Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 1.
Published in final edited form as: Soc Sci Med. 2022 Jan 13;294:114722. doi: 10.1016/j.socscimed.2022.114722

Comparison of three indices of relative income deprivation in predicting health status

Krisztina Gero 1,*,#, Aki Yazawa 2,3,#, Naoki Kondo 4,5, Masamichi Hanazato 6,8, Katsunori Kondo 6,7, Ichiro Kawachi 2
PMCID: PMC8844705  NIHMSID: NIHMS1773538  PMID: 35065345

Abstract

Relative income deprivation (RID) is a known risk factor for poor health. Previous research has proposed several measures to assess RID, e.g., Income Rank and the Yitzhaki Index. Hounkpatin et al. (2020) presented a new approach – the CR˜i index – to account for the observation that individuals are more sensitive to the differences in incomes of others who are closer to them, rather than to comparisons with incomes of others far above them. Using a Japanese nationwide cohort of older adults (n=62,438; mean [SD] age: 73.0 [5.6] years), this study compared the performance of alternative indices of RID in predicting health outcomes (depressive symptoms, functional capacity, and self-rated health), as well as the use of alternative CR˜i index weights (α weight range: −0.9 to 0.9). When 0< α <1, higher income differences lead to a more significant increase in relative deprivation, while when −1< α <0, excessively high incomes contribute less to the relative deprivation of lower income individuals in the same reference group. Results showed that all measures of relative income deprivation were associated with deteriorating mental and physical health among older Japanese adults. However, while the CR˜i index consistently outperformed the Yitzhaki Index, this did not hold true invariably when compared to the Income Rank – depending on the health outcome and the reference group. Also, while negative α parameters showed a good statistical fit in most models, the findings were not conclusive – the best-fitting CR˜i weight parameters ranged from −0.9 to 0.9. Therefore, a clear direction for the contribution of higher incomes to relative deprivation could not be established based on the study population.

Keywords: Relative deprivation, Income inequality, Depression, Functional capacity, Self-rated health, Gerontology

Introduction

Lack of income affects health not only through absolute deprivation – i.e., lack of access to essential material goods such as food and shelter – but also via mechanisms of relative deprivation (Boyce et al., 2010; Daly et al., 2015; Wilkinson, 1997; Wood et al., 2012). Relative income deprivation (RID) can harm health because even if the individual has sufficient income to avoid deprivation in the absolute sense (i.e., avoid hunger, homelessness), he/she can be still deprived of the opportunity to participate fully in society – for example, as a result of being unable to afford access to high-speed internet during a pandemic lockdown. A sense of relative deprivation is also triggered by upward social comparisons of income, which can give rise to feelings of unfairness, frustration, shame, stress, anxiety, and resentment (Kawachi et al., 2002; Marmot and Wilkinson, 2001; Wilkinson, 1996), resulting in deteriorating health through stress-coping behaviors (McEwen and Seeman, 1999).

The concept of ‘relative deprivation’ has been defined as “the extent of the difference between the desired situation and that of the person desiring it” (Runciman, 1967). Several measures have been proposed to assess relative deprivation, often based on differences in income among people. One such measure is the Income Rank, indicating an individual’s position in the income hierarchy (Brown et al., 2008). The concept of Income Rank is based on the notion that people usually do not know exactly how much others earn, and hence judgement becomes anchored by one’s rank conferred by income. Income Rank is therefore insensitive to the magnitude of the income gap between individuals (Boyce et al., 2010). Previous studies have shown that those with lower income rank had worse self-rated health, higher allostatic load, and were more likely to be obese and to experience mental distress and suicidal ideation or attempts (Daly et al., 2015; Hounkpatin et al., 2016; Wetherall et al., 2015; Wood et al., 2012). Another well-established RID measure – the Yitzhaki Index – calculates the average of the distance between the index individual and everybody else within the same reference group with higher incomes (Adjaye-Gbewonyo and Kawachi, 2012; Yitzhaki, 1979). Contrary to the income rank, the Yitzhaki Index assumes that the larger the gap between an individual’s income and that of others within a reference group, the more significant the psychological burden for the relatively deprived. The Yitzhaki Index has also been shown to be associated with worse health status, including higher mortality, poor self-rated health, more frequent mental-health services utilization, functional disability, smoking, and obesity (Kawachi et al., 1999; Kondo et al., 2008, 2009, 2015; Subramanyam et al., 2009).

