Infectious diseases and interpersonal trust: international evidence

doi:10.4236/health.2011.34037

Paper Menu >>

Journal Menu >>

Vol.3, No.4, 206-210 (2011) Health

doi:10.4236/health.2011.34037

Infectious diseases and interpersonal trust:

international evidence

Diego Gonzalez-Medina1, Quan V. Le2*

Seattle University, 901 12th Avenue, Seattle, USA; *Corresponding Author: lequ@seattleu.edu

Received 18 January 2011; revised 23 March 2011; accepted 27 March 2011.

ABSTRACT

The objective of this paper is to investigate the

relationship between infectious disease and

trust, hypothesizing a negative relationship. In-

terpersonal trust is defined as the aggregate

response that fellow citizens are trustworthy.

We explore stigma as a channel in the relation-

ship. We apply cross-country regression analy-

sis on a sample of 54 countries. We test our

hypothesis using data on selected infectious

diseases from the World Health Statistic s ( WHS)

published by the World Health Organization

(WHO) and data on trust from the World Values

Surveys (WVS). We cre ate an index of infectious

disease using factor analysis. The OLS regres-

sion equation includes control variables of in-

come inequality, per capita income and human

capital. The empirical results are considerably

robust showing that higher cases of infectious

diseases are negatively associated with trust

when controlling for macroeconomic and social

variables.

Keywords: Trust; Stigma; Social Capital; Infectious

Disease; HIV/AIDS

1. INTRODUCTION

The topic of trust is relatively new to the field of so-

cial science and public health. Trust positively facilitates

our social institutions, which rely considerably on col-

laboration. In their seminal paper on trust and growth,

Zak and Knack [1] explain trust as a driver for lowering

transaction costs by reducing oversight costs with as-

sumed aggregate effects. It can do this because trus t is a

force behind social capital formation rewarding coordi-

nation within social structures. Wang et al. [2] define

trust as a form of cognitive social capital, meaning that it

predisposes people to work in a mutually beneficial man-

ner. Their research shows that there is strong and com-

plex relationship between trust and health [2], which we

seek to investigate further. In this paper we focus on this

relationship in the context of in fectious disease. Previous

literature has provided strong empirical evidence that

social support promotes preventative action and treat-

ment of infectious disease [2,3]. Our objective is to in-

vestigate the correlation of infectious disease and aggre-

gate trust levels, hypothesizing a negative relationship.

Using a sample of 54 countries, we test our hypothesis

using data on selected infectious diseases from the World

Health Statistics (WHS) published by the World Health

Organization (WHO) and data on trust from the World

Values Surveys (WVS).

In previous studies, relationships between trusting

behavior and infectious disease were examined in the

context of public health [2-7]. Though samples were

limited to specific coun tries or diseases, stigma rose as a

common theme. The interaction between infectious dis-

ease, stigma and trust implies stigma as the main actor

between the others.

Stigma is created from “an attribute that is deeply dis-

crediting” of the infected population, provoking mistrust

or sympathy, fear or support [8]. Where interpersonal

trust is a reward of social collaboration, stigma disrupts

it by invoking widespread prejudice, discrimination and

disregard against infected members of society and those

associated with them [9]. Thus, the presence of stigma

undermines trust on an individual level, with assumed

aggregate effects1. Yet, stigma also leads infected people

to conceal their disease and making it unlikely for them

to seek testing, treatment and support [3]. Such inaction

generally increases the load of infectious disease since

effective treatment is not taken. Therefore, stigma has

the power to both undermine aggregate trust and in-

crease a country’s incidence of infectious disease.

The research on the stigma surrounding tuberculosis

1We elevate the relationship between stigma and interpersonal trust up

to aggregate trust levels because stigma, by definition, must be widely

held. Otherwise, it is indistinguishable from individual prejudice.

