State-building and the politics of scale in Solomon Islands

(image: Tobias Haque)
(image: Tobias Haque)
Written by Tobias Haque

In their new book International Intervention and Local PoliticsShahar Hameiri, Caroline Hughes, and Fabio Scarpello have made a useful contribution to current discussions regarding the successes and failures of state-building efforts in Solomon Islands. Application of new theoretical tools from the political geography literature is productive and thought-provoking. But their argument is weakened by strained interpretation of their own case study evidence and some internal inconsistencies in the use of core concepts.

Using politics of scale to explain state-building outcomes in Solomon Islands

Hameiri, Hughes, and Scarpello (henceforth HHS) view the ‘politics of scale’ as a useful framework for understanding the uneven outcomes of international state-building interventions, including in the Solomon Islands.[1] Drawing from the political geography literature, they define scales as “vertically differentiated, hierarchised social, political and economic spaces, each denoting the arena and moment, both discursively and materially, where socio-spatial power relations are contested and compromises are negotiated and regulated”. Examples of scales would include “political tiers within a state, such as a village or the nation, or… bio-regions, trans-governmental networks or ‘the global’’’. Different scales “afford different configurations of actors power, resources, and political opportunity structures”. Actors therefore face incentives to ‘rescale’ contests – moving contests upwards and downwards between scales – to a level at which their hands are strongest.

According to HHS, Solomon Islands’ political history (including the Tensions) has reflected structural contests between national and local actors, through which national elites have sought to build and sustain power by capturing rents and channelling them downwards through discretionary patronage networks to both undermine opposition from below, and mobilize local-level support. Since the RAMSI intervention, and reflecting this dynamic, national elites have worked to “maintain the primacy of the national scale” in opposition to both local actors (who seek to rescale contests downwards to the local level where local voices will have greater influence) and international actors (who seek to rescale governance upwards to an internationalized scale so that local and national institutions “become not so much responsive to local demands, which are often viewed as pernicious or dysfunctional, but to international governance agendas, geared towards meeting international targets and aspirations”). The rescaling objectives of national elites and state-builders therefore variously align and misalign, depending on whether donors’ upwards rescaling efforts strengthen or weaken the national scale.

Using politics of scale to explain public administration reform outcomes

The core contention of HHS’s Solomon Islands study is that contestation over scale explains the varying performance of public administration reform efforts in RAMSI-era Solomon Islands. According to HHS, this theoretical framing adds value over conventional political-economy analyses, which – in emphasizing incompatibilities between the goals of interveners and domestic political elites – predicts uniform failure rather than the observed uneven outcomes.

Certain public administration reform efforts under RAMSI provided important resources and tools through which national elites could rescale contests to the national scale. Such reforms were supported by national elites and generally succeeded. Other donor-driven reforms, however, failed because they threatened national elites by seeking to rescale institutions upwards to become accountable to global development and economic management agendas.

HHS argue that:

  • Improvements in tax collection reflected the interest of national elites in mobilizing resources for redistribution through national-level public sector patronage networks, while failure of reforms in the customs department reflected national elites’ dependence on the support of logging interests (which would have been threatened by improved taxation of log exports).
  • Failure to regulate or stem the growth of Constituency Development Funds (CDFs) reflected national elites’ resistance to any measures that would constrain discretion over their use of public resources to build and sustain national-level patronage relationships.[2]
  • Failure to implement decentralization through constitutional reform reflected alignment of interests around strengthening the national scale between national elites (who perceived decentralization as a threat to their direct control over patronage networks) and donors (whose global governance agendas would have been threatened by strengthened local-level accountability).

A better explanatory framework?

The ‘politics of scale’ framing is innovative and thought-provoking. But I question the claim that it provides greater predictive power than standard political economy analyses. That national elites selectively adopt and undermine donor-driven reforms in accordance with their own interests is clearly valid. However, this conclusion does not require recourse to a ‘politics of scale’ framing and has been well-established in broader literature. The argument that international intervention involves attempts to make local institutions accountable to international good governance and development agendas through insulation from local politics also rings true, but – again – is far from novel.

More importantly, a lack of conceptual clarity as to what precisely constitutes the ‘interests’ of national elites and how alignment with those interests should be assessed injects important weaknesses into both the internal consistency and predictive power of the framework.

