characteristics: adequate nutritional food and water, adequate protective housing, non-hazardous work and physical environments and appropriate health care. Autonomy is enhanced by people having access to security in childhood, significant primary relationships, physical and economic security, safe birth control and childbearing, and appropriate basic and cross-cultural education. These characteristics of satisfiers, also called ‘intermediate needs’, must be present in a society for people to be able to fully participate. However, the way ‘intermediate needs’ are provided through satisfiers (the goods, services and institutional arrangements used to meet basic needs) differ across cultures and socio-economic systems (Gough 2005). Thus, individual and societal wellbeing should not be assessed through satisfiers but through objective indicators showing the level of satisfaction of the universal components of the THN which are basic needs and/or ‘intermediate needs’.

Unlike objective approaches which address wellbeing as ‘valued functionings, typically identified by an authoritative other’ (Gasper 2005:186), SWB approaches define wellbeing from people’s own perspectives. SWB corresponds to “people’s multidimensional evaluation of their lives, including cognitive judgments of life satisfaction as well as affective evaluations of moods and emotions” (Eid and Diener 2004: 65). In order to capture SWB, researchers often use self-reported questions about happiness or life satisfaction. Within international development there has been increasingly strong support for using both objective and subjective indicators to evaluate wellbeing. The reason for this is that people’s wellbeing is influenced not only by what is available to them, but also by what they feel about the different options or constraints that they are facing (Camfield and McGregor 2005).

Studies of the relationship between subjective accounts of wellbeing and objective measures such as income, consumption or availability of housing, school or health facilities have not always produced the expected results (Gasper 2005). This is especially the case with regards to the relationship between income or wealth and SWB.Footnote 2 This relationship has been widely studied since Easterlin (1974) demonstrated that relative income could be more important than absolute in determining wellbeing. Recent research suggests that even in developing countries the relationship between wealth and wellbeing might depend on relative considerations (Fafchamps and Shilpi 2006; Kingdon and Knight 2007; Rojas 2008; Guillen-Royo 2010). These studies challenge the common assumption that basic need satisfaction is the priority of poor people and therefore, income is used predominantly to satisfy basic needs, which increases SWB. They suggest instead that income is important for SWB mainly if it implies higher status and that people’s satisfaction depends more on the symbolic meaning of income than on its use to satisfy needs. If this is the case, then income and wealth are poor indicators of basic needs satisfaction and should not be used interchangeably, as they capture different aspects of OWB.

So far, there have been few systematic attempts to distinguish between the effects of basic needs satisfaction and income/expenditure/wealth on SWB. Diener et al. (1995) used national indicators of basic health, safety and survival needs to study whether income would keep its significance once basic needs were considered. Their results underlined the importance of basic needs as the correlation between income and SWB was greatly reduced when basic needs such as health and nutrition were controlled for and there was a curvilinear relationship between basic need satisfaction and purchasing power. Moreover, in a study of SWB determinants in Europe, Lelkes (2005) found that ‘basic needs’ indicators such as health status, unemployment and friendships explained people’s wellbeing even controlling for income, which contributed significantly and independently to happiness. In addition, a recent study by Guillen-Royo (2010) showed that among poor Peruvians there were some domains (children’s education, family housing, family clothing) where relative expenditure was more important for SWB than absolute expenditure. However, in all domains an index of basic needs deprivation constructed using the THN was significantly and positively related to SWB.

The importance of basic needs for SWB is further studied in a recent work by Tay and Diener (2011) covering 123 countries. They expand the concept of basic needs as used by Doyal and Gough to include needs such as respect, social support, love, pride and mastery following Maslow’s categorisation of needs. They find that needs fulfilment is a stronger predictor of SWB than income, although both have a significant effect, and the relationship between income and SWB is only partially explained by its effect on needs fulfilment. This paper contributes to the literature by further clarifying the distinct role in determining SWB of wealth and basic needs satisfaction; two OWB measures that have often been used interchangeably, particularly in relatively deprived settings.