According to the prospect theory (Wakker 2010), however, individuals should be more sensitive to the differences in income of others who are closer to them, rather than to those further away (law of diminishing marginal sensitivity). To quote Hounkpatin et al: “the feeling of deprivation felt by a professor of economics may be increased more by the presence of a departmental colleague earning just a little more than they do than by the much higher salary of the university president.” This contradicts the assumption underpinning the Yitzhaki Index, whereby the larger the gap between the income of an individual and wealthier comparators, the more significant the degree of relative deprivation (i.e., directly proportional). To address this issue, Hounkpatin et al. (2020) proposed a new approach – the CR˜i index, which weights the income differences among people within a reference group. Based on the prospect theory (Wakker 2010) and the social comparison theory (Festinger 1954), Hounkpatin et al. suggested that people tend to compare themselves to those with similar though somewhat higher incomes instead of those with excessively high earnings since they provide the most diagnostic information for self-evaluation. Hence, incomes much higher than an individual’s own will receive less weight for his/her relative deprivation than incomes closer to his/her own.

To test Hounkpatin et al.’s theory, we used a Japanese nationwide cohort study established among older adults (1) to investigate which RID index (Income Rank, Yitzhaki Index, CR˜i index) best predicts health status, and (2) to find the CR˜i index weight that provides the best fit to capture the impact of relative deprivation on health status. We hypothesized that the contribution of higher incomes to relative deprivation might decrease with their distance above the income of the relatively deprived.

Methods

Study population

Data were obtained from the 2010 and 2013 waves of the Japan Gerontological Evaluation Study (JAGES), a nationwide cohort established in 2010 to identify factors pertinent for healthy aging. Participants aged 65 years or older who were free of disability were recruited from 31 municipalities of 12 prefectures, identified through residential registries and long-term care insurance databases. The baseline JAGES survey was mailed to 169,215 randomly selected community-dwelling older adults between August 2010 and January 2011. The response rate was 66.3% (n=112,123), with 102,869 participants providing valid information on their age, sex, and municipality of residence. Between October 1 and December 2 in 2013, 77,714 individuals who answered the baseline questionnaire were re-contacted, out of whom 63,462 participants filled-out the follow-up survey (response rate: 81.7%). After excluding 1,024 participants with discrepant information on their sex between the 2010 and 2013 waves, the JAGES panel data included 62,438 respondents (28,868 men and 33,570 women) who answered both the 2010 and 2013 surveys. The study protocol for the JAGES project was approved by the Ethics Committee for Research on Human Subjects at Nihon Fukushi University, Japan.

Health outcomes

In the JAGES 2010 and 2013 surveys, depressive symptoms were measured using the short version of the Japanese Geriatric Depression Scale (GDS-15). The GDS-15 is a self-reported questionnaire of 15 queries with “yes” (score of 1) or “no” (score of 0) as possible answers. Summing the score of all 15 questions, higher scores indicate higher depressive symptomatology (Nyunt et al., 2009). Changes in depressive symptoms were assessed by calculating differences in GDS-15 scores reported in 2010 and 2013.

The Tokyo Metropolitan Institute of Gerontology Index of Competence (TMIG-IC) was used to evaluate higher-level functional capacity – the ability to perform complex cognitive, physical, and social activities in daily life. The TMIG-IC is a self-administered questionnaire of 13 items with a simple “yes” (score of 1) or “no” (score of 0) format. Summing the score of all 13 items, higher scores indicate higher functional capacity (Koyano et al., 1991). A change score for functional capacity was calculated for each respondent by deducting the TMIG-IC score reported in 2010 from that recorded in 2013.

Respondents’ self-rated health was assessed with the following question: “How is your current health status?”. The possible answers ranged from “poor” to “very good” on a 4-point Likert scale. Poor subjective health was defined as a perceived “not so good” or “poor” health status. A binary variable was constructed for all analyses where those whose self-rated health changed from good in 2010 to poor in 2013 received a score of 1, while those with no change or improvement in their subjective health status received a score of 0.

In the analytic sample, the standardized Cronbach’s α was 0.83 for the GDS-15 scale and 0.73 for the TMIG-IC scale in the JAGES 2010 survey. The corresponding standardized Cronbach’s α in the 2013 survey were 0.84 and 0.77, respectively.

Relative Income Deprivation (RID)

Three alternative measures were used and compared to assess relative income deprivation based on self-reported household income information collected in 2010 and 2013.

