The quality of governance increases confidence in politics and institutions

The majority of the world's population lived in a democracy until the 2000s, when the proportion fell sharply to just 29%[1]. What's more, the opinions of citizens in some countries are so polarised that even long-standing democratic regimes are at risk of paralysis. The situation is worrying.

While debate and controversy are the hallmarks of democracies, confidence in national governments has fallen significantly, to around 40% of citizens according to OECD surveys in 2021 and 2023. Fortunately, in Luxembourg confidence in the government has reached 55% (2021 and 2023), well above the OECD average.

At what level does mistrust become dangerous?

Democracies are weakened internally by extremist parties and externally by autocracies waging an ideological war against the West and its humanist values. What are the mainsprings of trust? How can we defend democracy and strengthen trust in institutions?

In economics and political science, the rational choice paradigm prevails: voters assess the policies pursued by those in power and place their trust in them on the basis of the results achieved or the expectations placed in them. This approach has been formalised in the framework for analysing trust in institutions developed by the OECD (Brezzi et al. 2020). The building blocks of the theoretical model are: integrity, responsiveness, reliability, openness, fairness and satisfaction with public services. They are measured empirically using a questionnaire developed by the OECD and applied in 18 countries in 2021.

This study verifies the relevance of the conceptual model of trust that served as a guide for the OECD survey on trust in institutions and politics. The confirmatory factor analysis highlights three empirically validated dimensions: trust in politics, trust in institutions and the quality of governance. Our model shows a high degree of fit to the data across all countries and within each country.

These three dimensions - trust in politics, trust in institutions and quality of governance - allow us to rank countries on the basis of their median trust and quality of governance scores. We find that Denmark is consistently at the top, with the highest average trust scores, while Colombia is at the bottom in these dimensions. Luxembourg achieves a very respectable score. Although this disparity suggests a complex interplay of historical, cultural and socio-economic factors, our analysis is limited to measurement issues. To explore the reasons why some countries fare better than others, it would be appropriate to use a regression technique at two levels: individual, on the one hand, and macroeconomic (economic, social performance, transparency) apprehended at the level of each country, on the other.

Confidence scores, constructed using confirmatory factor analysis, are used as endogenous variables in a regression analysis in which one of the main explanatory variables is the factor score on the quality of governance. The analysis shows that governance has a significant positive effect on confidence in politics and institutions, all other things being equal. This result is important because it highlights the positive effect of the qualitative dimension of the services provided by the state and its agencies to the general public. Trust in institutions and government increases with the integrity, reliability, responsiveness, fairness and satisfaction of the services provided to the public!

We also examine the effect of respondents' socio-economic characteristics on the two trust factors (institutions and politics) and on the quality of governance factor. The analysis shows that women tend to have less confidence in institutions and politics, as do people who earn less or have financial difficulties. Young people and people aged between 30 and 49 are more sceptical. In all countries, natives have less confidence in their country's politics than foreigners. The approach adopted here is a general one, synthesising the different dimensions of confidence and examining the determinants in a multivariate analysis. The results are broadly in line with the conclusions of the OECD report (‘Building Trust to Strengthen Democracy’, OECD, 2022).

In addition, exploiting the continuous nature of the trust factors, we use distributional regression analysis to study the robustness of the coefficients relating to socio-economic characteristics. Our results indicate that the influence of gender, income and social status on confidence varies along the confidence distribution. More specifically, gender (female) and income have a negative and significant effect at the lowest confidence levels, but become insignificant and even change sign at the highest confidence levels. This calls for a nuanced interpretation of the socio-economic determinants at different confidence levels.

The main objective of this paper is to evaluate the model for measuring trust and the quality of governance in 18 OECD member countries. The next step will be to validate the confirmatory factor model described in this study using the new OECD survey conducted in 2023. In addition, we will take into account economic and social characteristics at the level of each country. The new data will enable us to pursue studies on the role of official statistics in enabling citizens to evaluate public policies (Allegrezza, Joxhe and Langer, 2023).

[1] https://www.statista.com/statistics/1379594/people-world-distribution-regime-type/