The new love triangle: quality of government, institutional trust and development

Recent weeks brought forward debates concerning governments and the quality and efficiency of their interventions and decision-making processes. While this discussion can be nuanced in many ways, my personal interest lays mainly with the basic understanding of the relation between the quality of the governmental act, the level of trust in international, national and sub-national institutions, and the GDP variation (perceived in this case as a unit of measure for development levels). I believe that the understanding of this relation is one of the keys in developing effective governance and policy-making processes.

Data sources and motivation

Developing a high-quality and accurate measurement of the relation between the quality of government, trust and GDP is a complicated matter as it involves correlating sets of qualitative and quantitative data, which are not always in sync in terms of meaning, therefore making the interpretation of the results debatable. My study represents a preliminary analysis meant to spark some (more) interest in further researching the relation between these three elements, as it might lead to interesting conclusions.

I based this short analysis on three data sets (secondary data). First, I needed an appropriate level of measurement for the governmental act and its effectiveness. A recent study regarding regional governance and quality of government (working paper available here) required the same. By combining the Worldwide Governance Indicators (WGI), provided by the World Bank (full description and data sets are available here), the authors were able to create a national level ranking system (see table below) based on the combined averages of four of the pillars provided by the WGI: Control of Corruption, Rule of law, Government Effectiveness, Voice & Accountability (Charron, Lapuente & Dijkstra, 2012, p.3-4).

Considering the regional level would have been too extreme for a blog post analysis due to the size of the data sets. As a result I decided to only use the national Quality of Government Index (QoG), as it compared well with the other data described below. I have to note that this QoG index is based upon indices for the year 2008.

The second data set was the “Public opinion in the European Union” survey published by Eurobarometer covering the spring of 2012. Among other issues the survey asks the interviewee what is the level of trust for his/her: national government, national parliament, regional and local authorities, and the EU (QA13.2, QA13.3, QA13.4, QA13.6). For the purpose of this analysis I have used the “Tend to trust” scores for each individual member state.

Even if the inquiry in itself is fairly clear (I would like to ask you a question about how much trust you have in certain institutions. For each of the following institutions, please tell me if you tend to trust it or tend not to trust it: […] ), I have to point out that in the case of the EU the results are questionable at some extent. This is because, “the European Union” in itself is not a single institution, but it is composed of several bodies, with different functions and roles.

Alas, the same survey actually points out that from the total interviewees, 9% have not heard of the European Parliament, 17% have not heard of the European Commission, 30% have not heard of the Council of the European Union, 16% have not heard of the European Central Bank, and 24% have not heard of the Court of Justice of the European Union (QA16.1 – QA.16.5).

The third data set represents the most recent GDP per capita in PPS for the year 2011, provided by Eurostat (full data set available here). I decided to use the 2011 GDP levels because the variation between 2008 (the year for which the QoG indices were used) and 2011 is relatively low (generally bellow 5%), the only exception being Greece with a 10% decrease. Because the analysis is constructed on the interpretation of the relation between data pairs (see below), I decided that the use of the 2011 is appropriate as the value can, at some extent, reflect the results of policy implementation in contrast with the QoG index measured for 2008, and at the same time motivate the trust levels reflected in the 2012 Eurobarometer survey. Even if causality might seem to run the train of though here, I have to emphasize that based on my research, causality cannot be implied in these circumstances, because of the lack of clear-cut proof. However, we can suspect it, and further analysis can confirm or refute it.

Putting it all together

The basic methodology of the analysis is fairly simple. I gathered the three data sets and combined them together in one table. Then I combined the columns in pairs of two and composed scatter-plots for each pair in an attempt to identify trends in the data. Additionally, because the ‘trust’ related data set is composed of three percentages a chart seemed the appropriate way to simultaneously visualize all three combined with the GDP values. Based on the scatter-plots’ distribution and the chart I was able to extract some interesting insights. (Country codes are listed here)

QoG vs Trust

Comparing the QoG index with the tendency to trust international, national and sub-national institutions resulted in four scatter-plots (see below) which if interpreted give a relatively consistent perspective. One major anomaly can be observed in the case of Romania and Bulgaria. Both have very low QoG indexes which positions them close to the Y Axis (i.e. vertical axis) in all four graphs.