  1. The Income Rank, calculated by:
    IncomeRanki=i−1n−1
    where i − 1 is the number of respondents with incomes lower than individual i’s, while n is the total number of respondents within i’s reference group. Representing the ordinal position of i’s income on the income hierarchy, values range between 0 as the lowest and 1 as the highest rank within the reference group.
  2. The Yitzhaki Index, defined as (Yitzhaki, 1979):
    YitzhakiIndexi=1N∑j(yj−yi)IijIij={1,ifyi<yj0,ifyi≥yj
    where yi is respondent i’s income, yj is the earnings of respondent j with income higher than individual i’s, and N is the total number of respondents in i’s reference group. The Yitzhaki Index represents the average difference between a respondent’s income and the earnings of all other respondents with higher income within the same reference group.
  3. The CR˜i index, estimated by (Hounkpatin et al., 2020):
    CR˜i=∑k=i+1k=n(yk−yi)α(n−1)
    where yi is respondent i’s income, yk is the earnings of respondent k whose income is higher than individual i’s, α is a weighting parameter, and n is the total number of respondents in i’s reference group. When α = 0, the CR˜i index equals to the complement of the Income Rank, while when α = 1, the CR˜i index approximates the Yitzhaki Index (the denominator is n − 1 for the CR˜i index and N for the Yitzhaki Index). Based on previous studies, the assessed values for the α parameter were restricted to a range between −1 and 1 in the analytic models to test for a concave increasing function (0< α <1), where higher income differences lead to more significant health consequences for the relatively deprived, or a concave decreasing function (−1< α <0), where excessively high incomes contribute less to the relative deprivation of lower income individuals in the same reference group (Figure 1) (Bossert and D’Ambrosio, 2014; Chakravarty and Chakraborty, 1984; Esposito, 2010; Hounkpatin et al., 2020; Paul, 1991; Podder, 1996; Stark et al., 2017). For example, when α = 0.5, the influence of the highest income on CR˜$40,000 will be larger when it is $100,000 ((100,000−40,000)0.5 = 245) than when it is $50,000 ((50,000−40,000)0.5 = 100), while when α = −0.5, the influence of the highest income on CR˜$40,000 will be smaller when it is $100,000 ((100,000−40,000)−0.5 = 0.004) than when it is $50,000 ((50,000–40,000)−0.5=0.01).

Figure 1.

Figure 1.

Open in a new tab

The impact of higher incomes to relative deprivation

yi: respondent i’s income, yk: the earnings of respondent k whose income is higher than individual i’s, α: a weighting parameter

The JAGES study did not ask the participants for information about reference groups, i.e., the people to whom individuals compare themselves. Instead, we created several alternative reference groups based on combinations of socio-demographic factors, including: a) same sex and age (mean group N [SD]: 16,906 [4,344] in 2010; mean group N [SD]: 15,700 [1,185] in 2013), b) same sex and municipality of residence (mean group N [SD]: 2,115 [1,460] in 2010; mean group N [SD]: 2,436 [1,551] in 2013), c) same sex, age, and municipality of residence (mean group N [SD]: 1,135 [827] in 2010; mean group N [SD]: 1,223 [776] in 2013), and d) same sex, age, education, and municipality of residence (mean group N [SD]: 422 [321] in 2010; mean group N [SD]: 464 [346] in 2013) (Ishida, 2011; Kondo et al., 2015). For example, it was assumed that a 65-year-old man living in metropolitan Kobe would compare his earnings to other men with similar characteristics, and not to an 80-year-old woman living in a rural area on the Chita peninsula.

Information on household income was collected in 15 pre-defined categories. The midpoint of each income category was set as the household income of each respondent. Since the highest income category was open-ended, a median-based Pareto curve estimate was calculated to retrieve an approximate midpoint (Parker and Fenwick, 1983). Equivalized household income was defined as the gross household income divided by the square root of the number of household members.

Covariates

The following time-varying covariates were included in the analytic models as potential confounders: age (continuous variable in years), marital status (married or other), employment status (having a paid job or being retired/never had a job), and equivalized household income (continuous variable in Japanese Yen). Time-invariant variables, such as sex and years of education, were omitted from our statistical analysis. Information on all covariates was collected at the baseline in 2010 and the follow-up in 2013.