3 The Thai Context and the Resources and Needs Questionnarie

Until recently Thailand’s economic success was manifested in its consistent growth, especially high during the ‘boom years’ of the mid 1980s to 1990s, when average growth in GDP was 9% per annum (Warr 2005). Economic growth, supported by planned developments in infrastructure and service provision, brought about rapid change, which was only slightly slowed by the economic crisis in 1997. This change encompassed not only the economy, but also social and political structures (for example, the focus on the rights of citizens in the 1997 Constitution), and its impact has been wide-ranging. Increases in real average incomes reduced overall poverty incidence from 45% of the population in 1986, to 10% in 2002 (ibid.), and caused the growth of an increasingly influential middle class. It was accompanied by increased household access to electricity and piped water, decreasing infant mortality, and rising life expectancy.

These changes were undoubtedly beneficial, but they were achieved through an unbalanced growth strategy, which created ‘uneven development’ in different sectors and locations (Parnwell and Arghiros 1996). Poverty reduction appears now to have plateaued, and there are persistent disparities in economic and welfare outcomes between different regions, rural and urban areas, and different population groups. This is especially remarkable in two of the most populous regions of Thailand; the Southern province of Songkhla and the Northeast which together account for 48% of the Thai population.

The seven communities investigated in this research were selected by the Wellbeing in Developing countries (WeD) ESRC research group in order to ‘give insights into poverty, inequality and quality of life in urban, peri-urban, and rural areas of Thailand’ (WeD 2007: 1). Thailand faces problems common to many countries experiencing rapid social change and increasing inequality, in Southeast Asia and elsewhere. For this reason an in-depth mixed-methods study that spans the rural–urban continuum in two contrasting regions enables generalisation of specific aspects of this experience. While the sample is not statistically representative, in combination the two regions cover nearly 50% of the population and 50% of those living in poverty (UNDP 2010). In that sense their experiences are more representative of the Thai population than those of people living in Bangkok where household income is more than double the median (ibid, p. 151). Table 1 shows the main characteristics of the seven communities highlighting their particularities as well as their commonalities. The first three sites cover the continuum rural- peri-urban- urban in the South, whilst the other four represent it for the Northeast. The three Southern sites have both Buddhist and Muslim residents whilst the Northeast is almost entirely Buddhist. All villages are involved to a certain extent in agricultural activities; rubber plantations in the south and rice and cassava in the Northeast. The two urban communities have a transient population living on railway land and show the greatest economic inequalities.

Table 1 The WeD communities in Northeast and South Thailand

3.1 The Resources and Needs questionnaireFootnote 3

The analysis of the differential impact of wealth and basic needs on happiness in the Thai communities draws on the WeD Resources and Needs Questionnaire (hereafter RANQ) which was based on the THN and the Resource Profiles Approach (McGregor et al. 2007). The RANQ collected information from 1,183 households in the seven Thai communities described in Table 1 during 2004. The questionnaire was interview-administered in the local language/s by trained enumerators from universities in Khon Kaen and Songkhla and the response rate was 100%. As communities were smaller in the Northeast, it was possible to take a census approach, which accounts for the differences in size; however, it was necessary to use random sampling in the South with a cap of 250 households in rural/peri-urban communities and 150 in urban.

RANQ gathered information on:

  • household resources (human, material, natural, social and cultural),

  • the level of needs satisfaction by household (income, health, education, food and housing),

  • long-term shocks and fortunes

  • happiness levels and satisfaction with life domains such as housing.

Most of the questions of the RANQ are at the household level, but the ones related to personal perceptions, for example, satisfaction with expenditure on food and global happiness, were only addressed to the head of the household.

3.2 Variables

The main variables used in this study are: happiness, material wealth and basic needs. Happiness is investigated through a three-point scale question asking “Taking all things together, how would you say things are these days? Would you say you are: very happy (3), fairly happy (2), not too happy (1)”.Footnote 4 Material wealth is approximated through the number of consumer assets owned by the household.Footnote 5 RANQ enquires about the assets the household has access to through a predefined list of 51 items classified under transport, electrical consumer goods and other household assets. This is a simple measure of household wealth that has been used in other studies of happiness determinants (see Graham and Pettinato 2002).

