The rise of international parliamentary institutions: Purpose and legitimation

For recent descriptive analyses, see, e.g., Cofelice (2012), Costa et al. (2013) and Kissling (2014). For an exceptional theory-oriented and comparative analysis, see Lenz (2013).

It could be argued that not only delegation but also pooling—the use of majority voting in IOs—encourages the creation of IPIs. However, empirical studies show that pooling is not independent of organizational purpose but, instead, predominantly a property of task-specific IOs. Indeed, it is seen as caused by organizational purpose. See Lenz et al. (2015). General-purpose IOs are more likely to deal with policies that have distributional implications and their open-ended, incomplete contracting may raise issues of national sovereignty and autonomy. Governments therefore want to keep their national veto. In addition, general-purpose IOs tend to have a smaller membership than task-specific IOs and are therefore less in need of pooling to overcome decision-making blockages. For these reasons, a correlation between IPI creation and pooling is likely, yet spurious: Member states of general-purpose IOs promote IPIs at the same time as they block or limit pooling. Our data features hardly any general-purpose IO that also engages in significant pooling, meaning that the real world of IOs does not afford the information necessary to disentangle the effects of pooling and organizational purpose.

We acknowledge the possible objection that governments might initiate treaty reforms with the express purpose of creating a parliamentary body. By selecting observations based on treaty reform, we might thus overstate the probability of IPI creation. Empirically, however, only about a third of the cases in our analysis feature the creation of a new IPI or maintenance of a pre-existing one.

The predicted probabilities are estimated based on 1,000 random draws from the estimated parameter distribution of the Model in Table 1. For each draw, we estimate the average predicted probabilities across the observations in our data for different values of our variables of interest. In Fig. 2, we show variability of the difference in predicted probabilities between task-specific and general-purpose IOs across the 1,000 simulations, discarding the top and bottom 2.5 percent (this can be interpreted as a 95 percent confidence interval). Figure adapted from Wucherpfenning et al. 2016.

It is worth noting that Fig. 2 does not show an interaction effect. That the difference between task-specific and general-purpose organizations widens over levels of democracy has to do with the non-linear transformation of the predictions of the logit model (i.e., log odds) into probabilities (see Berry et al. 2010).

Alternatively, nearest neighbor matching yields very similar results.

While ECOWAS was not exactly new to security declarations – in 1976 member states signed the Treaty on Non-Recourse to Aggression, in 1978 they signed the Protocol on Nonaggression, and in 1981 ECOWAS released a Protocol on Mutual Assistance on Defense (which was, however, not implemented) – ECOMOG was the first time ECOWAS was engaged in crisis management and had an active role in security issues. The significance of ECOMOG can also be glimpsed by the fact that it later served important roles in the Sierra Leone conflict, Guinea-Bissau, Liberia and Cote d’Ivoire.

The Mediation and Security Council marking the shift from ad hoc conflict resolution procedures to a more permanent structure was added only in 1997.

We are grateful to Densua Mumford for sharing this document with us.

New Zealand’s proposal most likely came at the behest of Mike Moore, a Member of Parliament in New Zealand at the time, who had written a book in 1982 advocating for a Pacific Parliament. In his book, Moore maintains that establishing a regional parliament would be beneficial both for PIF as well as the South Pacific Commission, as this would allow them to “rationalize their functions and roles” and help the two IOs become more efficient by dealing with details and administrative issues (Moore 1982: 41-42). In a sense, Moore did not want a parliament attached to PIF, but a regional parliament that would take over the functions of both PIF and the South Pacific Commission, in the hope that this would lead to further regional integration.

Personal interview with expert on Pacific region, 7 February 2017 (b).

Personal interview with individual affiliated with PIF, 7 February 2017.

Personal interview with expert on Pacific region, 7 February 2017 (b), personal interview with individual affiliated with PIF, 8 February 2017, personal interview with individual affiliated with civil society in the Pacific, 15 February 2017.

Similarly, Anderson et al. (2019) find that the procedural quality of international authority, rather than its strength, affects legitimacy perceptions.