Statistical Analysis

Sequential regression multiple imputation (SRMI, also called “imputation by fully conditional specification” or “imputation using chained equations”) was used to account for missing information (missingness: 0% − 12.8%), creating ten imputed datasets. Multivariable-adjusted first differences models were fitted to estimate the associations between changes in the three relative income measures and changes in depressive symptoms or functional capacity during the follow-up. All models were adjusted for changes in time-varying socio-demographic factors such as age, marital status, employment status, and income. Associations between changes in the levels of relative income deprivation and worsened subjective health (vs. no change or improved self-rated health) were assessed using multiple logistic regression models, adjusted for all above-described time-varying covariates. Combined effect estimates were computed based on the ten imputed datasets. In all models, change scores for relative deprivation indices were standardized, and Income Rank was reverse coded to improve comparability among the three relative income measures. A sensitivity analysis was also conducted using complete cases (Supplementary Table 1). SAS version 9.4 (SAS Institute Inc., Cary, NC), SAS-callable IVEware (Imputation and Variance Estimation Software) version 0.2, and Stata 16.1 (calculation of the Yitzhaki and CR˜i indices) (StataCorp, College Station, TX, USA) were utilized to conduct all statistical analyses.

Results

The descriptive characteristics of the study participants are summarized in Table 1. At baseline, the mean (SD) age of the respondents was 73.0 (5.6) years, and slightly more than half of them (53.8%) were female. The reported mean equivalized household income of the participants decreased from 2,343,000 (1,654,000) Japanese Yen (JPY) (approximately 23,430 USD) in 2010 to 2,203,000 (1,570,000) JPY in 2013, with the proportion of employed participants decreasing from 23.9% to 19.3% during the same period. The respondents’ mean GDS-15 score and the percentage of participants reporting poor subjective health slightly increased between 2010 and 2013 – from 3.3 (3.2) to 3.4 (3.3) and from 16.6% to 19.5%, respectively – while the mean functional capacity score decreased from 11.7 (1.8) to 11.4 (2.1) during the follow-up.

Table 1.

Participant characteristics

N = 62,438 Baseline Survey (2010) Follow-up Survey (2013) P a
N (%) Mean (SD) N (%) Mean (SD)
Mean age (years) 73.0 (5.6) 75.4 (5.6)
Sexb
  Male 28,868 (46.2)
  Female 33,570 (53.8)
Educationb
  < 6 years 1178 (1.9)
  6 to 9 years 28,025 (44.9)
  10 to 12 years 21,556 (34.5)
  ≥ 13 years 11,254 (18.0)
  Other 425 (0.7)
Employment status
  Having a paid job 14,925 (23.9) 12,069 (19.3) <0.0001
  Being retired/Never had a job 47,513 (76.1) 50,369 (80.7)
Marital status
  Married 45,835 (73.4) 43,555 (69.8) <0.0001
  Other 16,603 (26.6) 18,883 (30.2)
Mean equivalized income (JPY) 234.3 (165.4) 220.3 (157.0) <0.0001
Depressive symptoms 3.3 (3.2) 3.4 (3.3) <0.0001
Functional capacity 11.7 (1.8) 11.4 (2.1) <0.0001
Poor self-rated health 10,355 (16.6) 12,199 (19.5) <0.0001
Open in a new tab
a

One-Sample (Paired-Difference) t-Test for continuous variables and McNemar’s Test for binomial proportions.

b

Empty cells at follow-up for time-invariant variables.

Abbreviations: SD, Standard Deviation; JPY, Japanese Yen

Table 2 summarizes the effect estimates for the associations between relative income deprivation and three health outcomes: depressive symptoms, functional capacity, and subjective health. The results show that higher relative income deprivation was associated with worsening depressive symptoms, regardless of which RID measure was used or how the reference group was defined. However, while the Income Rank and the CR˜i index showed similar associations with functional capacity and self-rated health, higher Yitzhaki Index was statistically significantly associated with functional decline in only two out of the four reference groups and showed no significant associations with poor subjective health. Each standard deviation increment in the Income Rank change during the follow-up was associated with an 0.060 to 0.088-point increase in the depression change score and an −0.027 to −0.035-point decrease in the functional capacity change score between 2010 and 2013, as well as an 0.056 to 0.066 increase in the log-odds for worsening subjective health. The corresponding effect estimates were 0.044 to 0.076, −0.022 to −0.054, and 0.043 to 0.068, respectively, for the CR˜i index; and 0.038 to 0.078 for depressive symptoms and −0.025 to −0.029 for functional capacity changes for the Yitzhaki Index. Based on the mean Akaike Information Criterion (AIC) statistics, models with the Income Rank and the CR˜i index showed the best fit for depressive symptoms and subjective health, while the CR˜i index led to the best model fit for functional capacity with −0.9 as the best fitting α parameter. Also, the CR˜i index consistently showed a better model fit compared to the Yitzhaki Index for all three health outcomes.