Basic needs are measured through the INDI, an aggregate indicator of unmet needs at the household level generated by McGregor et al. (2007). They computed an index of lack of access to intermediate needs following Desai and Shah’s (1988) methodology and drawing on the THN’s classification of intermediate needs. The deprivation index (INDI) is generated as follows.

$$ Di = \sum {_{k} } I_{ik} $$

where

  • Ik = 1 if a household is facing a lack of access to the kth intermediate need

  • Ik = 0, otherwise

Thus, 0 ≤ Di ≤ 10, where an index of 10 would indicate that a household lacks access to all ten intermediate needs and an index of 0 that it has access to all of them.

4 Basic Needs and Wealth in North-East and Southern Thailand

Table 2 shows how happiness is distributed in percentages by site. Most household heads declare themselves to be fairly happy (76%) whilst only a minority state that they feel very happy (5%) and nearly 19% of the household heads describe themselves as feeling not too happy. Again, differences between regions stand out, North-eastern households being on average unhappier and more dissatisfied. The unhappiest households are in Ban Dong with 33% of the household heads declaring themselves to be ‘not too happy’.

Table 2 Happiness by site (%)

Table 3 below presents the indicators of intermediate need satisfaction used to construct the INDI. They are described by site and grouped under the universal basic need (physical health and autonomy) they contribute to. It illustrates how households in the sample have a relatively low level of satisfaction of their basic needs, particularly concerning physical health: 66% do not make use of health care, especially participants from the rural areas of the North-east; 63% use unsafe sources of drinking water which again affects strongly the population of the North-east together with the rural South; and food shortages threaten people’s health in the North-east as it is common to experience scarcity of staple food,Footnote 6 protein or fruit and vegetables.

Table 3 Indicators of the THN intermediate needs

Indicators of autonomy of agency are comparatively positive but show how the high exposure of households to economic shocks threatens stability and diminishes the possibilities of social participation. For example, flooding is a particular problem in Ban Tha and all the sites in the South, especially the urban site where over 90 percent of households were affected. The flooding has changed patterns of settlement, land cultivation, and migration (for example, a significant proportion of people from Ban Tha now work as taxi-drivers in Bangkok). Regarding education most people have primary schooling although there are still 21% of households that have either adult members that have not completed primary education, or children who are not attending school. Moreover, illiteracy is still a problem in rural Thailand, the rate being twice that of the population as a whole (8% compared to 4%). This is especially serious in Ban Dong where the illiteracy rate reaches 17%.

Despite the relatively high level of labour migration in the Northeast and commuting in the South (11% travel beyond nearby areas to work), family networks are very strong in rural communities with 88% of the households spending time with relatives on a weekly basis. Participants in WeD’s exploratory research into quality of life (Jongudomkarn and Camfield 2005) cited family relations as the most important aspect of a good life. However, the urban areas of the sample are populated by migrants and show, as expected, the highest percentage of households not spending time with close relatives. This suggests that progressive migration to urban centres might accelerate the breakdown of family networks, reducing self-esteem and emotional strength, which are likely to result in lower wellbeing.

Table 4 compares average wealth (assets) and intermediate needs deprivation (INDI) by site. It shows how Thai communities differ significantly in their level of basic needs satisfaction and material wealth. Concerning wealth, data on asset ownership illustrates that Ban Dong in the North-east is the poorest community and that Ban Chai Khao in the South is the richest.Footnote 7 Data on basic needs deprivation confirms that households in Ban Chai Khao, the wealthiest site, present high levels of basic needs satisfaction. It also shows that Nai Muang, a significantly poorer urban community, has nevertheless high satisfaction of basic needs.Footnote 8

Table 4 INDI, INDI components and asset by site

In general, households having a higher satisfaction of their basic needs are also wealthier than their counterparts. However, the INDI and the asset index are only moderately correlated (r = −0.32 at p < 0.001) for the whole sample and in Ban Tha the association is not significant. As Table 4 shows, INDI components linked to physical health are only weakly associated with material prosperity. In particular, the association is not significant in four out of seven communities in our sample. This illustrates the fact that in Thailand one can be relatively well-off in material terms and live in communities that do not have access to an efficient health care system and/or lack protection against economics shocks; thus, being non-poor and having needs scantily satisfied.