Contradictions and inconsistencies

The failure to tax logging activities, despite sustained programmatic support for improved logging governance from donors, is cited as evidence that such reforms cut against the ‘rescaling’ interests of national political elites. But national elites’ interests would surely have been better served by capturing a greater share of logging rents and distributing them through public sector patronage networks, rather than seeing them absorbed by Malaysian logging firms and a narrow group of provincial landowning elites. Loggers – operating at both a local and international scale and deliberately evading national scale regulation – seem to have more power here than national elites. Neither the national-level rescaling objectives of national elites nor international rescaling objectives of international donors were served by the failure of logging reforms. Through destabilizing political regimes and adroitly manipulating community-level governance mechanisms via the distribution of private sector rents, logging interests have successfully undermined both national and local level elites. It is difficult to square this reality against a theory that predicts reform success when donor and national elite incentives are aligned.

Analysis of public expenditure patterns is similarly flawed. There is no doubt that expansion of Constituency Development Funds was aligned with national elites’ goals of maintaining discretionary control over public resources to mobilize political support through distribution of patronage. But CDFs cannot be the whole story.

Sound macro-fiscal management was one of the unambiguous successes of the RAMSI intervention. Over the RAMSI period, institutions were reformed to deliver sustained surpluses and dramatic reductions in public debt. Government avoided central bank financing, which is frequently observed in many low-income countries, supporting macroeconomic stability. These outcomes were achieved through the international advisor presence in the Ministry of Finance and under an IMF program. If the ‘national level rescaling’ objective of national elites was to mobilize the greatest possible resource for distribution through patronage channels, why were such funds not mobilized through central bank borrowing and accumulation of public debt in contravention of the desires of interveners? The alignment of interests between national elites and interveners seeking to impose Washington Consensus economic policy norms is not clear through the ‘politics of scale’ framing.

The composition of public expenditure presents additional challenges to the predictive power and consistency of the argument. Government expenditure on basic public services (including health and education) grew exponentially over the RAMSI period, with such expenditures consistently accounting for a large share of the annual budget, and a much larger share than CDFs. Public expenditure, overall, seems to have been better aligned with international development goals than the sustainment of national patronage networks. Perhaps anticipating this argument, HHS suggest that public expenditure controls in Solomon Islands were so weak that all public resources allocated through the budget could be misused and diverted in the interests of national elites. But most international indicators suggest that Solomon Islands has both a relatively robust public expenditure management system and moderate levels of corruption relative to countries at similar income levels. Further, if we are to accept the argument that all public expenditure provides patronage channels for national elites, then why did national elites go to the trouble of establishing and resourcing CDFs in the first place? The establishment of CDFs cannot provide evidence that national elites resisted internationalizing rescaling in order to retain patronage channels if such patronage channels could have been maintained through all categories of public spending.

Finally, HHS argue that donors were consistently opposed to local-level accountability arrangements (such as would be served through the constitutional review process) because of their interest in rendering public institutions accountable to international development goals and agendas rather than domestic demands. But some of the largest donor programs during the RAMSI period (the World Bank Rural Development Program and the UNCDF Provincial Government Strengthening Program) explicitly sought to strengthen and distribute resources through local-level accountability mechanisms, including local government. While both projects arguably sought to bypass perceived pathologies of local political economy dynamics in the service of international development agendas, the rescaling strategies bypassed a national government that was perceived as ineffective and lacking sufficient downwards accountability. It is difficult to see why these projects should be viewed as serving donors’ ‘internationalizing’ agendas while broader decentralization through the constitutional review process should not.


Overall, HHS provide an innovative and thought-provoking theoretical framework. Their basic point – that alignment and misalignment of national elites’ incentives with the goals of international interveners was an important factor in determining the outcomes of specific reforms – is well made. But actual outcomes are only sometimes consistent with the predictions of the framework. Because the framework excludes important actors and their analysis omits the causal processes that led to observed outcomes, their argument ultimately proves difficult to specify in terms that would have any predictive power. Further, ‘alignment’ with the rescaling interests of a monolithic set of ‘national elites’ is simply not sufficient to explain reform outcomes in a context where all decisions are contested by multiple powerful actors tied into complex patronage networks, including in the private sector.