D. Gonzalez-Medina et al. / Health 3 (2011) 206-210

207207

in Nepal and Somalia illustrates how a widespread dis-

ease can sabotage interpersonal trust [6,8]. Since tuber-

culosis is a curable disease, it b enefits highly from effec-

tive social institutions. These would have the effect of

expediting successful identification and treatment of

tuberculosis. In one interview from the study, support is

clearly lacking: “…they want that you also have

TB…because they know that they will be isolated and

they become very vindictive. They want everyone to have

it…He will trick you…the person with TB will drink, spit

inside and throw the rema ining back in the pot…” [6]

Previous studies in Nepal, Réunion Island, and of

Caribbean immigr ants in the Un ited Kingdo m found fear

of contamination and presence of stigma [5,7,8]. One

woman from the Caribbean expressed how it caused her

to lose her job as a teacher: “You’ve got AIDS. I don’t

want nobody to come and burn my school down, so it’s

best if you leave.”[5]

Left to its own devices, widespread mistrust and fear

can introduce discrimination into the healthcare system.

This has large implications for the spread of infectious

diseases. For example, breaches of confidentiality and

refusal to exam an infected patient leads to delays in

treatment or complete evasion [5]. When social institu-

tions like these become a source of discrimination, the

infected are left on their own to find support [9]. Thus it

is clear that stigma has the effect of both diminishing

trust and raising levels of infectious disease. And given

its dominance in previous studies, it is assumed to be the

main actor behind the negative relationship between trust

and infectious disease.

In the following sections, we provide cross-national

evidence displaying a negative relationship between in-

fectious diseases and aggregate levels of interpersonal

trust. Following the two-tier methodology of Zak and

Fakhar [10], we use WVS data on interpersonal trust to

capture generalized trust and test it against country level

data of infectious diseases. We are, in essence, testing whe-

ther trust (at a personal level) has a relationship with

infectious disease that scales up to the country level. Ex-

planatory variables of per capita income, income ine-

quality and human capital are also taken into account.

Specifically, we use secondary education enrollment as

an adequate proxy for human capital, citing previous

growth literature [11].

2. DATA AND METHODS

We attempt to provide testable implications of our

hypothesis using relevant data on infectious diseases and

interpersonal trus t in a cro ss-section of countries. We use

second level data on selected infectious diseases from

the World Health Statistics (WHS) published by the

World Health Organization (WHO) in 2009 [12].2 There

are 18 officially reported infectious diseases based pri-

marily on the availability of data in 2007.3 This is the

most comprehensive data available despite its limita-

tion.4 The International Health Regulations reported

some diseases, while countries and the WHO monitor

other diseases in the context of specific control programs

[12]. The database reveals that the Eastern Mediterra-

nean region has the highest cases reported, while the

region of the Americas has the lowest number of cases

reported. The database also reports that lower middle

income group has the highest number of cases reported,

while the high income group has the lowest number of

cases reported.

Because the dataset includes a large number of infec-

tious disease cases, examining them one-by-one does not

provide sufficiently strong evidence to test the relation-

ship between infectious diseases and interpersonal trust.

It is more feasible to construct a statistical indicator us-

ing data reduction method on the cases of infectious

diseases. We construct an index of infectious diseases by

employing factor analysis. Principal component analysis,

the most common form of factor analysis, is used to ex-

tract the first principal component based on the largest

The data on trust are obtained from the World Values

Surveys (WVS) 1990-2000 [1,13-15]. The WVS con-

tains data from thousands of respondents from both de-

veloping and developed countries.5 The respondents in

each country respond to the question using the native

language, and the questions correspond to impressions of

the respondents’ own countries: “Generally speaking,

would you say that most people can be trusted, or that

you can’t be too careful in dealing with people?” This

question captures “interpersonal trust”, describing whether

two randomly selected individuals trust each other. This

general phrasing of trust captured in the WVS survey

question suggests that this is a reasonable cross-country

measure of trust. This trust variable has been used in a

2World Health Statistics (2009), Table 3: Selected Infectious Diseases,

p. 59-69. Geneva: WHO. The numbers are officially reported, but

vary greatly in quantity, representatives, comparability and informa-

tion value (WHS, p. 5 9 , 2009).

3H5N1 influenza cases were reported in 2008. WHS differentiates

between zero cases reported and no information available for a country

where possible.

