What the Public Wants to Know and What it Gets: Evidence from Quantitative Text Analysis of Freedom of Information Requests to the UK Government

From Bentham to Sen, philosophers, and political economists have agreed that access to government information is critical for democracy and development. In the past two decades, both democratic and authoritarian governments have adopted freedom of information acts (FOIAs) that enable citizens to ask for specific information of their interest that the government holds. Many scholars studied why governments pass the legislation that subjects them to greater scrutiny (Berliner, 2012, 2014; Berliner & Erlich, 2015; McClean, 2011; Michener, 2010, 2011; Scrollini, 2015). However, studies exploring what citizens seek to know once they have the legal right to do so are scarce. Understanding the public demand for government information and how the government responds to this demand is essential for assessing if FOIA is implemented as intended and fulfils its democratic oversight function. I study the public demand for the UK central government’s information and the government’s responsiveness to this demand using the data from Whatdotheyknow.com, an online participation platform for submitting information requests. Using unsupervised learning techniques, in particular structural topic modelling, I analysed the texts of all information requests sent to the central government bodies through this platform since its start in 2008 to the end of 2017, over 37 000 requests in total. I explored the proportion of the topics within these requests and their outcomes. I find that FOIA in the UK is very contemporary and reflects the major political events or policy changes that make newspapers’ headlines. Also, in theory, information as a public good assumes equality of the right of access to it. However, scholars demonstrated that the opportunities to access the information using FOIA are not equally distributed (Neuman, 2016). In practice, people with underprivileged background lack technical means to access the information. They also lack skills to use it autonomously as they are less likely to be exposed to innovations and have a support network (Dimaggio, Hargittai, Celeste, & Shafer, 2004; Hargittai, 2003; van Dijk, 2005). Information inequality is also present in the outcomes. Several scholars have already demonstrated that governments do not treat those who seek the information in the same way (Butler & Broockman, 2011; Michener & Rodrigues, 2018; Michener, Velasco, Contreras, & Rodrigues, 2019; A. Roberts, 2002; A. S. Roberts, 2005). While I do not investigate the effects of demographic characteristics on government responsiveness in this study, I explore the effects of the nature of requested information. I find that the government treats differently the requests asking for politically sensitive information, disclosure of which might have potentially damaging effects for the government (e.g. government spending, information about procurement, contracts or foreign relations). These requests are more likely to be associated with the negative outcome, i.e. withheld information than requests enquiring about noncontroversial issues. These findings have implications for fairness in the provision of information.