Income inequality, status seeking and household savings in rural vietnam

Nội dung text: Income inequality, status seeking and household savings in rural vietnam

1. INCOME INEQUALITY, STATUS SEEKING AND HOUSEHOLD SAVINGS IN RURAL VIETNAM Tran Dang Nhan*1- Ong Nguyen Chuong 2- Nguyen Le Dinh Quy3 Abstract: This paper uses household panel data from rural Vietnam to examine the role of status seeking in the relationship between income inequality and savings. Results show that households living in provinces with higher income inequality are likely to save more to improve their social status, which is consistent with statusseeking hypothesis. Specifically, we find thatpoorer, younger, unmarried-headed, women-headed households as well as those taking part in economic activities have stronger motives to increase savings, after controlling for other standard determinants of household savings. Our findings suggest that public policies that explicitly reduce inequality might have adverse economic outcomes. Key words: Income inequality, status seeking hypothesis, household and consumption savings JEL Classification: D14; D91; H31; E21 1. INTRODUCTION Traditional economic models assume the dependence of utility on absolute income and consumption and completely ignore the effects of social comparison on individual decisions. For example, both the permanent income hypothesis (Friedman, 1957)and the life-cycle hypothesis (Modigliani, 1970)predict that individuals seek to smooth consumption throughout their lifetimes; hence, savings rate remains barely changed over time andis independent of their relative position in the income distribution. The life cycle hypothesis also supposes that the accumulation of wealthwill exhibit a hump-shaped pattern – that is low near the beginning of adulthood and during retirement, while peaking in working age.Other studies focus onsubsistence consumption in individual preferenceswhose mechanism shows that income inequality might be associated with savings(Musgrove, 1980;Ben-David, 1998; and Kraay and Raddatz, 2007). The idea is that subsistence level prevents poor households from saving because they spend their income for basic consumption needs, once theseare solved, they can save more. Several papers investigate that precautionary motive is another mechanism explaining the relation between distribution and savings(Skinner, 1988; Well, 1993; and Carroll and Kimball, 2001). Some families withuncertain income, such as unemployment andbusiness risks, will save a larger portion of current income to cope withthose incidents. Recent studies explorethe influences of relative concern on the link between inequality and savings.4In the wealth-is-status model by Cole et al., (1992), individual utility depends not only on absolute wealth but * This research is granted by Science and Technology Development Fund, University of Da Nang. Corresponding author: Email: nhantran@due.edu.vn; Tel: +84 (0)903 509 157; Address: 71 Ngu Hanh Son street, University of Economics – the University of Da Nang (UD-UE), Da Nang, Vietnam. University of Economics – the University of Da Nang (UD-UE), Da Nang, Vietnam. Institute of Public Policy Management (IPPM), National Economics University, Ha Noi, Vietnam. 1 Weiss and Fershtman (1998) define that social status is the rank of an individual or a group of individuals in a given society; the rank relies on commonly agreed-upon standard such as wealth, education, origin and occupation. We here employ income distribution as a social ranking.
2. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 1241 also on relative position in the distribution of income, and increased inequality over time gives a strong incentive to accumulate that stems from the increased future consumption. By incorporating status seeking into the model of Glomm and Ravikumar (2014), Pham (2005)shows that income inequality triggers savings by highly motivated individuals toupward social status. When the economicdisparity between the low and high-status groups widens, there should beastronger increase in thesaving effort of low status individualsthat would help them get into high status clubs.According to a much different approach byVeblen (1992) and Corneo and Jeann (1997), instead of spending on productive goods and savings, consumers try to signal social status though conspicuous consumption that refers to purchasing visible positional goods and luxuries. This also implies the impact of economic inequality on a family’s savings. A more equal society acquires less entry consumption level for the high-status groups, thereby the poor spend more on conspicuous goods to overtake others in the race (Hopkins and Kornienko, 2004, 2009). However, some of studiesofferthe oppositethat states that inequality is negatively related to savings in the existence of status-seeking.Corneo and Jeanne (2001) present that more inequality can diminish status motivesto the extent that it make the relatively poor more difficult to accumulate wealth and, at the same time, weakens the incentive of therelatively rich to assert their social status. Long and Shimomura (2004) illustrate that ifthe strength of relative wealthis larger in utility compared to that of consumption, the poor who have a strong desire to save will catch up with those that are initially richer.Kawamoto (2009) also discuss that the condition of catching-up as well as decreasing inequality over time is the poor’ more benefit from increasing their own income that,in turn, encouragesexpenditure on education, when average income rises(keeping up with Joneses).Frank et al. (2014) find that each individual’s spending relies on both his/ her own income and the spending of those with higher income. The authors predict that the expenditure of high income groups urges low income groups to spend more money.Therefore, rising inequality will cause an increase in conspicuous consumption of all groups and may even lead to a