The material on Beyond Budgets is the responsibility of individual authors alone. It does not represent the views of any project, organisation or government.


Of institutions and butterflies

Author: ODI Researcher Philipp Krause

Ideas sometimes travel in strange ways. Isomorphism and isomorphic mimicry have been terms in biology since the 19th century. They refer to different organisms evolving to look similar without actually being related. In the case of mimicry the process is by, well, one organism mimicking another to gain some sort of an evolutionary advantage. Think of a perfectly edible species of butterfly that looks like another, not so edible, species to avoid being eaten.

The idea made its way into organizational sociology in the 1970s and 80s, most notably by the work of Powell and DiMaggio. Now, nearly 30 years later, it has become a buzzword of sorts in development policy. Pritchett, Woolcock and Andrews (PWA) in their work on “capability traps”, and Matt Andrews specifically with regards to PFM reform in Africa, have used isomorphism to describe the highly negative consequences of donor-assisted reform efforts to establish formal institutions in developing countries.

This mini-etymology matters, because along the way the connotation of isomorphic mimicry changed. In biology mimicry is a good thing, bestowing an evolutionary advantage on the mimic. In organisational sociology Powell and DiMaggio made the point that organisations can mimic other organisations without evidence that it actually increases performance. Organisations are imitated because they are perceived as successful, because others depend on them, or because they are seen as ‘the thing to do’ in a certain profession. But for PWA imitation of perceived success is unreservedly negative: part of the reason fragile states are hopelessly stuck is precisely because they try to mimic the formal institutions of success, rather than figuring out the functions of statehood on their own. According to PWA, mimicry is the expression of a teleological worldview, a futile chase for that one best-practice path towards development.

In the last year or two the PWA take on isomorphic mimicry has received a lot of attention, especially in the context of the 2011 World Development Report on fragile states. The paper was reviewed favourably in the blogosphere (see here, here and here). As far as I know, their idea that institutional change by mimicry is a bad thing stands unchallenged.

I think this is throwing the baby out with the bathwater.

Not to say that donors don’t deserve every bit of the opprobrium implied in this work. In Matt’s survey of African PFM reforms, the vast bulk of 31 governments surveyed pursued MTEFs, performance budgeting, top-down budgeting and so on, often at the same time. Given the variation of these institutions among OECD countries, it’s just inconceivable that any sound analysis of functional need could produce this incredible homogeneity. Inevitably some, if not most, of these 31 countries are ill-served by these reforms. So telling every government they need to mimic best practices is a very poor strategy for donors – but does that mean mimicry is automatically a bad thing for governments themselves?

Plainly, governments mimic one another all the time, and often quite successfully. Powell and DiMaggio cite the example of Meiji Japan. In the last decades of the 19th century, Japan copied the postal system from Britain, the police from France and the army from Prussia. Did they think the Prussian Army was most suited to Japan? No. Prussia’s army was reputed to be the best; though nobody knew why specifically (there were other modern, experienced armies around). Japan copied what they thought was the best case, and it worked very well for them. This is precisely the point. We hardly ever have good evidence for exactly why a certain institutional innovation works, or even if it really works. Given the complexity of social systems, we may never really know. Looking to good performers as models to imitate is just fine, as long as governments know how to adapt on their own.

What happens with successful reforms is of course far more complex than simple copying and pasting of organizational charts – which is what PWA find so offensive. But rarely can you find examples of historically embedded, problem-oriented institutional change without an element of mimicry as well. Matt Andrews has recently written about the 1990s budget reforms in Sweden, highlighting how they were anchored in Sweden’s fiscal history going back decades. True, but Swedish officials also looked at their European neighbours – specifically von Hagen’s index of budgetary institutions, which was very new at the time. This index is nothing if not formal-institutional. Swedish officials didn’t want to remain in the “spaghetti league” with Italy and reformed their institutions accordingly.

Even when governments in developing countries adopt formal institutions insincerely, to follow donor demands, it is not clear that such strategies are inherently bad for governments. Take PRSPs. A national poverty reduction strategy based on consultations with domestic groups, monitored according to international standards and focused on all things that donors hold dear clearly doesn’t fit most government’s institutional settings. The gap between PRSPs and regular government structures is not surprising, and in fact a perfect example of a “ceremonial institution”.