4There is a limitation when using this dataset. WHO recognizes tha

there are inadequacies in the measurement of cases of infectious dis-

eases [12]. The report states that it is difficult to understand the sever-

ity of infection in a country solely by the recorded number of occur-

rences. This inadequacy can be due to the nature of the disease itself or

due to the nature of data collection [12]. In addition, the disease itsel

can be difficult to report (e.g. H5N1 influenza, Japanese encephalitis)

without specific laboratory testing that is not always available [12].

Depending on the country, data collection can also be a significan

challenge.

5See the WVS’s website for further technical information about the

questionnaire at: www.wordvaluessurvey.org.

D. Gonzalez-Medina et al. / Health 3 (2011) 206-210

208

cross-country contex t in several studies [1,10,13] despite

its limitation. 6 Given the availability o f the data fro m the

WVS, we are able to compare the levels of trust to cases

of infectious diseases across countries in this study. The

data vary noticeably between 54 rich and poor countries.

In poor countries such as Uganda, Tanzania, and the

Philippines, less than 10% of the respondents say their

fellow citizens are trustworthy, more than 65% of the

respondents in Sweden and Denmark believe this is so.

If societies are at high risk of contracting deadly dis-

eases from the afflicted individuals, stigma and dis-

crimination can sabotage interpersonal trust. Thus, high

scores on the infectious disease index are associated with

lower trust levels. For instance, the trust levels in Swe-

den and Denmark are remarkably high with low scores

on the infectious disease index, whereas the trust levels

in Uganda, Tanzania, and the Philippines are considera-

bly low with high scores on the infectious disease index.

We also utilize other explanatory variab les that have a

strong effect on trust such as per capita income [13]. Per

capita income is expected to take a positive sign. Other

variables also included in the regression analysis are

income inequality and the level of human capital. In-

come inequality plays a role in discouragement of social

cohesion amongst individuals causing class distrust

through polarization [16]. Thus, societies with high in-

come inequality exhibit lower trust. We used the Gini

coefficient of income as a proxy for income inequality.

The sign is expected to be negative. Human capital ef-

fects trust in the way that it enhances social networks

and presents rewards for integrity. These rewards given

on the basis of meeting performance criteria with integ-

rity establishes the fundamental social infrastructure for

trust upon which an eco nomy can be built. We used per-

centage of gross secondary school enrollment as a proxy

for human capital. Secondary education provides not

only the tools of reading, writing and mathematics, but

also lays the foundations of on-going learning and de-

velopment, allowing individuals to make constructive

criticisms and rightful judgments. This leads us to expect

human capital to take a positive sign. Data for per capita

income, Gini coefficient, and gross secondary school

enrollment are the taken from the World Development

Indicators [17].

3. RESULTS

In this paper we hypothesize that societies with high

levels of infectious diseases are more likely to respond,

in aggregate, that their fellow citizens are less trustwor-

thy. Table 1 reports OLS regressions of infectious dis-

eases on trust controlling for other explanatory variables.

In equation 1, the index of infectious diseases explains

10% of the variation in trust. The coefficient estimate for

infectious diseases has a negative and significant effect,

as hypothesized. In equation 2, the infectious disease

index remains statistically significant with the expected

sign when per capita income is included in the regression.

By adding per capita income to the regression, the ex-

planatory power increases to 34% of the variation in

trust. Income inequality is added to equation 3 as a con-

trol variable. The result shows that the index of infec-

tious diseases has a negative and remains statistically

significant per expectation. The coefficient on income

inequality is negative and significant, suggesting dispar-

ity in income decreases the levels of trust. This equ ation

explains 20% of the variation in trust. Last but not least,

when human capital is taken into account in equation 4,

the infectious disease index has a negative, but insig-

nificant effect on trust. However, per expectation, the

coefficient on secondary education is positive and sig-

nificant, suggesting that the skills attainted in secondary

school are poised to develop critical thinking that are

necessary for trust to triumph through truth.7 This equa-

tion explains 29% of the variation in trust.