reduction in savings. Following these theories that have focused on status seeking, the empirical discussions has established mixed findings. Using aggregate quarterly US data for the years 1980 to 2003, Christen and Morgan (2005) find that increased income inequalityhas a positive impact on conspicuous consumption, and, instead of saving, poor families increase borrowing to keep consumption level relative to rich families.Jaikumar and Sarin (2015)present that poor rural households in India spend significantly on conspicuous goods and dissave owning to less access to alternatives to signal status like educational qualifications and savings,whenthey face the inequalityto rise.Drawing upon the theoretical work of Frank et al. (2014), Darku (2014) shows that there is a significant negative relationship between inequality and personal savings rates in Canada through what they call an expenditure cascade. However, there are several studies that provide different results. For example, Jin et al. (2011) employing Chinese household survey data reveals that poor and young people have a stronger motive to save that will put them in a higher social status. The authors also find that inequality has a negative impact on conspicuous consumption and it significantly influences education spending.Roychowdhury (2017)indicates that there isa negativelyconsistent relationship between conspicuous consumption and reference group inequality in rural India households, and that conspicuous spending is crowding out expenditure on education. To the best of our knowledge, no previous research empirically investigates the income inequality- status seeking-household savings relationship in a developing country like Vietnam.1Weadopt a simple 1 There are several different explanations of household saving behavior in Vietnam, including demographics (Nguyen et al., 2013), social capital (Newman et al., 2014), access to insurance (Gries and Dung, 2014).
3. 1242 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA model put forward by Pham (2005) and can analyze the role status seeking played in explaining thisrelation among rural Vietnamese households. Status-seeking hypothesis predicts that the positive influences of income inequality of reference group on household savings, for given income. We try to test this hypothesis grounded on the observations that according to the World Bank, the Gini index increased from 35.4 to 39.3 between the year 1998 and the year 2012, then considerably declined to 34.8 in 2014, while gross savings as percentage of gross national income raised by almost 10% between 1998 and 2012 and lightly reduced by approximately 1% to 32% in 2014.1We also hypothesize the effect of income inequality on savings to be higher in poorer, younger, women-headed, unmarried-headed and economic activity-engagedfamilies because they assignmore importance to going ahead. The underlying mechanism can be explained as follows. Agrowth in the reference group income inequality increases marginal returns fromwealth accumulation that thereby gives more satisfactionto households who lies at the bottom status groups. They are more concerned with savings for the future consumption to enter higher status groups faster. This,in turn, strengthens the saving incentives for those belong to the middle and top status groups to keep their social status. A possible consequence is thatsocial ranking positions of everyone may remaintied to the same, whilehousehold savings rise in theeconomy.2 However, because people have different status concerns,the level of inequality of the reference group may enlarge through time, which againincreaselow-status households’ motivation to save more in the status competition. We use household level panel data from the Vietnam Access to Resources Household Survey, conducted by the United Nations University World Institute for Development Economics Research (UNU-WIDER) between 2008 and 2014 in 12 provinces.In this paper, the reference group of a family is presented as allfamilies living in the same province. Our findings are that income inequality has a positive and significant impact on household savings, after controlling for income and other demographic characteristics; and that poorer, younger, unmarried-headed, female-headed and economic activity- engaged households are those whom stronger desire to upgrade their low social status through saving when facing a raise in income inequality in the province-reference group. In our paper, several measures of economic inequality, including Gini, Theiland Palma ratio are used to check robustness of its relationship with household savings. Some important policy implications can be drawn from this study. Evidently, province-income inequality increases household savings due to the stronger incentivesfor higher status seeking in the society. This also implies that at some certain level, inequality has a positive and indirect effect on economic growth. Policies aimed at inequality reduction maycause unintended side-effects. Furthermore, as noted above, disadvantaged groups (exception to households that do not engage in economic activities) are more susceptible to social comparison in Vietnam.The taxation of capital income and wealth on these groups may discourage this source of economic growth, whilepolicy interventions aimed to increasing their savings for productive purposes to benefit the poor and thus helping to fight poverty. The paper is organized as follows. Section 2 presents the theoretical model this paper follows. In next section, we describe the econometric model, survey data and provide measures of household savings and income inequality. Section 4 shows results of research and discussion, subsample analyses and robustness tests. In final section, wegive the conclusions. 1 See World Bank Open Data at 2 Hopkins and Kornienko (2004) also discuss the race of conspicuous consumption that keepseveryone in the same place.
4. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 1243 2. A SIMPLE MODEL ON THE RELATIONSHIP BETWEEN INCOME INEQUALITY, STATUS SEEKING AND HOUSEHOLD SAVINGS We draw upon the theoretical model developed by Pham (2005) which finds that an increase in income inequality will have a positive impact on savings through what is called status-seeking behavior. It can be assumed that there are two groups of agents in the economy. The size of first group and second group is and constant over time. All agents of the same group are identical, so that an agent may represent others of her group. Instead of using the Cobb-Douglas production function, for simplicity purposes, the stock of public capital does not enter the total output as following: Here are constant parameters.The variables and held by agent iare capital and labor force, respectively. As in Barro (1990),capital includes both physical and human capital. Each agent plays a role as both household and producer in the economy and supplies one unit of labor inelastically.Assume that capital is fully depreciated in one period. In other words, capital accumulated by agent iat period t+1, , is equivalent to her investment expenditure at t, : . Agentihas the utility function that consists of her consumption and social status: Where is the weight assigned to social status, is the weight placed on consumption, and is exogenous.Notice that two groups of agents have a different attitude towards social status, which become higher as the wealth of an agent ( rises and the average wealth of the society declines. Utility from consumption is while utility from social status is . The latter term shows that accumulation of wealth is associated with agent i’s welfare due to her enhanced social status. It can be written as: Where is the weight assigned to relative wealth and is the weight assigned to absolute wealth. Agent i maximizes (1) subject to the lifetime budget constraint: and By solving the first order conditions, then Combining equations (2) and (3) gives the following condition The left-hand term is the marginal cost of capital, which is equal to multiply the marginal utility of consumption by tradeoff between capital and consumption, at time t. The right-hand term is the discounted marginal benefit of raising each additional capital at timet +1. The marginal benefit equals the marginal product of capital times the marginal utility of consumption at time t+1 added to the marginal utility of capital at time t+1. Under optimal choice of agent i, the marginal cost and the marginal benefit must be equated.
5. 1244 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA Combining (4) and budget constraint gives the following solution: Let us define new variables Then, combining (5) and (6) gives the system At the steady-state, all variables, such as consumption and capital, rise at the same rate g. The steady state of the economy is given by Let be the share of wealth owned by agent 1 compared to total wealth, . We can write ., for any t , with . From equation (10) by considering gives us the system. Solving this system gives us Noting that h(s) is increasing with s For , implying , then we will get inequality . In that case, as the status-seeking motive of agent 1 is stronger, agent 1 will hold more wealth than agent 2 in the steady state, . Moreover, it explicitly verifies that
6. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 1245 This means that if there is an increase in the status-seeking motive of agent 1, then she will obtain the larger fraction of total wealth. Equation (9) implies that, X(s) is increasing with sof agent 1, independent with that of agent 2and a change in X(s) is increasing with a change in s. This suggests that if there is an increase in the status-seeking motive of agent 1 and/or agent 2, implying the reducing importance of consumption, correspondingly, then the savings level would increase in the economy. These results show that the incentives for status-seeking play an important role in explaining the wealth inequality.1 If agents have the same motive to wealth accumulation, they launch the same wealth level. The possible explanation is as the following. Firstly, the poor people feel more satisfied with the marginal increase in wealth than the rich people, when the marginal status utility of wealth, which is equal to , is inversely related to . Secondly, agent 1 with a higherstatus motiveemphasizes more on the utility from status than the utility from consumption. This leads her to higher accumulation of wealth. Taking an instance, agent 1 have a lower initial wealth endowment than agent 2 (i.e. , if , agent 1 will catch up with agent 2 as she has more satisfaction from the marginal increase in wealth. However, if , she catches up with agent 2 before she obtains a larger fraction of total wealth. This finding indicates that wealth inequalityrises over time due to agents with difference in status- seeking savings.Long and Shimomura (2004)argue that in the presence of status-seeking, income inequality declines because of poor agents’ catching up with rich agents. This similar result is found in Pham (2005)’s model, if agents assign the same importance to saving;or by using the special casewith the parameter (the AK-type production function) we can see all agents end up holding the same wealth whatever the initial distribution.Kawamoto (2009) investigates thatincome inequality is increasing withstatus attitudes to human capital accumulation. If agents with high-income spend more on education that eventually negatively affect the learning efforts of the low-income agents, the catching up will not occur.In the current model, the status motive of agent1 is independent with that of agent 2. If agent 1 has stronger desire to accumulate wealth, this would affect her savings and fraction of whole wealth. The latter in turn influences the agent 2’ wealth ratio but has no impact on the savings of this agent,and vice versa. In summary,our approach simulates every household with low-status as an agent (poorer, younger, female-headed households, etc.) that have stronger incentives for status-seeking savings, becausetheyget more satisfaction from an increase in marginal returns to wealth.This in turn encourages those belong to the higher- status groups to increase savings to maintain their social status. Consequently, in the economy, household savings is set to rise, while social status positions may stay at the same level. However, when people have 1 In this study, we consider income inequality as alternative measure for inequality of wealth.
