A central motive behind the creation of the QoG Institute was a suspicion (or hypothesis) that in all societies, the quality of government institutions is of the outmost importance for the well-being of its citizens. Instead of seeing political institutions as mere reflections of the underlying social and economic forces, inspired by development in various types of institutional theory, we wanted to investigate the opposite idea, namely that economic and social development in a specific society is caused by the quality of its government institutions.
When the QoG institute was started in 2004, our ambition was to address three interrelated problems - one theoretical, one methodological and one empirical. The theoretical problem we identified came in two forms. The first was that there was not a generally established definition of how "good governance" or "quality of government" should be defined. Without a proper definition, QoG could not be operationalized and measured. The other had to do with our limited understanding of how and under what circumstances high quality political institutions can be created. The problem as it was stated in the literature was the following: From an interest-based perspective, every political and economic organization/actor would prefer to have political institutions (regulations and laws) that work in the favor their special interests. Such institutions, however, are per definition partial, or discriminatory, or outright corrupt. Thus, from an interest-based understanding of politics, there was no solution to the problem of how high quality government institutions could be created although this had happened. There was thus a lack a theory of how QoG could be create.
In part of the literature, this is known as the problem of "credible commitments" Nobel Laureate Douglass C. North expressed this as that universal institutions and administrative behavior are precluded within a strict utility maximizing model. If political leaders successfully shape a state that is administratively strong enough to protect the rights of individuals from corruption and the abuse of power, they will also have the capacity to an administrative machine that can violate those rights. Elinor Ostrom, Nobel Laureate in 2009, defined this problem as a collective action problem of the second order: the achievement of social norms of trust and confidence, without which impartial institutions cannot be created, is in itself a problem of collective action.
In developing countries, there are many cases where strong political leaders with a legacy from the struggle for national liberation have been unable to establish reliable and competent government institutions. In the Eastern European countries, the research showed a mixed picture. However, the literature did not only show that the genesis of universal and impartial institutions was hard to explain. Influential political theorists such as Russell Hardin presented a strong argument that even when impartial institutions had been established, we should expect them to be weak because from a rationalist perspective, few agents would have an in defending them.
The second problem we identified was methodological. Much of the research on changes of government institutions that had been done was based on single case studies of different countries. This research did give valuable information, but as is well known, it is hard to generalize from such studies. In order to make generalizations possible and test hypothesis, we wanted to move this research in two directions. The first was theoretically focused comparative and historical case studies, the second were large-n statistical studies.
The third problem we identified was empirical. Several different sets of data intended to measure QoG had been made available. Among those were, for example, data on levels of corruption, on systems of legal protection, on freedom of information, on the quality of electoral systems, on the constitutional structure, and survey based data on the perceptions people have in different countries of their government institutions.
We identified a number of major problems with the available data that we wanted to address. One was validity and reliability. There were stark differences in the quality of the available data that we wanted to address. Another problem that we identified was that aggregate and individual level data sets were usually kept separate. A third problem was that the data was scattered, literally, around the world and thus difficult to get access to, not least for time-constraint graduate students. A fourth problem was that data on quality of government were, for the most part, kept separate from two other theoretically relevant types of data. The first type we called “What You Get” data. Simply put, if a country has high QoG, what do it get in increased human well-being in areas such a population health, economic prosperity and literacy to name a few. In addition to such “hard” measures, we also wanted to relate measures of QoG to more “soft” data such as social trust, feelings of “happiness” and subjective well-being. A second type of data we wanted to assemble was the “how to get it” data. In this set, we wanted to include all possible variables that could explain the huge variation between countries in their QoG.
We came to understand that the disorganization of data was a real problem for the advancement of research within this area. This problem had become an importance entrance barrier for people who want to do studies in this field, including of course many PhD-students. Consequently, in addition to our theoretical focus on how to define and understand variations in QoG, an important part was to create a world-leading and freely accessible database where the quality of the data was highlighted and where it was made ready for easy use.
We emphasized that the datasets we wanted create would not only be valuable for those who wanted to do quantitative research. It would be of equal importance for students who wanted to do in-depth case studies of one or a few countries. For this type of research, our data set would be used to identify interesting cases/countries from different theoretical angels. These could be either so called "outliers", i.e., cases/countries that deviate from what was expected from known theories. Or it could be used to identify what is known as "most different" or "most similar" research designs in the comparative case methodology.