Table 2.

Associations between depressive symptoms / higher-level functional capacity / self-rated health and relative income deprivation

RID reference group Depressive Symptoms Functional Capacity Self-Rated Health
Estimate (p-value) Mean AICa Estimate (p-value) Mean AICa Estimate (p-value)b Mean AICa
Gender and age-group (N = 62,438)
 Income Rank (1SD) 0.080 (0.0001) 291594.80 −0.035 (0.0052) 229523.22 0.059 (0.0298) 40267.79
 Yitzhaki Index (1SD) 0.078 (<0.0001) 291591.15 −0.025 (0.0463) 229526.84 0.039 (0.1401) 40270.54
 CR˜i (1SD) 0.072 (<0.0001) 291589.09 −0.025 (0.1359) 229522.13 0.048 (0.0283) 40267.01
  Weight 0.2 −0.9 −0.2
Gender and area (N = 62,438)
 Income Rank (1SD) 0.087 (<0.0001) 291590.87 −0.027 (0.0347) 229527.68 0.066 (0.0137) 40266.14
 Yitzhaki Index (1SD) 0.074 (<0.0001) 291594.00 −0.019 (0.1068) 229530.08 0.034 (0.1817) 40271.29
 CR˜i (1SD) 0.073 (<0.0001) 291593.91 −0.022 (0.0556) 229525.80 0.043 (0.0464) 40267.41
  Weight 0.9 −0.9 −0.9
Gender, age-group, and area (N = 62,438)
 Income Rank (1SD) 0.088 (<0.0001) 291590.03 −0.035 (0.0045) 229523.27 0.066 (0.0144) 40266.07
 Yitzhaki Index (1SD) 0.077 (<0.0001) 291591.42 −0.029 (0.0184) 229524.46 0.039 (0.1420) 40270.46
 CR˜i (1SD) 0.076 (<0.0001) 291579.59 −0.054 (0.0004) 229482.93 0.068 (0.0045) 40257.30
  Weight −0.9 −0.9 −0.9
Gender, age-group, education, and area (N = 62,395) c
 Income Rank (1SD) 0.060 (0.0013) 291409.51 −0.021 (0.0953) 229359.99 0.056 (0.0251) 40241.15
 Yitzhaki Index (1SD) 0.038 (0.0158) 291416.10 −0.013 (0.2509) 229362.06 0.022 (0.3357) 40246.23
 CR˜i (1SD) 0.044 (0.0010) 291409.68 −0.036 (0.0025) 229338.34 0.052 (0.0121) 40236.60
  Weight −0.9 −0.9 −0.9
Open in a new tab

Models are adjusted for changes in age, marital status, income, and employment status.

a

Mean AICs were calculated based on the goodness-of-fit statistics of ten imputed datasets.

b

Regression log odds coefficients (p-values) for worsening self-rated health vs. no change or improved self-rated health.

c

Respondents who were alone in their comparison group were excluded from the analysis.

Abbreviations: RID, Relative Income Deprivation; AIC, Akaike Information Criterion; SD, Standard Deviation

When conducting complete case analysis, the number of eligible participants markedly decreased from 62,438 to 28,623. The results for depressive symptoms did not change significantly compared to the findings based on multiply imputed data. However, while the direction of the associations between changes in relative income deprivation and changes in functional capacity scores and self-rated health remained the same, most of the effect estimates did not reach statistical significance (Supplementary Table 1).

Discussion

In this sample of older Japanese adults, all three measures of relative income deprivation were associated with mental and physical health status. Our main findings are twofold. First, in our study, the novel CR˜i index consistently outperformed the Yitzhaki Index, while its performance was comparable to the Income Rank measure depending on the health outcome and the reference group. Second, we did not find strong evidence on the best-fitting α parameters, though negative weighting showed a good statistical fit in most analyses, especially for models predicting functional capacity and self-rated health. However, this was not the case for depressive symptoms, and overall the best-fitting α parameters ranged from −0.9 to 0.9 in the statistical models. Given that mental health (depressive symptoms in this study) by its nature is thought to be the most reflective of one’s perceptions and is consistently reported to be associated with relative deprivation (Gero et al., 2017; Ladin et al. 2010), we conclude that a clear direction for the contribution of higher incomes to relative deprivation could not be established based on the study population. More specifically, we did not find conclusive evidence to support our initial hypothesis based on the prospect theory proposing that the contribution of higher incomes to an individual’s relative deprivation might decrease with their distance above (i.e., −1< α <0), or its alternative, stating that as the income gap increases, the contribution to relative deprivation rises (i.e., 0< α <1).