7. 1246 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA different attitudes to social status, the level of income inequality will change over time. Therefore, households seeing rising income inequality within their reference group are likely tohave more savings. 3. EMPIRICAL STUDY Econometric model In our simple model, the status-seeking hypothesis is that for given income, increasing income inequality with any province-reference group leads to a rise in household savings. To test this hypothesis as well as other standard determinants of household savings in rural Vietnam, firstly, we estimate a baseline regression model: Where S is the annual average saving for household h in time t; income is monthly real per capita income; is a vector of household-specific variables, including characteristics of the household head, income, land ownership, the presence of a family enterprise, and the incidence of natural and economics income shocks; inequality represents the four different measures of income inequality (namely, Gini index, Theil L index, Theil T index and Palma ratio); are household fixed effects to control for all time-invariant household-specific characteristics; are time dummies; and is a statistical noise term. An increase in income inequality will lead to a rise in savings, hence, is expected to be positive. It can be seen that also shows the effect of inequality on the average propensity to save or savings rate, sincelog(savings rate) equals log(savings) minus log(income). is the income elasticity of savingsand expected to be positive, and if it is less than 1.0 means thatthe average propensity to save or savings rate declines with income.The variables such as savings, incomeand durables expenditure are under log transformation to reduce the impact of outliers and for convenience in interpreting parameter estimates. Data In this section, we use the balanced panel of data to perform the household fixed effects analyses of the impact of income inequality on the household savings. All the data used for the study were obtained from the Vietnam Access to Resources Household Survey (VARHS) implemented by the United Nations University World Institute for Development Economics Research (UNU-WIDER) between 2008 and 2014 in Vietnam.1The VARHS spreads 12 provinces (Dak Lak, DakNong, Dien Bien, Ha Tay, KhanhHoa, Lai Chau, Lam Dong, Lao Cai, Long An, Nghe An, PhuTho, and Quang Nam). To establish the balanced panel of 2,181 households, the same households interviewed every two years are included.However, in our paper, the number of households may be smaller due to missing data. Since VARHS did not collect total expenditure data, the standard measure of saving that equals income minus expenditure is unavailable. Alternative measure of saving collected reliable is the self-reported levels of savings. Families are asked about their saving stock at the interview date, twelve months ago, and how much they saved during the past 12 months. This allows to check misreporting of financial information. To measure income inequality, we use the Gini coefficient, which is one of the most commonly used income inequality indicators. Other measures of inequality are used to check robustness of our regression results, consisting of the Theil L index, Theil T index and Palma ratio - the ratio of the income share of the richest 10% divided by the income share of the poorest 40%.Both the savings and income data are adjusted 1 The data is publicly available at
8. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 1247 for price differences over time and between the different provinces in Vietnam. While the saving measure was expressed on an annual household basis, information is annually collected on per capita income from different sources (agriculture, wage, non-farm non-wage, transfers, etc.) that then was transferred to monthly figures. 4. ESTIMATION AND DISCUSSION OF RESULTS Baseline regression results The theoretical model shows a positive relationship between income inequality and savings in the existence of status seeking.Table 1 shows the baseline panel data regression results empirically examining the effect of income inequality on household savings. We focus on the results using the Gini coefficient as the measure for income inequality and control for income, household fixed effects and time varying household characteristics in all regressions. Column 1 describes the household savings-income relationship before any measures of income inequality are included.After controlling the effect of income in column 2, the results