If governments are rational actors, why wouldn’t they adopt these insincere institutions that are required to access debt relief, grants and cheap loans? The cost-benefit calculation would surely beat raising taxes. The crucial point here is that governments are actors. If the analogy to biology is to make any sense, governments should not be assumed to strive for development but for their own survival. Survival for organisations is enhanced by greater resources, so developing country governments will continue to implement superficial best practice reforms as long as donors ask – and pay – for them.

Along with genuine innovation by mimicry and insincere mimicry there is a third kind. This occurs when fragmentation and informality in government is so high that purposeful action is very difficult. When the centre of government doesn’t have enough autonomy and capability to learn and adapt, externally sponsored formal reforms become very dangerous. This is not isomorphic mimicry, though, but institutional ventriloquism. It happens when best practice reforms are articulated, planned and implemented following external prompting and via externally funded advisers and consultants. Genuinely local government units are further fragmented as they align themselves to available funding and manpower.

Institutional ventriloquism worsens capability traps because it keeps states from developing the autonomous capability to adapt and change. For states that are able to adapt and change on their own, mimicry isn’t a problem.

PrintView Printer Friendly Version

EmailEmail Article to Friend

References (12)

References allow you to track sources for this article, as well as articles that were written in response to this article.
  • Response
    Response: minecraft gratuit
    Now Air pocket Variation of indie mega-sentience Minecraft was purchased a stupefying 3100 times across the iOS and Android application sales in France.
  • Response
    Response: visit homepage
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    Response: sports betting
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
  • Response
    Response: more info here
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies
  • Response
    beyondbudgets - Beyond Budgets - Of institutions and butterflies

Reader Comments (8)

Thanks for these comments. In our most recent work on this subject -- see working paper 299 at the Center for Global Development (cgdev.org) -- we try to articulate a positive operational response to 'capability traps'. In so doing, we also note that mimicry is at least a plausible strategy for those development problems which have a technical/universal answer but itself the problem when the challenges (those as those in governance and legal reform) require context-specific solutions that are unknown (perhaps unknowable) ex ante, and thus must be arrived at through an adaptive, iterative process.

December 3, 2012 | Unregistered CommenterMichael Woolcock

Hi Philipp
Great post. I agree with you on many ways. In Sweden, for instance, you are totally correct. The group of officials who advocated change and were instrumental in designing the change were seriously impacted by Von Hagen et al. They used the work on top down systems and the indexes to draw attention to the problems in sweden's system and to feed the search for alternatives. In my forthcoming book I argue that external examples can be vital fuel for change in the way that they allow agents of change to draw attention to problems in a comparative sense and then use these examples in the process of finding and fitting new change alternatives. I seldom see functional solutions emerging from mimetic reproduction of external ideas and models, though. The key, as you note, is how external examples can become valuable in contexts that are adaptive. Governments in such contexts don't mimic. Rather, they use external ideas to foster change as part of their adaptive process. In my forthcoming book I argue that most functional solutions are a blend of past internal ideas, new internal ideas and external practice. The key is not one or other, however, but the process through which they are blended together into functional hybrids. So, if external ideas are the baby...I say lets keep it...but work out how it matters in the process of change.
Great post, by the way.

December 4, 2012 | Unregistered CommenterMatt andrews

In the context of isomorphic mimicry, I agree with your view that it is indeed risky to throw baby with the bathwater. The colonization in last two centuries provides the best example of isomorphic mimicry – institutions and practices of the occupied country were discarded in wholesale while institutions and practices from the occupying power were implanted. This happened regardless of whether the occupying power was England, France, Spain or Portugal – no wonder we have Anglophone, Francophone, Spanish or Lusophone legacies. Legacy institutions and practices continued to flourish even after countries attained independence. The question to ask is how alien institutions and practices became ingrained into the DNA of national systems of these newly independent countries? I think the key answer is: tone at the top and political will – the occupying power was convinced about the superiority of their systems, never sought approval from the occupied population before implanting their systems, and used their might to ensure successful implantation – over a period of time people not only accepted these institutions and practices but became defenders of these systems (for example, the Indian Civil Service is essentially the same as British left it in 1947).

December 4, 2012 | Unregistered CommenterSanjay Vani

These are great comments - thanks Michael, Matt and Sanjay.