It is clear that for Romania and Bulgaria the low level of trust in national and sub-national institutions is a reflection of the low quality of national governance, which cannot be countered by the higher level of trust that is associated with the EU institutions. Leaving this exception aside, it is clearly visible that a correlation between institutional trust and good governance is true in the case of the EU member states. The most evident, I believe, is the link between the tendency to trust the national parliaments and QoG. This can be assigned to the fact that usually parliament members are elected through democratic vote. In the case of national governments as well as regional and local authorities, these are also correlated in an ascending trend with QoG, while the trust in EU institutions seems not to have a very important relation with the QoG.

Quality of Government vs Trust in Institutions

In conclusion, we can state that it is possible that a higher level of trust in national and sub-national authorities can lead to a higher quality of governance at national levels.

Trust vs GDP

Building upon the trust levels emphasized by the Eurobarometer survey, using the chart below we can clearly observe the following. First, by comparison there is an overall difference between the trust in national institutions and sub-national authorities, the latter being in some cases two or even three times higher than the first (e.g. Czech Republic, United Kingdom, Latvia, Hungary). Second, by ordering the member states according to the trust in regional and local authorities and then by GDP, they can be divided in 3 categories. From left to right, according to the tendency to trust regional and local authorities: (1) 0% – 29%: IT, ES, EL; (2) 30% – 49%: IE, LT, RO, PT, BG, SI, PL, HU, MT, LV, UK, CZ, CY, SK; (3) 50% – 75%: EE, NL, FR, DE, FI, BE, SE, AT, DK, LU. (Click on the chart for full size)

Strangely enough the GDP variation between the 3 categories is not as obvious as I first suspected it would be. The first category is compose of three countries which all have a GDP of over 75% of the EU average, moreover Spain is close to 100%, while Italy has a GDP of 101%. The second category is the most varied in terms of GDP, the lowest GDP being of 45% (i.e. BG) while the highest reaches 127% (i.e. IE). Lastly, the third category has the highest GDP average, as only one member state in this category has a GDP below 100% (i.e. EE, 67%), while the rest span from 100% to 125%, with a peak of 274% in the case of LU. It has to be noted that the case of LU is an outlier in this analysis and because of the disproportionate difference in GDP and population, its case cannot be considered as a standard.

Trust in Institutions vs GDP

Another interesting observation is that in the case of the third category we can see a close relation between all the trust levels; in the case of all the member states the trust levels for national institutions span from 35% to 74%. Compared to the second category in which large differences in trust can be identified between national institutions and regional and local authorities, it can be stated that a direct link between institutional trust and GDP can be suspected. However, this chart would imply that in order to ensure a high GDP level a higher trust level in regional and local authorities (2) is not enough; as the member states in the third category show, concurrent trust in all institutional levels is more probable to create the premise for higher GDP levels.

As you may have notice I neglected the first category in the analysis above. The reason behind this (as in the case of LU) that there is a notable difference between the levels of the GDP and the levels of trust. Basically this means that with low levels of trust from the population, these countries’ national and sub-national institutions managed to implement policies that achieved economic growth. If we also consider the QoG index for these countries, and the recent macro-economic developments we can clearly observe that these three countries (i.e. IT, ES, EL), are the ones that, at the moment, suffer the most from the economic crisis. In conclusion, we can suspect an artificial inflation of the GDP.

In conclusion, a relation between the composite level of trust in national and sub-national institutions and the GDP does exist. Moreover, it is also clear that high levels of trust of only national or sub-national institutions are not enough in order to correlate these to a certain level of GDP.

QoG vs GDP

By verifying the relation between the QoG index and the GDP levels, a clear emphasis of the direct relation between good governance and GDP is evident. If we consider the four pillars of the QoG (see above), and their implicit effect on the perception of the individual and his/her trust level, it can then be confirmed that trust can be one of the elements that creates the premise for high GDP levels.

Quality of Government vs GDP


The most clear-cut conclusion that can be drawn from this analysis is that trust in the sub-national and national authorities is of utmost importance, and can be suspected as an important player in the GDP evolution. The link is also confirmed by the QoG index, which coincides with high (I use the term loosely here) trust levels, as well as with high GDP levels. This creates the opportunity to develop the inter-dependency between the three in a tool, aimed at explaining public policy impacts and effectiveness at national and sub-national levels.

While I believe the above analysis depicts a useful and accurate image of the inter-dependent relation I described, it is, however, limited in scope and depth. National GDP levels can sometime show distorted images, which ignore sub-national disparities in development levels, and why not, even trust disparities. Even though I do not believe that the end image might change in entirety, further research should extend this framework to sub-national levels, as it might bring forward some new scale related issues and differences not considered here.