Tobias Haque is currently a PhD student at the State, Society and Governance in Melanesia Program at ANU, where his research focuses on the interaction between economic and political development under state-building interventions.

Shahar Hameiri and Fabio Scarpello presented their book International Interventions and Local Politics as part of a panel discussion at the Crawford School on 24 August 2017; listen to the podcast here


[1] Their Solomon Islands case study is the focus of my review, but the book includes additional case studies from Cambodia and Aceh. 

[2] Constituency Development Funds are discretionary funds provided for the direct use of Members of Parliament through the development budget. They are not subject to the same controls and accountability mechanisms as general government spending.

Tobias Haque

Tobias Haque worked on economic development in fragile states for the World Bank between 2009-2017, including two years living in Honiara and – more recently – serving as the Senior Country Economist for Afghanistan. He is currently a PhD student at the State, Society and Governance in Melanesia Programme at ANU, where his research focuses on the interaction between economic and political development under state-building interventions. He currently lives in Washington DC where he works in the Development Policy and Finance team at the Bill & Melinda Gates Foundation.

He holds Masters degrees in finance (from the University of London) and sociology (from the University of Auckland).


  • Thanks Shahar

    As you say, an interesting debate and I can easily get behind ‘politics of scale’ providing an interesting and new perspective. I just don’t really see how it ‘fully accounts’ for variability in outcomes – especially not in a way that is substantially more useful than previous approaches.

    I’m interested in your statement: “For us, actors are not ‘scales’; actors make scales and associated institutions…”

    A very quick ‘find’ in Acrobat turns out 21 instances in which the term ‘national elites’ is used in the book, to describe a group of actors and the incentives or imperatives influencing that group. “National elites” even gets an index entry! Doesn’t this indicate that actors and scales are being used interchangeably, at least in shorthand? You seem to acknowledge this yourself for Solomon Islands when you say “the overwhelming reliance of all national politicians on a system that centralises rents from extractive activities within Honiara and then redistributes some resources back to the periphery to win elections, this elite’s class interests are remarkably cohesive.” I guess my claim is that these interests are not at all as cohesive as you state and that you lose more than you gain when bundling elites using a scale framing.

    I’m also interested in the claim that you (as above) don’t intend to present a ‘model’. In your text you claim to be seeking to rectify a failing in existing literature that ‘agency of those targeted by intervention had previously been relatively neglected’. Through doing so you seek to ‘explain diverse intervention outcomes in given contexts’. I wonder how you can do these things without a model?

    More specifically, I wonder what exactly can be explained through reference to ‘politics of scale’ and ‘scalar strategies’ that cannot be explained by a more traditional political economy perspective (even one that resorts to simplifying public choice assumptions)? Such analysis would equally point to the need for national MPs to raise discretionary resources for distribution through patronage networks (this is a common finding across many countries via the neo-patrimonialism literature).

    Great discussion.


  • I’ve really enjoyed this debate so far. I think there’s great value in what Tobias says in both the original piece and in the comments, though I stand by the utility of the framework for analysing the uneven outcomes of intervention. I think that some of Tobias’ misconception of the framework’s utility stems from a misunderstanding regarding how we approach scale, and subsequently its role in political analysis. This is clear from this comment above:

    ‘Much more generally, I’m just not convinced that talking in such monolithic terms as “national”, “local” or “international” is very helpful. Most recent work on state-building emphasizes that ‘national elites’ have divergent and competing objectives. Is a Solomon Island MP a ‘national’ actor? And what does that mean in tangible terms?’

    If there’s one aspect distinguishing our approach from hybridity, which does treat actors somewhat monolithically as ‘national’, ‘international’, ‘local’ and so on, and makes assumptions about how they behave subsequently, is that we don’t focus on the scalar position of the actors themselves, but on how they construct, reshape, and contest scales and associated institutions. Hence a village chief could be instrumental in constructing a ‘national scale’, while an international donor could promote the construction of a ‘customary’ local scale. For us, actors are not ‘scales’; actors make scales and associated institutions, but not under conditions they completely control. As such, scales are anything but monolithic in our approach – they may change over time, being mutually constitutive with power relations in society. To know what social groups are dominant and thus better able to produce scales that support their interests, for us, requires careful historical-sociological analysis, which we undertake in all of our case study chapters, including the Solomon Islands one. What the ‘national scale’ is in the Solomon Islands and what particular groups dominate it and its associated institutions has not been static, but is the dynamic product of historical processes of the kind we describe in the early part of the chapter.