Table 1. OLS Regression: Infectious Diseases on Interpersonal

Trust (excluded HIV/AIDS/STIs cases)

Independent

variables (1) (2) (3) (4)

Constant 12.278**

(5.312) 10.832**

(4.571) 39.099**

(11.263) –7.606

(7.062)

Infectious diseases –58.718***

(22.266) –33.065*

(20.069) –40.098*

(22.806) –0.792

(25.469)

Per capita

income 0.001***

(0.0002)

Income

inequality

–0.640***

(0.239)

Secondary

school

enrollment

0.371***

(0.097)

No. of

observations 54 53 50 52

Adjusted R2 0.10 0.34 0.20 0.29

Notes: The dependent variable is Interpersonal Trust. White heteroskedas-

ticity-constant standard errors in parentheses. ***Significant at 1%; **sig-

nificant at 5%; *significant at 10%.

6For example, Wang et al. [2] survey trust by measuring levels o

trustworthiness and mistrust separately. They find that a low level o

trust does not necessarily mean a high level of mistrust. They suggest

that when surveying trust, researchers asking questions regarding trust

should be careful to separate those emphasizing mistrust from those

emphasizing trust and report them differently. This is because their

research shows that the effects of low levels of trust can differ from

those of hi

h mistrust

[

]

7We also included primary education in the regression.The result

reveals that the infectious disease index has a negative and significant

effect, as expected. However, the coefficient on primary educationis

negative and insignificant. Possible explanation for this is that primary

education is more universal than secondary education across countries.

Thus, secondary education was much more effective in explaining

trust.

D. Gonzalez-Medina et al. / Health 3 (2011) 206-210

209209

Table 2. OLS Regression: Infectious Diseases on Interpersonal

Trust (included HIV/AIDS/STIs cases).

Independent

variables (1) (2) (3) (4)

Constant 16.691***

(3.882) 12.666**

(3.297) 42.630***

(11.468) –7.291

(7.209)

Infectious

diseases –40.052***

(15.454) –21.913*

(13.286) –22.112*

(12.784) –4.479

(16.777)

Per capita

income 0.001***

(0.0001)

Income

inequality

–0.623**

(0.260)

Secondary

school

enrollment

0.358***

(0.094)

No. of

observations 52 51 49 51

Adjusted R2 0.10 0.40 0.20 0.29

Notes: The dependent variable is Interpersonal Trust. White heteroskedas-

ticity-constant standard errors in parentheses. *** Significant at 1%; **

significant a t 5 %; * s i gnificant at 10%.

WHO does not report HIV/AIDS/STIs in the WHS

[12], yet it is one of the serious infectious diseases.

Stigma and discrimination are common problems facing

people living with HIV/AIDS/STIs. Because of such

stigma and discrimination, HIV/AIDS/STIs is in turn

associated with decreased trust. We extend our analysis

to include HIV/AIDS/STIs in the principal component

analysis to extract a new index of infectious diseases.

The OLS regression results are reported in Table 2 with

the new index. In equation 1, the index of infectious

diseases is significantly correlated with trust per expec-

tation at the 1% level. This equation explains 10% of the

variation in trust. The index of infectious diseases re-

mains negative and statistically significant at the 10%

level when income per capita is included in equation 2.

This equation explains 40% of the variation in trust.

When income inequality is added to equation 3, the in-

dex of infectious diseases remains robustly significant at

the 10% level. When human capital is tak en into accou nt

in equation 4, the infectious disease in dex has a negative,

but insignificant effect on trust.

The statistical results indicate that high levels of in-

fectious diseases have a significant effect on lowering

trust levels, suggesting that infectious diseases can sa-

botage interpersonal trust causing a decline in social

capital.

4. DISCUSSION AND CONCLUSION

The empirical results indicate that higher cases of in-

fectious diseases are negatively associated with trust in a

sample of 54 countries with robust estimates. The find-

ings are strong enough to motivate a shift in perspective

when studying factors of infectious disease. We seek to

underscore the relevance of social capital in this discus-

sion.

Specifically, the literature surveying experiences of

infected people suggests that stigma and low interper-

sonal trust are caused by factors of fear and perceived

threat of infection [3,4,6,8,9]. And since risk perception

necessarily precedes stigma, we affirm that it is the in-

cidence and subsequent risk perception that causes stig-

matizing reactions and falling levels of interpersonal

trust.