The associations between Yitzhaki Index/relative rank and health outcomes echo previous studies conducted in Japan and other countries, e.g., on depression (Gero et al., 2017), functional disability (Kondo et al., 2009), and self-rated health (Kondo et al., 2008). Moreover, the results were also in keeping with the findings of Hounkpatin et al. (2016) and Gero et al. (2017), showing that Income Rank better predicted health status than the Yitzhaki Index, indicating that among older people, health was more closely related to an individual’s relative position in the income hierarchy than the magnitude of income differences within a reference group. There is some evidence for sensitivity to relative rank in judgments made in areas of psychology, which supports our finding (Stewart et al. 2006). People may not be that sensitive to the quantitative difference with others, but rather judge themselves by relative position (i.e., the number of people above and below the target item on the dimension of interest).

The finding that Income Rank better predicted health status than the Yitzhaki Index might also partly explain the inconclusive findings on the α parameters; older – especially retired – adults might no longer compare themselves with others based on current income differences. Suls and Mullen (1982) proposed that the tendency to engage in social comparisons weakens during the life course, peaking around young adulthood and decreasing from middle- to older age. Callen et al. (2015) showed that older adults had lower levels of social comparison tendency and personal relative deprivation (i.e., individual differences in people’s beliefs and feelings associated with comparing their outcomes with the outcomes of similar others) as well. Therefore, future studies incorporating younger age groups, especially working-aged adults, are warranted. It is also possible that the value of the α parameter – and whether the CR˜i index outperforms income rank – varies across datasets and reference groups, potentially depending on factors such as the cultural context of populations or actual income levels. Future work needs to explore to whom people compare their incomes.

The following limitations should be considered when interpreting the study findings. First, although we have used longitudinal data, we cannot make strong causal inferences about the observed associations since those with ill health may have lower income (i.e., reverse causality). Second, although we have presented the results for several sets of hypothetical reference groups, these may not be the actual groups to whom individuals compare themselves (Guyer and Vaughan-Johnson 2020). For example, some people might be more sensitive to social comparisons across generations, e.g., whether they are doing better than their parents. Third, we only focused on social comparisons based on income. In reality, individuals make comparisons based on additional factors, including wealth and consumption (which we did not assess). A further limitation is that we used the midpoint of the income categories instead of the actual income. This may have prevented us from considering the impact of comparisons with people in the same income category who have slightly higher incomes, which could have the strongest impact on health in Hounkpatin et al.’s theory.

In conclusion, the alternative formulations of RID consistently outperformed the Yitzhaki Index, while its correlation was to a similar extent with health outcomes to the existing indices (i.e., Income Rank) with no additional consistent predictive power as a result of weighting the social comparisons. Future research should also focus more sharply on people who are more sensitive to social comparisons, e.g., people who routinely engage in social media (Instagram, Facebook).

Supplementary Material

Supplementary Material

Highlights.

  • Relative income deprivation was consistently associated with mental and physical health

  • The novel CR˜i index consistently outperformed the Yitzhaki Index

  • The performance of the CR˜i index was comparable to the Income Rank

  • No conclusive evidence was found on the direction or magnitude of the best-fitting α parameter

Acknowledgements

Funding: NIH grant R01 AG042463 and 2 R01 AG042463-06; Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (KAKENHI 23243070, KAKENHI 22390400, and KAKENHI 24390469); Health Labour Sciences Research Grant from the Japanese Ministry of Health, Labour, and Welfare H22-Choju-Shitei-008 and H24-Choju-Wakate-009; and grant S0991035 from the Strategic Research Foundation Grant-Aided Project for Private Universities from the Japanese Ministry of Education, Culture, Sports, Science, and Technology.

