06 August 2014


'Let the Sunshine in: Why Countries Adopt Freedom of Information Acts' by Patrick Brown, Jerg Gutmann and Stefan Voigt comments
Over the last decade, some fifty countries around the world have adopted Freedom of Information Acts (FOIAs), which allow citizens to force the release of information from public offices. Here, we are interested in analyzing the determinants for adopting FOIAs. We apply a set of survival analysis techniques to identify the cultural, political and economic determinants that lead countries to adopt a FOIA. We further distinguish the adoption of constitutionally entrenched information rights from weaker forms of self-commitment to government transparency. We find that there are marked differences in the variables explaining the existence of FOIAs and their inclusion in the constitution.
Governance, more recently also referred to as quality of government, plays a prominent role in contemporary discussions of institutional reform and economic development. If one aims at designing institutions that ensure impartiality in the exercise of public authority, the transparency of government actions is a key ingredient. While a free media is a central instrument of civil society to provide transparency, governments themselves can contribute to the transparency of their actions – and particularly those of the administration – by promising citizens access to information. Freedom of information acts (FOIAs) guarantee this access in different shapes and to a varying degree. A FOIA regulates the right to access information that is held by public authorities. Individuals and groups covered by a FOIA have the right to request and receive information that is deemed public record. 
By the end of 2012, 91 countries the world over had adopted FOIAs. Countries in Europe and North America are at the forefront with an adoption rate above 80 percent, followed by a 70 percent adoption rate in Latin America. Yet, to this day, no country in the Pacific region has implemented such a law and only some have on the African continent. Looking beyond the mere introduction of laws and ordinances, we find that 57 countries worldwide have even entrenched freedom of information rules in their constitution. Again, over 50 percent of the countries in Latin America belong to this group, but no countries located in the Caribbean. The current distribution of FOIAs, either as simple laws or constitutional clauses, is illustrated in Figure 1.
While Sweden was the first country to adopt a FOIA in 1766 and remained the only country for over a century, FOIAs have spread rapidly over the past ten years with almost 50 countries adopting such laws between 2001 and 2010. xxxxx The rise of “good governance” occurred simultaneously with the rise of FOIAs. In good governance, both accountability and transparency are key. FOIAs are to increase the transparency of the administration and thereby improve the accountability of the entire executive. FOIAs can be interpreted as one means to achieve good governance and this paper would, hence, constitute a contribution to the governance literature. As we are interested not only in the introduction of FOIAs, but also in their possible entrenchment in the constitution, this paper also seeks to contribute to a recent literature on endogenous constitutions (Hayo and Voigt 2011, 2013). 
Listening to advocates of FOIAs, one might get the impression that these laws are some kind of panacea. Banisar (2006), e.g., argues that FOIAs are a prerequisite for informed democratic participation. Further, they would allow countries in transition to distance themselves more credibly from their recent past. By allowing individuals to protect their own rights and by curbing government mismanagement and corruption, they could also increase the level of trust that citizens have in their government. 
In one of the first studies concerned with the effects of FOIAs, Islam (2006) finds a positive correlation between two indicators for FOIAs on the one hand – namely their mere existence as well as the number of years that they have been in existence – and a set of governance indicators on the other hand. Her results seem to indicate that by introducing FOIAs, governments can promote good governance. Using instrumental variables regression, she finds that better governments do not necessarily promote more economic transparency and they are not more likely to adopt FOIAs. However, in the absence of suitable instrumental variables for FOIAs, the question of whether greater transparency promotes better governance cannot be answered conclusively. More recently, Costa (2013) finds that countries that have adopted a FOIA subsequently experience a decrease in the quality of governance and an increase in perceived corruption. Prima facie, the latter results appear more credible because they are based on a differences-in-differences approach, whereas Islam presents simple correlations based on a cross-section and is, hence, not able to establish any causality. 
These results are somewhat contradictory, but it seems that FOIAs can have important and far-reaching effects. This is why we are interested in the determinants that lead countries to introduce a FOIA. Taking into account that FOIAs can be adopted as an ordinary law (FOIALEG) or in a country’s constitution (FOIACON), we ask two questions: (1) What are the determinants for the adoption of FOIAs tout court? And (2) what are the determinants for the inclusion of FOIAs in a country’s constitution? Under the assumption that constitutions are more difficult to amend than introducing ordinary legislation, giving FOIAs constitutional status could be indicative of the importance attributed to the promises contained in FOIAs. 
Banisar (2006) also speculates on possible motives for the introduction of FOIAs. He suggests three reasons, namely (1) corruption and scandals, (2) international pressure, and (3) modernization and movement towards an information society. These reasons would reflect an increased demand for transparency from within civil society as well as from other countries and a reduction in the costs of processing information and thus of supplying transparency. 
Along the lines of the “varieties of capitalism” literature à la Hall and Soskice (2001), McClean (2010) argues that the difference between competitive vs. coordinated market economies is a crucial determinant of transparency: The more coordinated a country’s economy is, the less transparent it will likely be. In coordinated economies, the most powerful interest groups should be hostile to sharing their exclusive access to the government. McClean illustrates this political economy-argument and its limitations by drawing on the U.S. and the U.K. as examples of competitive economies and Germany as well as Sweden as examples of more coordinated economies. 
The only paper to date that systematically analyzes the determinants of FOIA- adoption in a large cross-section of countries is Berliner (2014). He argues that the adoption of FOIAs provides political and expressive benefits for political actors and conjectures that FOIA-passage is more likely in countries where political competition and uncertainty create an incentive to institutionalize transparency. As political actors are facing strong opposition and a highly competitive political environment, political benefits can supposedly be derived from access to information in the future. Hence, when the likelihood of re-election is low due to intensive political competition, the incentive to institutionalize transparency would be high. Expressive benefits could be derived from earning the recognition of international stakeholders. When both these benefits outweigh the substantial costs for political actors – namely an increased risk of exposure of corrupt activity and the loss of profits from having a monopoly on government information – then the likelihood of adoption is – according to Berliner – particularly high. However, one could make the argument regarding political competition in exactly the reverse direction: When the potential costs of losing an election are high due to an uncompetitive political system, a risk averse government will be tempted to insure itself against the loss of political control (analogous to Ginsburg’s (2003) argument regarding judicial review). Politicians might consequently introduce a FOIA even at the cost of giving up some of the rents from their time in office. 
Berliner (2014) uses an event history approach to model the timing of FOIA adoption across countries. He analyzes the post-Cold War period from 1990 to 2008 and confines the analysis to developing countries, arguing that government transparency was already high in the most developed countries before the formal adoption of FOIAs. Many autocratic regimes are removed from his sample. His empirical results seem to support the role of political and expressive benefits in explaining variation in the timing of FOIAs. Furthermore, Berliner (ibid.) shows that transnational civil society, democratic transitions and IMF conditionality have no systematic effect across countries. In an attempt to elucidate the determinants of FOIAs, Berliner’s pioneering contribution on FOIAs has put a magnifying glass on post-Cold War developing countries. However, his work has some limitations, including (1) limited geographic scope, (2) a limited time frame and, most importantly, (3) a questionable operationalization of some determinants, including his modelling of spatial and time dependence. We seek to improve upon some of these limitations and provide a more comprehensive analysis based upon survival analysis techniques applied to a sample of 136 countries.
The rest of the paper is organized as follows: In Section 2, we present our theoretical framework and develop a number of hypotheses. Section 3 presents the data and the empirical model. In Section 4, the results are discussed. Section 5 concludes and provides an outlook.