Correlations between happiness and the objective measures of wellbeing are low: well below the 0.45 (p < 0.05) correlation coefficient found by Biswas-Diener and Diener (2001) in their study of deprived people in Calcutta. For the whole sample, happiness and material wealth present a Spearman correlation coefficient (r S) of 0.2 at the 0.1% level of significance and the needs deprivation index shows an even lower correlation coefficient of −0.1 at the 0.1% level. People’s happiness seem only to increase in conjunction with basic needs satisfaction in Ban Tha (r S =  −0.168, p < 0.1) and Ban Laow (r S = −0.341, p < 0.05) in the North east of Thailand. In contrast, happiness increases with material wealth in most communities except in Ban Dong and Nai Muang in the North-east. These results suggest diversity in the relationship between needs deprivation, wealth and happiness in Thailand.

5 Happiness in Thailand: the Role of Wealth and Basic Needs Satisfaction

The way wealth and deprivation of basic needs are related to happiness in Thailand is explored further through an ordered Probit model, which also accounts for selected socio-demographic variables.Footnote 9 This is designed to model the choice between discrete alternativesFootnote 10 and has been applied extensively in previous economic studies (Ferrer-i-Carbonell and Frijters 2004). In general, it is assumed that there are N individuals (i = 1,….,N), with a vector xki containing observations on K independent variables that explain individuals’ perception of happiness.

The empirical specification is formulated in terms of a latent response variable, \( y_{i}^{*} \), which depends on individual perception and is defined by the following structural equation:

$$ y_{i}^{*} = \sum\limits_{k = 1}^{K} {\beta_{ki} } x_{ki} + \varepsilon_{i} \quad \varepsilon_{i} \sim NID(0,1) $$
(1)

where:

  • i: The surveyed individual

  • x ki : Independent variables that explain happiness

  • β k : Parameter that indicates the effect of x k on \( y_{i}^{*} \)

  • ε i : A normally distributed independent error term for household i

Let yi be a discrete random variable whose value ranges from 1 to 3. The happiness question’s categories are “not too happy”, “fairly happy” and “very happy”. Therefore, the ordered 3 alternatives is defined as follows:

$$ y_{i} = \left\{ {\begin{array}{*{20}c} 1 & {{\text{if}}\quad y_{i}^{*} < d_{1} } \\ 2 & {{\text{if}}\quad d_{1} \le y_{i}^{*} < d_{2} } \\ 3 & {{\text{if}}\quad d_{2} \le y_{i}^{*} } \\ \end{array} } \right.$$

where d 1 < d 2

The parameter d is the “threshold parameter” and the model is estimated using maximum likelihood method.Following the definition from Eq. 1, the empirical model uses the following exogenous variables.

  1. (a)

    Personal characteristics of the respondent such as age, gender, marital status and employment status.

  2. (b)

    Characteristics of the household such as the number of children, the dependency ratio and the presence of chronically ill in the household.

  3. (c)

    Basic needs levels estimated through the Intermediate Needs Deprivation Index (INDI).

  4. (d)

    Material wealth through an asset index.

  5. (e)

    Community of residence.

Thus, the empirical model is represented as:

$$ \begin{aligned} {\text{Y}}* & = {\text{F}}({\text{Household head characteristics}},{\text{Household characteristics}},{\text{basic needs}}, \\ & {\text{material wealth}},{\text{community}}) \\ \end{aligned} $$
(2)

where the dependent variable, Y*, is defined to take the values of 1, 2 or 3.Footnote 11

Model (2) was estimated in three steps in order to assess the additional information provided by the basic needs and material wealth variables as determinants of happiness. The first model consists of all the controlling variables plus both material wealth and INDI. The second model only excludes the material wealth variable and the third model considers all the controlling variables and the material wealth variable.

The results of the three Ordered Probit models, including parameter estimates and corresponding z-statistics are given in Table 5.