    An international intervention invariably steps into that kind of context, shaped by uneven power relations in society and their manifestation in scalar hierarchies. What makes the politics of scale an important dynamic in the context of intervention is that international actors come in and seek to change how power and resources are produced and distributed by attempting to ‘rescale’ key institutions (usually of the national government, but often lower level governments or even informal, customary ones) so that they’re disconnected from the political and popular pressures that have hitherto shaped them. This invites, as we say, a response from recipients that (a) is also often framed in terms of what the most appropriate scale and institutions are that key issues should be governed through; and (b) adopts or rejects particular aspects of the intervention also because of how they are seen to affect pre-existing scalar hierarchies and the social power relations embedded in these.

    Crucially, we don’t see Solomon Islands’ national politicians as base rent-seekers who are always striving to maximise pecuniary gains for themselves. This seems to be the test Tobias has in mind to prove our framework is sound: any evidence for the use of funds for anything but patronage is used against us. If anything, we set our argument against ‘public choice’ and similar economistic approaches that see all politics as nothing more than market calculations for benefit/cost. Some of Solomon Islands’ politicians I’m sure would prefer to do things differently, but struggle against the structural constraints they face. We see them as political actors operating in a socio-political context, which has developed over decades, whereby for them to be elected, for governments to form and for the national scale not to collapse under the constant pressures for devolution, even dissolution, they must adopt scalar strategies that promote sufficient access to discretionary funds for local redistribution. Granted, there are still important differences between more personally decent and venal politicians, which we didn’t discuss much in the chapter (you can’t include everything…). For instance, we know that some MPs use CDFs relatively well to genuinely promote local development; others do not. Some may be more concerned about deploying budgets towards real improvements in services (as long as this doesn’t undermine the scalar hierarchy discussed above); others do not. Some are happier to fund provincial governments (again as long as they remain dominated from the centre); others do not. And so on.

    In summary, we’re not offering a politics of scale ‘model’ here. We’re offering an approach that explains important aspects of international intervention and its outcomes by drawing attention to the scalar aspect of social and political conflict and power relations, but which requires careful context-specific historical-sociological analysis to operationalise. As we make clear in the book, the approach explicitly links up with, and seeks to add to, existing political economy approaches in this area. It hasn’t emerged in a theoretical vacuum. It does, however, we think, deal with some of the blindsides of these approaches.

  • Hi Tobias and Terence. It was an interesting challenge for me to read something so technical like that piece and then to try and relate to what I know on the ground. Still pondering but thank you both for the valuable insights and robust discussions.

    I am sorry if I miss the point with this piece Tobias, but I then wondered if this review into the types of expenditure more (prone?) to patronage considered the scholarships directly managed by MPs.

    As well as the infrastructure sector (not only the shipping grants but also the build contracts).

    Both not at-first-glance CDF type expenditure lines but with parallel substantial injections through grants/pooled funding or direct budget support, this could be more about more money freed up to divert elsewhere?

    Again, sorry if I am missing the point.

    • Hi Atenasi,

      Thank you for your comment.

      You make a really good point. I imagine Tobias would agree with you. I certainly do. It’s hard in Solomons to accurately quantify the amount of funding that is directly available for patronage spending by MPs because it comes in so many different forms including — as you mention — scholarships, and shipping grants and the like.

      I think in your final paragraph you’re suggesting that aid money, in effect, frees up SIG revenue for patronage spending?

      This, I think, is more complicated. The argument you make is often made by economists (who talk about it as a form of fungibility). It’s very plausible in theory and must be true to an extent in Solomons. At the same time, however, donors do push against the money their spending frees up simply being used in unproductive ways by the government. As a result, fungibility is probably less than it might initially seem.

      I think one of the strengths of the book (at least based on what I’ve read thus far) is that the concept of scale makes explicit the fact that any nation’s political economy involves contestation at many levels including between international and domestic actors.