Footnotes

The authors have no conflicts of interest to declare.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Adjaye-Gbewonyo K, Kawachi I, 2012. Use of the Yitzhaki Index as a test of relative deprivation for health outcomes: A review of recent literature. Social Science & Medicine 75, 129–137. 10.1016/j.socscimed.2012.03.004 [DOI] [PubMed] [Google Scholar]
  2. Bossert W, D’Ambrosio C, 2014. Proximity-sensitive individual deprivation measures. Economics Letters 122, 125–128. 10.1016/j.econlet.2013.11.009 [DOI] [Google Scholar]
  3. Boyce CJ, Brown GDA, Moore SC, 2010. Money and Happiness: Rank of Income, Not Income, Affects Life Satisfaction. Psychol Sci 21, 471–475. 10.1177/0956797610362671 [DOI] [PubMed] [Google Scholar]
  4. Brown GDA, Gardner J, Oswald AJ, Qian J, 2008. Does Wage Rank Affect Employees’ Well-being? Industrial Relations: A Journal of Economy and Society 47, 355–389. 10.1111/j.1468-232X.2008.00525.x [DOI] [Google Scholar]
  5. Callan MJ, Kim H, Matthews WJ, 2015. Age differences in social comparison tendency and personal relative deprivation. Personality and Individual Differences 87, 196–199. 10.1016/j.paid.2015.08.003 [DOI] [Google Scholar]
  6. Chakravarty SR, Chakraborty AB, 1984. On indices of relative deprivation. Economics Letters 14, 283–287. 10.1016/0165-1765(84)90096-X [DOI] [Google Scholar]
  7. Daly M, Boyce C, Wood A, 2015. A social rank explanation of how money influences health. 34, 222–230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Esposito L, 2010. Upper boundedness for the measurement of relative deprivation. Review of Income and Wealth 56, 632–639. 10.1111/j.1475-4991.2010.00396.x [DOI] [Google Scholar]
  9. Festinger L, 1954. A theory of social comparison processes. Human Relations 7, 117–140. 10.1177/001872675400700202 [DOI] [Google Scholar]
  10. Gero K, Kondo K, Kondo N, Shirai K, Kawachi I, 2017. Associations of relative deprivation and income rank with depressive symptoms among older adults in Japan. Soc Sci Med 189, 138–144. 10.1016/j.socscimed.2017.07.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Guyer JJ, Vaughan-Johnston TI, 2020. Social Comparisons (Upward and Downward), in: Zeigler-Hill V, Shackelford TK (Eds.), Encyclopedia of Personality and Individual Differences. Springer International Publishing, Cham, pp. 5011–5015. 10.1007/978-3-319-24612-3_1912 [DOI] [Google Scholar]
  12. Hounkpatin HO, Wood AM, Brown GDA, 2020. Comparing indices of relative deprivation using behavioural evidence. Social Science and Medicine 259, 112914. 10.1016/j.socscimed.2020.112914 [DOI] [PubMed] [Google Scholar]
  13. Hounkpatin HO, Wood AM, Dunn G, 2016. Does income relate to health due to psychosocial or material factors? Consistent support for the psychosocial hypothesis requires operationalization with income rank not the Yitzhaki Index. Social Science & Medicine 150, 76–84. 10.1016/j.socscimed.2015.12.008 [DOI] [PubMed] [Google Scholar]
  14. Ishida A, 2011. A quantitative model of relative deprivation and reference group: An application of Yitzhaki’s individual relative deprivation index. Sociological Theory and Methods 26, 371–388. 10.11218/ojjams.26.371 [DOI] [Google Scholar]
  15. Kawachi I, Kennedy BP, Wilkinson RG, 1999. Crime: social disorganization and relative deprivation. Social Science & Medicine 48, 719–731. 10.1016/S0277-9536(98)00400-6 [DOI] [PubMed] [Google Scholar]
  16. Kawachi I, Subramanian SV, Almeida-Filho N, 2002. A glossary for health inequalities. Journal of Epidemiology & Community Health 56, 647–652. 10.1136/jech.56.9.647 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kondo N, Kawachi I, Hirai H, Kondo K, Subramanian SV, Hanibuchi T, Yamagata Z, 2009. Relative deprivation and incident functional disability among older Japanese women and men: prospective cohort study. Journal of Epidemiology & Community Health 63, 461–467. 10.1136/jech.2008.078642 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kondo N, Kawachi I, Subramanian SV, Takeda Y, Yamagata Z, 2008. Do social comparisons explain the association between income inequality and health?: Relative deprivation and perceived health among male and female Japanese individuals. Social Science & Medicine 67, 982–987. 10.1016/j.socscimed.2008.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kondo N, Saito M, Hikichi H, Aida J, Ojima T, Kondo K, Kawachi I, 2015. Relative deprivation in income and mortality by leading causes among older Japanese men and women: AGES cohort study. J Epidemiol Community Health 69, 680–685. 10.1136/jech-2014-205103 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Koyano W, Shibata H, Nakazato K, Haga H, Suyama Y, 1991. Measurement of competence: reliability and validity of the TMIG Index of Competence. Archives of Gerontology and Geriatrics 13, 103–116. 10.1016/0167-4943(91)90053-S [DOI] [PubMed] [Google Scholar]