The point I find most important is that imitation of success happens all the time, and it's not necessarily a bad thing.

Institutional change is a complicated business, and the work of Matt, Lant and Michael rightly puts the spotlight on how this actually happens in developing countries. To properly understand how adaptation works should be a huge priority for development community.

To me, mimicry definitely is a form of adaptation - if only because that's what it is in biology. It assumes there is one agent A mimicking a model agent B in order to better survive. I think that's entirely consistent with problem-driven adaptation, and entirely possible to happen in governance-related areas.

Sanjay, a very important point - although this form of coercive isomorphism isn't likely to happen much anymore. It would be interesting to trace how exactly the ICS turned into the IAS, since, as you say, it is now an Indian institution. I can think of other inherited civil services that seem to work quite well (like Nepal), but many cases where the colonial institutional legacy really doesn't serve the successor country well at all.

The worst examples of "mimicry", the copying-and-pasting of formal institutions without any understanding of (or being bothered by) context, shouldn't be called mimicry, they should be called institutional ventriloquism. If the reforms are prompted by external actors, then planned and implemented by external advisors, and funded externally as well, where's the agency?

I don't think we know enough about why some governments seem perfectly capable of taking an international model and then make sense of it domestically, while others don't. Is this just a function of income? I don't think so...

December 4, 2012 | Unregistered CommenterPhilipp Krause

Thank you for sharing the post. It’s well argued.

However, allow me to propose a different view than the ones above. Doesn't the key trouble with the mimicry etymology arise, if you take it to mean that mimicked institutions merely become paper tigers? In other words, that it misses that what from a programming view is considered to be failed or empty institutions rarely are without functions.

I’ve recently taken inspiration from Avner Greif and Frances Cleaver, among others, to argue that institutions are complex and dynamical emergent properties of socially positioned actors’ organising of rules, internalised- and strategic beliefs, norms and organisations. From such a viewpoint, and interpreting what most of you have written above, isomorphic mimicry isn’t as much good or bad as it is vulnerable to being a mischaracterisation of what institution building actually entails.

Thanks again.

December 5, 2012 | Unregistered CommenterSøren

Great post and interesting discussion. There seems to be agreement on not throwing out the baby – that adopting external ideas/practices/forms can be positive in at least some cases – but that the bathwater still needs to go. The key question then becomes: what distinguishes good mimicry from bad mimicry?

Answering this requires us to disaggregate the concept. Philipp has done this with his typology of genuine mimicry, insincere mimicry, and institutional ventriloquism (a great term). There are some useful distinctions here, but for another approach we can go back to the source:

“We identify three mechanisms through which institutional isomorphic change occurs, each with its own antecedents: 1) coercive isomorphism that stems from political influence and the problem of legitimacy; 2) mimetic isomorphism resulting from standard responses to uncertainty; and 3) normative isomorphism, associated with professionalization.” (DiMaggio and Powell 1983, 150)

Importantly, for each of these categories the relationship between isomorphism and organizational performance is ambiguous. DiMaggio and Powell note that dependence on external parties for resources is one cause of coercive mimicry – this is the sense in which Pritchett, Woolcock, and Andrews use the term ‘isomorphic mimicry’, and it is plausible that in this case isomorphism is typically a bad thing for organizational performance.

But isomorphism also occurs for other reasons that may be more positive. DiMaggio and Powell note that mimetic isomorphism is often a response to uncertainty, and so organizations model themselves on other organizations that are perceived to be successful. They hypothesize that the degree of mimicry will increase “the more uncertain the relationship between means and ends” (154) and “the more ambiguous the goals of the organization” (155). When an organization’s outputs and/or outcomes are not very visible and its goals are ambiguous, the organization is very difficult to manage (c.f. Wilson 1989), and so in many causes mimetic isomorphism might actually be the optimal strategy.

DiMaggio and Powell’s typology of isomorphism highlights how complex the issue is - they identify at least three broad categories of isomorphism, each of which can be good, bad, or indeterminate for performance. Philipp’s breakdown is also useful, and introduces even more aspects to the analysis. I think this moves us away from a blanket condemnation of isomorphism and toward questioning exactly when and how context matters, and examining more closely the relationship between particular contexts and universal mechanisms. Michael and Matt’s comments also hint at one somewhat process-based way of approaching this question. I look forward to seeing how Matt’s new book addresses these issues.