Table 5 Determinants of happiness in Thailand

The first model (1a) exhibits the highest explanatory power (pseudo R 2), denoting that both INDI and wealth variables are significant determinants of happiness. When comparing the findings from model 1b and model 1c, it is observed that adding INDI to a basic model (without assets or INDI) contributes less to explain happiness variations than adding assets (see changes in McFadden R 2).

5.1 Discussion of Findings

In this section we discuss the results of the preferred Probit model (Model 1a) which includes both INDI and asset variables. As the Likelihood ratio test shows, the independent variables contribute significantly to explaining variation in people’s happiness. They explain circa 8–10% of its variation (see McFadden R 2 and Pseudo R 2), which is common in studies of this type using discrete choice models (Verbeek 2000).

Relationships between dependent and independent variables are considered significant at the 0.05 level of significance unless indicated otherwise. The focus of the discussion is on the direction of the relationship and not on the marginal effects of each single variable. Main findings are discussed in relation to previous studies.

With regards to household head characteristics, age and type of job are the only significant variables.Footnote 12 Self-reported happiness in rural Thailand is affected negatively and significantly by age, as has been found in subsequent WeD research using the Positive and Negative Affect Scale (Woodcock et al. 2008) and in previous analysis of happiness determinants focussing on the Thai rural sites (Guillen-Royo and Velazco 2006). This finding is not uncommon in happiness studies although the econometric method used and the type of the dependent variable are said to explain some of the divergences found in the literature (Ferrer-i-Carbonell and Frijters 2004).Footnote 13 In Thailand old age is rarely a time for relaxation as the absence of state-provided social security and scarcity of pensions mean that older people need to continue working to support themselves, often in physically demanding roles (older people are disproportionately represented in the agricultural workforce). In the Northeast in particular older people not only have to look after themselves (and often a sick or disabled partner), but also care for the grandchildren left behind by migrant parents.

Studies in wealthy countries suggest that having a salaried job has a positive impact on people’s happiness compared to being self-employed or working at home. This also applies in developing countries where self-employed workers are paid little, often work in poor conditions, and cannot socialise or argue for better conditions with other workers (Graham and Pettinato 2002; Camfield et al. 2007). This is the case for our Thai sample where jobs with a relative degree of authority, office jobs and non-traditional occupations are the ones regarded as providing a better quality of life (Jongudomkarn and Camfield 2005).

Concerning household characteristics, this research finds that having a chronically ill person in the house (dummychronic) has a negative effect on the household head’s happiness. Half of the sample reported having someone suffering from chronic illnesses such as (in order of frequency): joint and muscle pain, diabetes, hypertension, allergies and heart disease. While the first category of chronic illness reflects a lifetime of agricultural and other manual labour, the remainder indicate Thailand’s growing prosperity and the rapid changes in diet, smoking, and drinking that have resulted from this (reflected in rising obesity rates among adults and children in Bangkok). These have begun to be addressed by the Universal Healthcare Scheme introduced in 2002; however, its effects will not be reflected in our survey, which also indicates that at time many people were still loathe to use formal healthcare services (Camfield 2009).

Table 5 shows how the location of the household matters for happiness. Compared to living in Nai Muang, the congested slum of Khon Khaen, living in all the southern communities in addition to living in Ban Dong and Ban Laow in the non-urban North-east makes a positive contribution to people’s happiness. This occurs controlling for income and level of basic needs satisfaction and might therefore be representing other structural variables inherent in Nai Muang such as the poor quality of the drainage which results in floods and the lack of social integration of the community. It might also be capturing the negative effect of living in a poor marginal neighbourhood of Khon Khaen, which provides a depressingly wealthy reference group.

As Table 5 indicates, having a lower level of basic needs satisfaction (a high value of the index) is negatively linked to happiness. Taking into account the geographical location and the wealth of the participants, basic needs matter for Thai people’s happiness. This finding highlights the relevance of accounting for basic needs levels when analysing people’s SWB in developing countries together with using a comprehensive measure such as the one derived from the THN. General indicators of socio-economic level usually disregard family or close relationships and focus on material measures such as asset ownership, education levels or sanitation facilities (Diener et al. 1995; Sahn and Stifel 2003; Rojas 2006), thus not accounting for the depth and breadth of basic needs.

