Non-recognition is the symptom, not the cause

My PhD thesis – What is the effect of non-recognition? The external relations of de facto states in the post-Soviet space – is now online and can be downloaded freely.

Aware of the limited allure of a 300-pages pdf file, I decided to outline here some key outcomes (not necessarily the conclusions), as well as some additional thoughts.

1. Non-recognition is the symptom, not the cause

First, a one-paragraph summary of my conclusions. Once post-Soviet de facto states are conceptualised as small, dependent, and partly isolated jurisdictions, many seemingly unusual features of their external relations can be explained without reference to lack of recognition. Indeed, prevalent dynamics of external relations found in these territories, such as strong dependence on a patron, are compatible with those found in uncontested territories on both sides of the sovereignty divide. Most features of external relations described in different parts of my research can be related to underlying dynamics, rather than – strictly speaking – lack of recognition. Ultimately, non-recognition is a symptom rather than the cause of the complex environment that surrounds these entities and determines the boundaries of their external relations.

I developed this argument further in an article published on Ethnopolitics, where I introduce a comparison with nominally sovereign Pacific island states such as the Marshall Islands and Palau. Another, closer set of terms of comparison (that is not really discussed in the paper, and only briefly outlined in the thesis) it that of sub-state entities in the Russian federation. Along the key set of indicators I refer to in my research (migration, remittances, aid, and bureaucracy – i.e. a high share of civil servants), republics in the Northern Caucasus apparently present remarkably similarities with post-Soviet de facto states. The fact that а significant part of budget expenditure in Abkhazia and South Ossetia is determined jointly by the local authorities in Sukhumi and Tskhinvali, and the Russian Ministry for North Caucasus affairs is also telling.1

The comparison between post-Soviet de facto states and Pacific islands is one based on similarities among jurisdictions that are strikingly different in most respect. A structured comparison that is on the contrary focused on differences, aimed at analysing the political economy (or political competition, or decision-making processes) of post-Soviet de facto states and republics in Russia’s northern Caucasus may be particularly helpful in understanding what is peculiar about the functioning of these entities, and singling out other aspects that, after all, are heavily influenced by non-recognition (or perhaps, by not being fully incorporated into the patron). Thinking of potential new venues for research, this is certainly one.

2. How much does Russia spend there? Is it affordable?

In my dissertation, I approach the question “Where does the money come from?“, and I detail how most of the budget and the pension system of these jurisdictions is financed by their patron. In the relevant chapter, I include also a number of graphs that present data on how much money is involved, and puts these relatively high figures in context (I have posted some of the graphs on Twitter with the hashtag #deFactoStats, and I plan to release them in the form of a structured dataset.)

To cover for all direct expenses such as pensions payments  and budget assistance to Transnistria, Abkhazia, and South Ossetia, Russia currently spends significantly less than 1 billion USD a year. It was probably close to 1 billion USD in 2013 – excluding military expenditure – but due to the fall of the exchange rates as well as other priorities (such as Crimea and the war in Ukraine) it decreased substantially if calculated in USD in the following years. Is this a lot? For the recipients, it certainly is, both as a share of budget, and per capita (according to my estimates, South Ossetia would likely top the world ranking of jurisdictions by aid per capita in most years). But for a country with an economy such as Russia’s, not really. Even after the economic crisis and the fall of the exchange rate in late 2014, as of 2016, Russia’s GDP stood at about 1283 billion USD (Source: World Bank). As of 2016, Russia’s central budget expenditure stood at 232 billions (Source: World Bank).2 From these figures, it appears that less than 0,3% of Russia’s budget expenditure goes to sponsor Transnistria, Abkhazia, and South Ossetia. This is significant, but probably affordable in the long term as long as there is some political capital associated with these territories.

3. What kind of external assistance reaches these territories?

In the thesis, I detail some of the ways in which different international actors provide assistance for state and capacity building. Again, unsurprisingly, the patron is the most important source of assistance in developing local capacities. Indeed, even beyond direct financial transfers, authorities in these territories receive from their patron multi-faceted capacity-building assistance in quite literally all sectors of state activity. The pervasiveness of such interactions (and the ensuing alignment of practices) became increasingly apparent as I advanced with my research.

