Comment on “How Inequality Affects Trust in Institutions: Evidence from Indonesia”

Suryahadi et al. (2022) is an ambitious and innovative paper. Highlighting the importance of trust as a key ingredient in the process of economic and political development, the authors examine the impact of inequality on trust, and by extension institutions, at the village and district levels in Indonesia.

The paper has several distinctive features. First, the authors carefully disaggregate the concept of institutions, into economic, social, and political (“the state”) dimensions, plausibly conjecturing that each of these may have different behavioral relationships to the key variables of interest. Second, the analysis is also geographically disaggregated. This adds richness to the study given that Indonesia is the world's largest archipelagic state. Third, they introduce an intermediate variable in the analysis, education, finding that more highly educated individuals' trust in the political and state institutions is more sensitive to higher inequality.

The authors' main conclusions are reassuring. In general, the levels of trust are high (Suryahadi et al.'s figures 2 and 3), in some cases arguably higher than might have been expected, and mostly rising. Not surprisingly, trust is particularly high for social ties. Evidently, the village respondents are least trusting only toward “strangers.” Financial institutions enjoy high trust; one might surmise that the absence of any major bank crashes in Indonesia this century might contribute to this finding. Trust in governments and the civil service is very high, approaching 80% in 2018. This is perhaps a little unexpected given the endless “coffee shop” discussions of corruption. Nevertheless, in discussing their figure 5 the authors add an important qualifier, that “lower trust does not pertain to political institutions like elections or the parliament, but more on state apparatus.”

The exceptions to the conclusion of high trust include that across religious communities, not a major surprise in view of the country's occasional religious tensions, and some decline in the press and media, which is a global phenomenon in this era of proliferating “fake news.”

The finding that trust is comparatively high is also of interest given Indonesia's inequality outcomes. Historically expenditure inequality was moderately low, but it has risen significantly for much of this century. Not surprisingly, therefore, the authors conclude that keeping inequality “in check” is important for healthy institutional development.

I have several comments on this fine paper, which might be explored in future work on the subject. First, it would be interesting to set out some analytical “priors,” of what one might hypothesize to be the likely relationships. For example, Indonesia has had episodes of quite serious conflict over the past 50 years, including ongoing unrest in the two Papua provinces (which presumably were not in the survey). But they have generally been contained, and in addition the level of reported criminality is quite low. In passing, both these variables, conflict and criminality, could be introduced as intermediate variables in the analysis.

Moreover, presumably the fact that Indonesia has successfully conducted five national elections in the democratic era, that is, from 1999 onward, with mostly credible processes and outcomes, and similarly so (on most occasions) for sub-national elections, also provides prima facie support for the high recorded trust. And even if people are unhappy with the leadership at various tiers of government they know they can vote them out at the next election (as they do quite often).

It might even be inferred that the high trust levels were a factor in Indonesia navigating the Covid crisis without major stress on its political and social fabrics. In this context, the authors document that the Covid pandemic increased inequality in Indonesia, especially for workers in the informal sector. But overall the effects were relatively mild and, importantly, about two-thirds of the negative effects of Covid were mitigated by the government's social protection programs. Connecting to the paper's central topic, they might want to emphasize that this significant achievement has presumably been a trust-enhancing outcome, as has the almost continuous reduction in head count poverty since the 1970s, apart from during the Asian financial crisis.

A second set of questions relate to the World Value Survey (WVS) data, which are employed to generate the paper's most original empirical contribution. The Indonesian WVS data were obtained from interviews with 3200 adults, who were geographically dispersed and also identified by education, gender, age, and location (urban/rural). For those (like this reviewer) who have not used the WVS data, some brief commentary on its origins and construction would be helpful. For example, how were the data collected (phone, internet, face-to-face interviews), who were the interviewers, and how well do these potentially value-laden concepts travel across international boundaries, particularly from developed to developing countries?

Third, given Indonesia's great subnational diversity, it would be interesting to explore whether there are significant spatial variations in the relationships. For example, one might hypothesize that the relationship between trust and inequality would differ between densely-settled, industrialized Java districts and those in lightly-settled, resource-rich, remote regions, where personal connections are assumed to be more important. But the sample size is presumably not large enough to undertake such analysis, and in any case the authors do employ island dummies.