      Thank you for sharing your on the ground perspective.


  • Thanks Tobias, excellent comment.

    I think your second to last paragraph is particularly useful. I take the point about actors being potential actors at a range of scales. Nevertheless, I still think it useful to think about actors with respect to different arenas and to think about why different actors might find different scales more or less empowering.

    On public expenditure, the World Bank resource is great, thank you for pointing it out, and thanks for a more detailed answer to my questions about education and health spending. I have modified the spreadsheet somewhat interested readers can download my modified version here.

    You correctly point out that in absolute terms expenditure on health and education have increased (I’m not sure the increases actually warrant the term ‘exponential’ but I won’t be a bore). However, we are talking about a government recovering from a catastrophic shock. Much of its spending has increased. This means what we’re interested in is relative increases. If we look at SIG funded health spending as a share of total SIG government spending over the period for which there are data it has remained constant at about 12%. If we look at education spending in the same way we find it to stayed constant from 2006 to 2012 (about 23% to 22%). It lurches up to 32% in 2013. This is a result (as best I can tell) of increased funding going to training (including overseas training), a questionable budget line.

    You stated in your review that “Public expenditure, overall, seems to have been better aligned with international development goals than the sustainment of national patronage networks.”

    I think public expenditure patterns are a product of (a) a degree of path dependency (b) MPs’ seeking funding for patronage (c) contestation between donors and SIG over spending priorities, and (d) some national orientation on behalf of some MPs and many civil servants.

    I’d be interested to know what aspects of my take you disagree with?

    • Thanks Terence – and sorry for the misspelling in my last.

      I’d agree with your take. My pretty simple point is that this wasn’t a case of all available resources being diverted to patronage, as we seem to agree. Perhaps also worth noting that Solomons spends a relatively large share of the budget on health and education relative to other countries at similar levels of income (or at least did so back in 2014 when I last looked closely). I guess this also speaks to donor influence, and arguably against any complete dominance of patronage.

  • Hi Tobias,

    Thanks for the review. As someone who’s reading the book it’s useful to encounter other people’s takes while reading. As someone who works on Solomons it’s great to know I can weigh up the authors’ take, my take, and your take on that particular case study.

    I haven’t gotten to the Solomons chapter yet; perhaps when I do, your take will make more sense to me. However, as a conceptual insight the idea that different actors gain competitive advantage from shifting the scale of contestation strikes me as very useful.

    For example, it helps highlight potential problems with decentralisation (and not just in developing countries.) This an improvement given decentralisation is still often sold as an efficient technical fix.

    To give another example, aid actor’s who were cognisant of the scale of different power dynamics might have been considerable more alarmed about the decentralisation of (some) electoral powers to provinces in the lead up to the 2017 elections in PNG. Donors, unless they want to invest very heavily in placing their staff on the ground, have much less ability to work as a countervailing power against electoral subversion at a provincial level than they do when dealing with a single national electoral commission. On the other hand, powerful local actors will probably find it much easier to capture provincial electoral commissions than they will the national body. Decentralisation was only partial in the lead up to the 2017 election, and we are yet to fully diagnose the problems that occurred, but the dynamics of scale ought to be brought to bear as actors undertake diagnosis and look to future proposals.

    To give another example, scale does a good job of explaining why federal models of governance are often talked about in Solomons and PNG, but are unlikely to ever be legislated for by national MPs, unless regional political blocks develop or perhaps during crises of succession. Why would actors at a national-level give resources and power to actors at the provincial level?

    Scale seems useful to me.

    Briefly, on empirical matters, you write:

    “Government expenditure on basic public services (including health and education) grew exponentially over the RAMSI period, with such expenditures consistently accounting for a large share of the annual budget, and a much larger share than CDFs.”

    With respect to government expenditure I think the most appropriate baseline is pre-state collapse. Even setting this aside, are you sure that *non-aid-funded* government expenditure in health and education has grown exponentially? Also, in education’s case, it’s worth noting that much of the growth that did occur was in scholarships, which are a tool of patronage plain and simple.

    Also, you write,

    “Further, if we are to accept the argument that all public expenditure provides patronage channels for national elites, then why did national elites go to the trouble of establishing and resourcing CDFs in the first place?”