Non-patron assistance is focused on fewer sectors and to some extent is complementary to Russian assistance. The label of “confidence building” covers a significant part of it, even when assistance is not strictly conflict-related. I have argued in a separate post (“Confidence building, by any other name? Surpassing the triple bottleneck of assistance to de facto states“) that a different framing may bring positive externalities. Indeed, this is already happening to some extent, and I believe this is a trend that should be fully embraced by forward-looking international donors.

4. There’s a lot of evidence online, and often Google won’t help

In a recent article based on my PhD research, I have outlined some examples of “Quantitative Analysis of Web Content in Support of Qualitative Research“. I don’t want to dismiss the “quantitative” part, but I really think my main argument is about “structured analysis of web contents.” There’s plenty of evidence to be found in online sources, official websites, local media, etc.

Parsing these contents in a systematic way can be more effective, and often quicker than alternative approaches.3 A structured approach also makes it possible to subset and reduce even hundreds of thousands of news items or press releases to a manageable number that can be parsed manually to find additional data points or information.

I feel that many colleagues and area studies experts find the technicalities involved off-putting. This is understandable, but regrettable. My hope is that, in due time, an ecosystem will grow whereby it will be easier to conduct structured analysis of online contents, as well as to outline how we conduct our research in a way that can be easily accessed and directly scrutinised by fellow researchers. I think we all have to gain by increasing the transparency of what has by now become a considerable part of many research projects, yet remains hidden in most research outputs: looking up stuff online.

I suppose (I hope?) that some larger organisation will take up the effort to develop the necessary infrastructure. In the meantime, I have started developing a package for the R programming language that facilitates all of the above: it allows to transform online sources in textual datasets, and allows to share the results in an interface that can easily be used without requiring specific training. Here’s interactive example based on press releases issued by the Kremlin, here’s a post outlining a potential output on Russian media, and here’s another one focused on elections in Abkhazia; these examples are all based on word frequency, but again, I believe that the “systematic” rather than the “quantitative” component is what really matters.

As, quite clearly, software development is not my main occupation, this package is functional, but still rough. In due time, I will hopefully be able to add introduce new functionalities, and enable the creation of textual datasets from arbitrary websites directly from a visual interface that expects very little technical competence from the user.

5. Concluding thoughts

Unsurprisingly, I am not completely satisfied with my dissertation, and I already look rather critically at some of its parts. With a smile, I blame all shortcomings to my becoming a parent in the early stages of this research. More seriously, I believe that becoming a parent has fundamentally changed my approach to research. Among other things, it led to less fieldwork, probably less readings, and of course less sleeping, but it also gave me unusually lengthy stretches of time for thinking freely, which I think had a positive impact on my research.

[Together with this post, I am publishing online a brief excerpt from my thesis where I explain some of the reasons that brought me to study post-Soviet de facto states, and partly determined some of the methodological choices of my PhD research.]

Again, the thesis is available in its full glory, with pictures and plenty of coloured graphs from this link. Overall, I do believe it represents an original contribution to the literature on post-Soviet de facto states, theoretically, empirically, and methodologically.

I must take this chance to thank Donnacha Ó Beacháin for accompanying me through this journey as my supervisor, as a fellow researcher with insatiable curiosity, and as a friend. Gerard Toal, for pushing me to think further and harder about many aspects of my research as my external examiner. John Doyle for ensuring a positive and friendly research environment through my stay at Dublin City University. And so many other fellow researchers I have met through these years.

The PhD journey is never easy. Which is why it is not without some genuine surprise that I find myself typing the final words of this post: it’s been a joy.

  1. It is worth highlighting that joint negotiation of budget expenditure is common in nominally sovereign small dependent jurisdictions, as well as in sub-state entities with meaningful self-government. []
  2. If the exchange rate remains stable at 2017 levels, the planned budget expenditure for 2018 would however be significantly higher, or about 283 billion USD []
  3. Of course, it needs then to be complemented by other approaches, most prominently, interviews and fieldwork. []