present that the Gini coefficient is positively associated with household savings, which suggests that the greater the income inequality of province-reference group, the greater the household level of savings. Column 3includes the differences in assets (land area, durables expenditure, redbook), in addition to income and province- level inequality.Column 4 reports the results for the specification in which we control for demographic characteristics, such asage, family business, marital status, the number of children, and head education. Column 5reports the results for the specification in which we control for economic and natural shocks, but do not include the differences in assetsand demographic characteristics.The estimates of these added variables give further evidence relating to the appropriateness of using our modelto estimate and thus the robustness of the income inequality effect. Our preferred specification reported in column 6 which consists of the full set of variables. Overall, we find a positive and statistically significant effect of income inequality on savings across all specifications with magnitudes of the coefficients ranging between 1.092 and 1.281 at the 5 % significance level, as expected.These magnitudes are economically meaningful. More specifically, our preferred specification indicates that whenthe Gini coefficient increases by 0.1, household savings or the average propensity to save rises by 9.2%.Notice that this does not necessarily imply that rural families in Vietnam seek to reduce conspicuous expenditure and education spending (important indicators of social status).Unfortunately, VARHS dos not record information that would enable us to measure and test these hypothesizes.1 Furthermore, the results are noticeable with the two variables: income and shocks. Firstly, per capital income per familythat strengthens income inequality has a positive and significant relation with household savings in all columns. More specially, a 1% increase in the per capita income of a household is associated with a 0.58 – 0.63 % increase in savings, which implies that the average propensity to save or savings rate rises by 0.37 – 0.42%. Secondly, all parameter estimates for economic and natural shocks have the unexpected sign and are highly significant. On the contrary to precautionary motive, the rural Vietnamese household strategies for coping with negatively economic and natural shocks are completely stronger enough to smooth consumption. Alternatively, it can suggest that the possibility of giving up of households 1 In our comparison paper, Jin et al. (2011) induce that reference group inequality has a negative (positive) effect on conspicuous consumption (savings) and it significantly (insignificantly) affects investment on education (subsistence consumption) in China due to the incentives of status seeking.
9. 1248 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA living in the rural areas is higher when they are more susceptible to shocksthan other households. Table 1: Regressions estimating the effect of income inequality on household savings Dependent variable: Log(household savings) 1 2 3 4 5 6 GINI 1.281 1.184 1.252 1.246 1.092 (0.533) (0.544) (0.535) (0.533) (0.545) Log(income) 0.618 0.621 0.579 0.634 0.623 0.594 (0.034) (0.033) (0.036) (0.035) (0.034) (0.037) Log(land area) -0.081* -0.071a (0.044) (0.044) Log(durables) 0.130 0.131 (0.022) (0.022) Redbook -0.056 -0.053 (0.082) (0.081) Family enterprise -0.111 -0.104 (0.055) (0.055) Age -0.004 -0.003 (0.004) (0.004) Married 0.127 0.123 (0.100) (0.099) Kids 0.066 0.068 (0.061) (0.063) Higher education -0.129 -0.125 (0.083) (0.085) Ethnic minority -0.010 -0.065 (0.221) (0.235) Household size -0.061 -0.065 (0.024) (0.024) Natural shock -0.098 -0.100 (0.038) (0.039) Economic shock -1.191 -0.170 (0.048) (0.049) Time dummies Yes Yes Yes Yes Yes Yes Constant 3.398 2.815 2.903 3.051 2.898 3.146 F stat 317.48 254.67 142.38 117.87 186.76 78.89 R2 within 0.216 0.217 0.226 0.219 0.221 0.230 R2 between 0.534 0.546 0.509 0.541 0.553 0.523 R2 overall 0.346 0.353 0.344 0.353 0.360 0.356 Number of household 2,181 2,181 2,152 2,181 2,181 2,152 Number of observations 7,236 7,236 6,950 7,214 7,236 6,931 Notes: Each model includes household and time fixed effects. Robust standard errors clustered at the household level in parenthesizes. denotes significance at the 1% level, denotes significance at the 5% level, * denotes significance at the 10% level, and a denotes significance at the 11% level. Source: Author’s calculations based on VARHS 2008-14 survey data.
10. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION Table 2: Regressions estimating the effect of income inequality on household savings: subsample analysis Dependent variable: Log(household savings) Income Age Gender Marriage Status Economic activity Not Poor Middle Rich >39.5 Female Male Married Yes No 39.5 married Gini 2.338 0.291 1.32a 1.67 1.178 2.71 0.815 2.719* 0.936a 1.339 5.136 -1.001 -0.905 -0.958 -0.73 -0.945 -1.393 -0.609 -1.457 -0.604 -0.549 -4.374 Log(income) 0.578 0.752 0.712 0.575 0.609 0.6 0.576 0.467 0.597 0.582 0.815a -0.06 -0.066 -0.074 -0.052 -0.064 -0.095 -0.04 -0.09 -0.041 -0.037 -0.52 Time dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Constant 1.69 2.366 2.645 3.246 2.585 2.492 3.558 4.555 3.102 3.12 -7.102 F stat 37.35 26.65 23.71 38.92 28.91 13.42 65.54 10.24 70.91 79.33 3.4 R2 within 0.297 0.251 0.215 0.247 0.214 0.209 0.239 0.195 0.238 0.237 0.332 R2 between 0.301 0.38 0.264 0.345 0.453 0.404 0.508 0.279 0.538 0.508 0.087 R2 overall 0.29 0.301 0.226 0.284 0.36 0.347 0.352 0.237 0.365 0.354 0.127 Number of household 693 713 712 1,450 1,344 541 1,768 527 1,839 2,132 138 Number of observations 2,058 2,354 2,436 3,632 3,299 1,407 5,524 1,238 5,693 6,708 223 Notes: Each model includes household and time fixed effects. Robust standard errors clustered at the household level in parenthesizes. denotes signifi- cance at the 1% level, denotes significance at the 5% level, * denotes significance at the 10% level, anda denotes significance at the 11% level.Controlling for income, shocks and demographic characteristics, according to the theories addressed. Source: Author’s calculations based on VARHS 2008-14 survey data. 1249
11. 1250 Table 3: Regressions estimating the effect of income inequality on household savings: different measures of income inequality Dependent variable: Log(household savings) Theil L Theil T Palma ratio Whole Whole Whole Poor Middle Rich Poor Middle Rich Poor Middle Rich sample sample sample Inequality 0.583* 1.359 -0.038 0.894* 1.034 1.297 0.987 1.022 0.116 0.202 0.042 0.159 (0.315) (0.572) (0.521) (0.544) (0.196) (0.386) (0.314) (0.328) (0.042) (0.079) (0.070) (0.070) Log(income) 0.62 0.686 0.754 0.694 0.624 0.682 0.768 0.696 0.621 0.686 0.757 0.695 (0.034) (0.056) (0.063) (0.063) (0.033) (0.056) (0.063) (0.063) (0.034) (0.056) (0.063) (0.064) Time dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Constant 3.175 2.284 2.156 2.6 2.973 2.309 1.678 2.519 3.1 2.287 2.023 2.531 HỘI THẢO KHOA HỌC QUỐC TẾ F stat 254.27 109.18 84.67 71.29 259.18 112.04 84.87 72.31 254.84 109.59 83.85 71.4 R2 within 0.217 0.281 0.242 0.192 0.221 0.285 0.247 0.195 0.218 0.282 0.243 0.193 R2 between 0.543 0.379 0.342 0.393 0.55 0.381 0.367 0.388 0.545 0.377 0.35 0.391 R2 overall 0.351 0.315 0.258 0.251 0.357 0.318 0.272 0.252 0.353 0.314 0.261 0.251 Number of household 2,181 716 716 715 2,181 716 716 715 2,181 716 716 715 KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA Number of observations 7,236 2,224 2,395 2534 7,236 2,224 2,395 2,534 7,236 2,224 2,395 2,534 Notes: Each model includes household and time fixed effects. Robust standard errors clustered at the household level in parenthesizes. denotes signifi- cance at the 1% level, denotes significance at the 5% level, and * denotes significance at the 10% level. Controlling for income, shocks and demographic characteristics, according to the theories addressed. Source: Author’s calculations based on VARHS 2008-14 survey data.
12. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 1251 Subsample analysis In this section,we will investigateif the relationship between economic inequality and household savings can vary across subgroupsbetween 2008 and 2014 in rural Vietnam. Table 2 presents the regression results. Firstly, the regressions based on these three income subsamples exhibit a positive and significant impact of income inequality on savings of the poor (p< 0.05), the effect on the middle is not significantly different from zero, and the coefficient of the rich is significantly positive at the 11% level.1 It should be noted that the impact of inequality on the rich (1.32) is much smaller than that for the poor (2.34).The incentive for status savings is stronger for poorer householdssince they receive more benefit from increasing wealth. By contrary, the rich have weaker motive to defend their social status that has stayed at the highest group. Particularly, the middle seems to be not sensitive to increased inequality in their reference group; this may be because they feel satisfied with the current social position. These results are different compared to previous studies (Corneo and Jeanne, 2001;Christen and Morgan, 2005; andJaikumar and Sarin, 2015) that show that poor households engage more in conspicuous consumption than savings. Banerjee and Duflo (2007) argue that lack of access to financial and labor market is not a significant explanation for saving behavior of the poor. They have strong motive to save more by spending less on alcohol, tobacco, festivals and food items such as sugar and tea. A key policy challenge for the poor is to increase the share of savings for investment purposes (for example, buying a new machine, storing food in the family shop and expenditure in education of their children). Secondly, the effect of income inequality on savings is much higher for households headed by younger individuals, female, and unmarried individuals. Prior research also supports these findings. The analysis by Jin et al. (2011) show more incentives for wealth accumulation of poor and young people. Moav and Neeman (2012 ) present women’s more self-control over recourses and less conspicuous consumptionthan men. Unmarried people seem to be more responsive to others’ assessment of their social status as they search for spouses and/or to social pressure to moving up(Roychowdhury, 2017). Next, table 2 also shows that households taking part in economic activities such as agriculture, non-farm enterprise are likely to be more concerned about their social status.2The effect of inequality on savings of households having economic activity is statistically significant (1.339). This implies the possibility of giving up of those who did not hastheir own earnings because they stay much far from the entry wealth level for a higher status group (Corneo and Jeanne, 2001), but there may be due to small observations in the sample of this latter type of household. Robustness check Table 3 describes the regressions estimating the effect of income inequality of reference group, by different measures of inequality that are used to check robustness for our baseline results.For whole sample, using other measures of inequality, we find the similar pattern of the effects of inequality on household savings. For instance, when the Theil L (Theil T) index increases by 0.1, household savings rises by 5.8% (10,34%). Using Palma ratio as the measure of inequality presented on table 3 gives the same results. For subsample by income group, overall, the positive and significant impact on the poor also shows the stronger status seeking motive of this group compared to the middle class of which the coefficient is 1 Three income groups are measured by per capita income per household in 2008. The poor, middle, and rich are the bottom, middle, and top one-third of the income distribution, correspondingly. 2 Households with no economic activities that do not earn an income from any of the economic activities. Their main income source was from public and private transfers.
13. 1252 HỘI THẢO KHOA HỌC QUỐC TẾ KHỞI NGHIỆP ĐỔI MỚI SÁNG TẠO QUỐC GIA statistically insignificant, except for the Theil T index of which the influence appears to three groups. The coefficient of the rich became positive and statistically significant for three measures of inequality, but still weaker than that of the poor (p < 0.1). This is not surprising that while the poor people have stronger desire to enter the high-status group by increasing savings, the rich people may try to keeptheir social status. CONCLUSION The literature on social status motive and economics growth highlights the mechanism that social status seeking can stimulate more savings and thus promote economic growth. It can be assumed that people care about both consumption and social status. The former shows the dependence of their welfare positively on consumption, while the latter is an increasing function of both absolute wealth and relative wealth. We empirically examine whether the rising income inequality of the reference group causes savings of household to increase, particularly in developing countries. The data is mainly derived from the Vietnam Access to Resources Household Survey conducted by the United Nations