December 11, 2012 | Unregistered CommenterMartin Williams

Hello Philipp,

As the other commentators say, this is a very nice piece and it flags a number of issues that truly deserve to be further discussed. Governments, as you say, have always "copied" (but also "learned" from) other governments, and they will certainly continue doing so.

I would just add a few comments for the sake of further "cleaning" the water in which the baby is bathed... PWA (both in their original piece and in their 2012 one) adequately point at the risk of adopting (and, before that, spreading) best practices just for the sake of it, or for unclear "legitimacy" purposes. To begin with, how do we actually know for sure that a best practice is actually best, or that it even is a real practice and not just talk. Those (and the copy-pasting possibility you flag) are indeed risks that should not be overstimated, as Andrews and other authors in the policy diffusion/transfer have shown.

Now, apart from the points that you have already mentioned, let me suggest the following (without any specific order), mainly based on my doctoral thesis on cross-national policy learning about management by results policies in Chile and Mexico:

- isomorphism, and the discussion about it, suggests a focus on an output (an organizational model or a best practice), which is copied by other countries. Yet in many cases one can observe that "best practices" are not copied, but merely used as reference and then combined with many other inputs before a new institution/organization/practice is built (a point Andrews also mentions above). So perhaps it might be better to speak about isomorphic trends, rather than isomorphic mimicry?

- DiMaggio and Powell discuss three mechanisms that lead to isomorphism, which are certainly mentioned both by you and the literature you discuss. Yet they are only that, mentioned in both places, and then get conflated into one term (isomorphic mimicry). I would suggest things might be different depending on whether isomorphism was led by mimetic, epistemic, or coercion means...

- Then the discussion about the spread/adoption of "best practices" seems to imply developing countries were forced to follow an institutional/organizational/legal "form" without an appreciation of their national needs/conditions. But I would contend that a) not developing countries are equal, and thus not are similarly "forced" to adopt a foreign "best practice", perhaps they actually promote the process; b) not all developing countries are equal and some have a higher capacity to learn from abroad (as you suggest); c) sometimes a "best practice" arrived to the country before the international organization...; d) "best practices" are sometimes copied/used as a reference both because of "form" and "function" considerations (e.g. performance based practices, but also transparency laws); e) a number of good or best practices in the field of public management/governance do not necessarilly spread through donor/country relationships, but through other difussion channels (e.g. international conferences, regional networks, etc.) as "butterflies" (your metaphor); and f) "best practice" imitation without a clear prior assessment of the copied practice's effectiveness also happens among developed countries (e.g. central bank independence, regulatory agencies, as Al Roberts has shown)

- Then two last points, which are certainly mentioned in your post (and in PWA 2012) but which might require more attention in the future, relate to the question of a) learning (how, what and how much are governments capable of learning about international "best practices", and thus how do they actually build those capacities); b) temporal factors (that a "best practice" worked in Sweden does not mean it can equally work in Mexico; but that a certain "best practice" (or idea) transferred, say, to Chile did not work in the first 5 years after its implementation simlarly might does not mean it cannot work a few years later, and even become an international "best practice" in its turn (e.g. monitoring and evaluation system).



December 11, 2012 | Unregistered CommenterMauricio Dussauge

Thanks Martin and Mauricio, these are very interesting points. From your posts, I take away two issues in particular.

First, whether you follow Powell and Dimaggio's or some categorization, isomorphic mimicry is not automatically a bad thing. That's far from saying it's automatically a good thing, either. And it's worth a lot more investigation to figure out how different kinds of mimicry, or other types of isomorphism for that matter, really work.

Second, this stuff plays out differently in different countries. I have little doubt that Mexico and Chile are able to use mimicry for their advantage (being OECD members, it's good to compare well on "OECD best practice"), without burdening themselves with copy-paste institutions that make no sense. PWA, of course, did their work on capability traps with fragile states in mind, which is a different story altogther...

December 14, 2012 | Unregistered CommenterPhilipp Krause

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>
« The unintended consequences of civil service reforms in developing countries | Main | It's politics, we’re not stupid: in defence of a division of labour between technocrats and politicians »