Targeting stigma, as the literature suggests, may be an

effective measure to improve interpersonal trust and de-

crease infectious disease. Targeting stigma enhances in-

terpersonal trust through open dialogue and approval of

social bonds once broken by stigma. Similarly, targeting

sources of stigma reduces the levels of infection diseases

by making healthcare more approachable and trustwor-

thy in areas where healthcare was once a source of stigma.

To reap the benefits of trust, efforts are required to coun-

teract discrimination and stigma against those targeted.

REFERENCES

[1] Zak, P. and Knack, S. (2001) Trust and growth. The Eco-

nomic Journal, 470, 295-321.

doi:10.1111/1468-0297.00609

[2] Wang, H., Schlesinger, M. and Hsiao, W.C. (2008) The

flip-side of social capital: The distinctive influences of

trust and mistrust on health in rural China. Social Science

& Medicine, 68, 133-142.

doi:10.1016/j.socscimed.2008.09.038

[3] Gupta, A.K. (2004) Integrated approach for leprosy

elimination in India. Journal of Development and Social

Transformation, 1, 31-36.

[4] Gostin, L. and Hodge, J. (1999) Reforming state public

health law. Turning Point.

http://www.publichealthlaw.net/Resources/ResourcesPD

Fs/Alaska.pdf.

[5] Anderson, M., Elam, G., Gerver, S., Solarin, I., Fenton, K.

and Easterbrook, P. (2008) HI V/AIDS-related stigma and

discrimination: Accounts of HIV-positive Caribbean

people in the United Kingdom. Social Science & Medi-

cine, 67, 790-798.

doi:10.1016/j.socscimed.2008.05.003

[6] Citrin, D. (2006) Somali tuberculosis cultural profile.

http://ethnomed.org/clinical/tuberculosis/somali-tb-cultur

al-profile.

[7] Setbon, M., Raude, J. and Pottratz, D. (2008) Chikun-

gunya on Réunion Island: Social, environmental and be-

havioural factors in an epidemic context. Population, 63,

491-518. doi:10.3917/pope.803.0491

[8] Baral, S.C., Deepak, K.K. and Newell, J.N. (2007)

Causes of stigma and discrimination associated with tu-

berculosis in Nepal: A qualitative study. BMC Public

Health, 7, 211. doi:10.1186/1471-2458-7-211

[9] Herek, G.M., Mitnick, L., Burris, S., Chesney, M., De-

vine, P., Fullilove, M.T., Fullilove, R., Gunther, H.C.,

D. Gonzalez-Medina et al. / Health 3 (2011) 206-210

210

Levi, J., Michaels, S., Novick, A., Pryor, J., Snyder, M.

and Sweeney, T. (1998) Workshop report: AIDS and

stigma: A conceptual framework and research agenda.

AIDS Public Policy Journal, 13, 36-47.

[10] Zak, P. and Fakhar, A. (2006) Neuroactive hormones and

interpersonal trust: International evidence. Economics

and Human Biology, 4, 412-429.

doi:10.1016/j.ehb.2006.06.004

[11] Barro, R. (1991) Economic growth in a cross section of

countries. Quarterly Journal of Economics, 106, 407-443.

doi:10.2307/2937943

[12] World Health Organization: World Health Statistics. 2009:

50-69.

http://www.who.int/entity/whosis/whostat/EN_WHS09_

Full.pdf.

[13] Knack, S. and Keefer, P. (1997) Does social capital have

an economic payoff? A cross-country investigation.

Quarterly of Journal Economics, 112, 1251-1288

doi:10.1162/003355300555475

[14] Inglehart, R. et al. (2000) World values surveys and

European values surveys, 1981-1984, 1990-1993 and

1995-1997. ICPSR Study No. 2790, Inter-University

Consortium for Political and Social Research.

[15] Inglehart, R., Basáñez, M., Díez-Medrano, J., Halman, L.

and Luijkx, R. (2004) Human beliefs and values, a

cross-cultural sourcebook based on the 1999-2002 values

surveys. siglo XXI editores, S. A. de C.V., Mexico.

[16] Keefer, P. and Knack, S. (2002) Polarization, politics and

property rights: Links between inequality and growth.

Public Choice, 111, 127-154

doi:10.1023/A:1015168000336

[17] World Development Indicators 2009. Washington, D.C.:

The World Bank.