University World Institute for Development Economics Research(UNU-WIDER) from 2006 to 2014 in 12 provinces. In this paper, we define the reference group is a resident group to which a household is compared in the same province. Our study confirms the importance of the motive for social status in the relationship between income inequality and household savings. Particularly, ourmain finding is thatincome inequality in household’s province- reference group is positively associated withsavings, while income and other demographic characteristics are held unchanged. We also find that poorer, younger, unmarried-headed, women-headed, economic activity-engaged households have stronger motives for the sake of status seeking savings. Relating to inequality measurement, the study uses the Gini coefficient and other measures including Theil and Palma ratio to check robustness of its relationship with household savings. Since savings is among important factors of household’s welfare and income in the long run, studying on its determinantsbrings meaningful recommendation. The study suggests that public policies explicitly aimed to reduction in inequality might have unexpected economic outcomes. REFERENCES Banerjee, A. V, & Duflo, E. (2007). The economic lives of the poor.Journal of Economic Perspectives, 21(1), 141–167. Barro, R. J. (1990). Government spending in a simple model of endogeneous growth. Journal of Political Economy, 98(5), S103–S125. Ben-David, D. (1998). Convergence clubs and subsistence economies. Journal of Development Economics, 55, 153–169. Carroll, C. D., & Kimball, M. S. (2001). Liquidity constraints and precautionary saving. Retrieved from nber.org/papers/w8496.pdf Christen, M., & Morgan, R. M. (2005). Keeping up with the Joneses: analyzing the effect of income inequality on consumer borrowing. Quantitative Consumer Marketing and Economics, 3(2), 145–173. Cole, H. L., Mailath, G. J., & Postlewaite, A. (1992). Social norms, savings behavior, and growth. Journal of Political Economy, 100(6), 1092–1125. Corneo, G., & Jeann, O. (1997). Conspicuous consumption, snobbism and conformism. Journal of Public Economics, 66, 55–71. Corneo, G., & Jeanne, O. (2001). Status, the distribution of wealth, and growth. The Scandinavian Journal of Economics, 103(2), 283–293. Darku, A. (2014). Income inequality, status seeking, and savings rates in Canada. Canadian Studies in Population, 41(3), 88–104. Frank, R. H., Levine, A. S., & Dijk, O. (2014). Expenditure cascades. Review of Behavioral Economics, 1, 55–73. Friedman, M. (1957). A theory of the consumption function. Princeton, New Jersey: Princeton University Press.
14. INTERNATIONAL CONFERENCE STARTUP AND INNOVATION NATION 1253 Glomm, G., & Ravikumar, B. (2014). Growth-inequality trade offs in a model with public sector R&D. The Canadian Journal of Economics, 27(2), 484–493. Gries, T., & Dung, H. Van. (2014). Household savings and productive capital formation in rural Vietnam : insurance vs. social network. Modern Economy, (5), 878–894. Hopkins, E., & Kornienko, T. (2004). Running to keep in the same place: consumer choice as a game of status. American Economic Review, 94(4), 1085–1107. Hopkins, E., & Kornienko, T. (2009). Games and economic behavior status, affluence, and inequality: rank-based comparisons in games of status. Games and Economic Behavior, 67(2), 552–568. Jaikumar, S., & Sarin, A. (2015). Conspicuous consumption and income inequality in an emerging economy: evidence from India. Marketing Letters, 26(3), 279–292. Kawamoto, K. (2009). Status-seeking behavior, the evolution of income inequality, and growth. Economic Theory, 39(2), 269–289. Kraay, A., & Raddatz, C. (2007). Poverty traps, aid, and growth. Journal of Development Economics, 82(2), 315–347. Moav, O., & Neeman, Z. (2012). Saving rates and poverty: the role of conspicuous consumption and human capital. Economic Journal, 122(563), 933–956. Modigliani, F. (1970). The collected papers of Franco Modigliani. Cambridge: The MIT Press. Musgrove, P. (1980). Income distribution and the aggregate consumption function. Journal of Political Economy, 88(3), 504–525. Newman, C., Tarp, F., & Broeck, K. Van Den. (2014). Social capital, network effects, and savings in rural Vietnam. Review of Income and Wealth, 60(1), 79–99. Nguyen, T. M., Hong Nhat, N., Trinh, T. A., Phung, M. D., & Le, T. S. (2013). Demographics and saving behavior of households in rural areas of Vietnam: an empirical analysis. Journal of Econonimics and Development, 15(2), 5–18. Pham, T. K. C. (2005). Economic growth and status-seeking through personal wealth. European Journal of Political Economy, 21(2), 407–427. Roychowdhury, P. (2017). Visible inequality, status competition, and conspicuous consumption: Evidence from rural India. Oxford Economic Papers, 69(1), 36–54. Skinner, J. (1988). Risky income, life cycle consumption, and precautionary savings. Journal of Monetary Economics, 22(2), 237–255. Van Long, N., & Shimomura, K. (2004). Relative wealth, status-seeking, and catching-up. Journal of Economic Behavior and Organization, 53(4), 529–542. Veblen, T. (1992). The theory of the leisure class. New York: Routledge. Weiss, Y., & Fershtman, C. (1998). Social status and economic performance: A Survey. European Economic Review, 42, 801–820.