  21. Ladin K, Daniels N, Kawachi I, 2010. Exploring the Relationship Between Absolute and Relative Position and Late-Life Depression: Evidence From 10 European Countries. The Gerontologist 50, 48–59. 10.1093/geront/gnp065 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Marmot M, Wilkinson RG, 2001. Psychosocial and material pathways in the relation between income and health: a response to Lynch et al. BMJ 322, 1233–1236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. McEwen BS, Seeman T, 1999. Protective and Damaging Effects of Mediators of Stress: Elaborating and Testing the Concepts of Allostasis and Allostatic Load. Annals of the New York Academy of Sciences 896, 30–47. 10.1111/j.1749-6632.1999.tb08103.x [DOI] [PubMed] [Google Scholar]
  24. Nyunt MSZ, Fones C, Niti M, Ng T-P, 2009. Criterion-based validity and reliability of the Geriatric Depression Screening Scale (GDS-15) in a large validation sample of community-living Asian older adults. Aging Ment Health 13, 376–382. 10.1080/13607860902861027 [DOI] [PubMed] [Google Scholar]
  25. Parker RN, Fenwick R, 1983. The Pareto Curve and Its Utility for Open-Ended Income Distributions in Survey Research. Social Forces 61, 872–885. 10.2307/2578140 [DOI] [Google Scholar]
  26. Paul S, 1991. An index of relative deprivation. Economics Letters 36, 337–341. 10.1016/0165-1765(91)90043-K [DOI] [Google Scholar]
  27. Podder N, 1996. Relative Deprivation, Envy and Economic Inequality. Kyklos 49, 353–376. 10.1111/j.1467-6435.1996.tb01401.x [DOI] [Google Scholar]
  28. Runciman W, 1967. Relative deprivation and social justice : a study of attitudes to social inequality in twentieth-century England. 10.2307/2091734 [DOI]
  29. Stark O, Bielawski J, Falniowski F, 2017. A class of proximity-sensitive measures of relative deprivation. Economics Letters 160, 105–110. 10.1016/j.econlet.2017.08.002 [DOI] [Google Scholar]
  30. Stewart N, Chater N, Brown GDA, 2006. Decision by sampling. Cognitive Psychology 53, 1–26. 10.1016/j.cogpsych.2005.10.003 [DOI] [PubMed] [Google Scholar]
  31. Subramanyam M, Kawachi I, Berkman L, Subramanian SV, 2009. Relative deprivation in income and self-rated health in the United States. Social Science & Medicine 69, 327–334. 10.1016/j.socscimed.2009.06.008 [DOI] [PubMed] [Google Scholar]
  32. Suls J, Mullen B, 1982. From the cradle to the grave: Comparison and self-evaluation across the life-span. Psychological Perspectives on the Self 1, 97–125. [Google Scholar]
  33. Wakker PP, 2010. Prospect Theory: For Risk and Ambiguity. Cambridge University Press, Cambridge. 10.1017/CBO9780511779329 [DOI] [Google Scholar]
  34. Wetherall K, Daly M, Robb KA, Wood AM, O’Connor RC, 2015. Explaining the income and suicidality relationship: income rank is more strongly associated with suicidal thoughts and attempts than income. Soc Psychiatry Psychiatr Epidemiol 50, 929–937. 10.1007/s00127-015-1050-1 [DOI] [PubMed] [Google Scholar]
  35. Wilkinson RG, 1997. Socioeconomic determinants of health. Health inequalities: relative or absolute material standards? BMJ 314, 591–595. 10.1136/bmj.314.7080.591 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wilkinson RG, 1996. Unhealthy Societies: The Afflictions of Inequality. Routledge. [Google Scholar]
  37. Wood AM, Boyce CJ, Moore SC, Brown GDA, 2012. An evolutionary based social rank explanation of why low income predicts mental distress: A 17 year cohort study of 30,000 people. Journal of Affective Disorders 136, 882–888. 10.1016/j.jad.2011.09.014 [DOI] [PubMed] [Google Scholar]
  38. Yitzhaki S, 1979. Relative deprivation and the Gini coefficient. Quarterly Journal of Economics 93, 324. 10.2307/1883197 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS1773538-supplement-Supplementary_Material.docx (19.4KB, docx)

RESOURCES