    There’s an obvious answer to this question. CDFs provide MPs with funds that are more electorally useful than central resources. The state has very little reach into rural areas in Solomons. And most Solomon Islanders are not going to be employed by the state. Therefore it is hard, notwithstanding the special case of scholarships, for MPs to put conventional state tools to work to their electoral benefit. As a result, it’s difficult for MPs to win votes through these state-based avenues of patronage. On the other hand, cash for solar panels, tin roofs, and OBMs is very useful. It can be targeted and it’s easy to deliver. No mystery.

    Finally, now in the realms of outer nerd land, I want my regressions to have “predictive power”, but I’m not so sure that this is a necessary demand of an initial theory based on case studies. I’ll be happy if–when I read the book–I find that the idea of scale adds additional explanatory power to the cases vis a vis the counterfactual of conventional political economy analysis. On the basis of the authors’ talk last week, I’m hopeful.

    • Hi Terrence

      I have to say I admire your willingness to engage in robust discussion even before reading the case study ; )

      Firstly, on public expenditure. Check out the government expenditure database available from the World Bank open data portal. Yes, I’m sure that government non-grant expenditure on health and education increased exponentially over the post RAMSI period. Health expenditure increased from around SBD90 million in 2006 to around SBD290 million in 2014. Education spending increased from SBD174 million to SBD695 million – a huge increase even excluding scholarships (which accounted for about 25% of education spending in 2016). Overall patterns of public spending are fairly well-aligned with MDG-type outcomes, and were especially so prior to the recent explosive growth of scholarships and CDFs since around 2012. If ‘national elites’ are only interested in ‘strengthening the national scale’ why were they so responsive to donor priorities in setting expenditure allocations?

      Please take another look at my review: I don’t think anyone who has worked on Solomon Islands in the past decade could fail to understand how CDFs function as patronage tools, and I wasn’t questioning that. I was trying to point out that if national level MPs are only interested in distributing patronage, then it is difficult to explain the radical expansion of non-CDF expenditure aligned with international goals. If we explain the radical expansion of MDG-type expenditure on the (incorrect) assumption that all government expenditure can be used to distribute patronage and therefore supports the ‘primacy of the national’, then there is no particular need for CDFs. You can’t have it both ways.

      I am also very confused by what value this framework brings to discussions of decentralization. As I tried to point out, the big-picture constitutional review process has fizzled, but national MPs quite explicitly provided additional resources to provincial governments through increasing provincial grants (both in nominal terms and as a share of the budget) and through co-financing capital grants under the PGSP. Again with this example, MPs acted to protect the ‘primacy of the national scale’ – except when they didn’t!

      As for electoral commissions – do we need a ‘politics of scale’ framing to understand that it’s easier to monitor centralized functions and agencies than decentralized/deconcentrated ones? This is a consistent finding from the decades-old literature on public administration reform. Check out Nick Manning’s review of the Machinery of Government Program in Sols for a discussion of this.

      Much more generally, I’m just not convinced that talking in such monolithic terms as “national”, “local” or “international” is very helpful. Most recent work on state-building emphasizes that ‘national elites’ have divergent and competing objectives. Is a Solomon Island MP a ‘national’ actor? And what does that mean in tangible terms? He might be connected to Chinese businessmen or Malaysian loggers – so presumably he also cares about security at the local scale where they do business? He might have business interests in the expat rental market or a water supply business in Honiara, in which case he will have an interest in the presence of donors, presumably linking him to the international scale, and giving him an interest in the quality of the local business environment. He might even be seeking to build credibility with the international development community to mobilize aid resources, in which case he presumably has an interest in achieving the global goals and targets that we are allegedly fixated upon. He could be in a coalition with or seeking to undermine any number of other MPs who have various configurations of these interests. This just all seems to suggest we are better off mapping real elite coalitions and interests, including trying to understand specific patronage networks, rather than relying on the categories provided by the scale framing.

      To me, all work in this area is valuable and welcome. I have no problem with ‘politics of scale’ as a way of kicking up some new ideas. But I’m not at all convinced by the authors’ claim that it is superior to other frameworks of analysis (whether measured in terms of either predictive or explanatory power).

Leave a Comment