As expected, household material wealth is also positively linked to happiness. This is consistent with most studies in developing countries that find significant relationships between the economic level of the household or individual and SWB (Biswas-Diener and Diener 2001; Graham and Pettinato 2002; Guillen-Royo and Velazco 2006). However, in this study material wealth is significantly associated with happiness controlling for the level of basic needs of the household. This implies that having more possessions is intrinsically valued by people beyond their use as basic needs satisfiers.

A popular Thai saying is that ‘money is king’ and this relates to the psychological security it offers—no need to ‘think too much’ or worry about fulfilling responsibilities to family or paying for healthcare. In fact, wellbeing was initially translated as ‘kin dee, yuu dee’, or ‘eat well, live well’, which emphasises its material basis. Money also enhances opportunities for social participation—making merit at the temple, hosting ‘papa’ or ‘kathin’ for community members, entering local politics, and even travelling for pleasure (‘bpai tiaow’). ‘Convenience goods’ such as fridges, motorcycles, and mobile phones have become increasingly desirable for the young, and also enable people to bypass inadequate local services (for example, the paucity of landlines).

The rise of materialist values in an increasingly unequal society has already begun to reduce the happiness of poor people in Thailand. Recent studies in Latin-American countries and Nepal (Graham and Felton 2006; Guillen-Royo 2007; Fafchamps and Shilpi 2006) point at the detrimental effect on happiness of ‘upward’ social comparison, which is more prevalent among the poor. Further research is needed to quantify this effect, but it serves to illustrate that, even for poor people, wealth has a range of subjective and relational meanings that extend beyond basic needs satisfaction.

6 Conclusion

This article has analysed how in the Thai context of recent economic growth and unequal development, objective wellbeing indicators such as wealth and basic needs have a significant and distinct link with happiness. It has drawn on the THN to approximate basic needs through a single indicator (INDI), which also enables disaggregation. Wealth has been estimated through an asset index collecting the number of assets owned by the household.

The descriptive analysis has shown that in Southern and North-east Thailand basic needs and wealth are weakly associated, although in most communities being richer implies a higher level of basic needs satisfaction. However, the low level of correlation and the fact that in the two poorest communities wealth and the INDI are not associated shows that using them interchangeably might hide important disparities such as: availability of public services, job opportunities, and the social security provided to those working in the formal sector.

The study of happiness determinants in Thailand clarifies the link between satisfaction of basic needs and wealth with happiness, accounting for the relevant socio-demographic variables. Basic needs are important for happiness, controlling for the location and wealth levels of the household. Wealth is also significantly linked to happiness taking into account the level of basic needs satisfaction and the socio-demographic variables included in the model. This supports our claim that satisfaction of basic needs and wealth should not be used interchangeably, even in poor contexts, as local structures and people’s understandings determine their relationship to SWB.

In Thailand basic needs as defined by THN represent the aspects that people consider most influential on their quality of life. Wealth implies psychological security, respect and opportunities for social participation. All these aspects are important for happiness but have different policy implications. For example, recognising the trade-off between increasing economic growth and incomes by encouraging manufacturing in urban and peri-urban centres and the effect this has on family relations in rural areas, an equally important determinant of SWB. The effects of illness on SWB, both directly through people’s experience of life as a patient or carer and indirectly through its impact on income, underlines the importance of continuing support for the Universal Healthcare scheme. This needs to be accompanied by a greater focus on the quality of basic services outside urban areas and government-funded equivalents to the social insurance offered by formal sector employers.

Finally, further research is needed to refine the indicators used, for example, we do not have measures of personality traits such as extroversion and neuroticism that are important determinants of SWB (Schimmack et al. 2002 in Camfield 2004). Moreover, the THN does not capture needs such as the need for social support or the need to be respected which are strongly associated with SWB (Tay and Diener 2011) and might also be linked to specific levels of wealth. This, together with the implications of wealth for social status, could explain the independent effect of wealth on happiness found in our research. Refining the study by expanding our concept of needs to incorporate non-basic or higher needs will no doubt contribute to our understanding of the meaning of basic needs and wealth in the Thai context.