Complexity in Development Cooperation

Complex network of ropes
The Overseas Development Institute (ODI) recently published a working paper by Harry Jones, called ‘Taking Responsibility for Complexity’ (Working Paper 330, June 2011).  The text is not an easy read, but makes it up in interesting content.    I created a Concept Map from the text, outlining and connecting its main ideas.  It doesn’t substitute for reading the text and I would recommend giving it a thorough read.  However, it may convince you to read the text and help with its digestion.
 

Complexity in Development Concept Map (click to enlarge)

Concept Map of ODI Working Paper “Taking Responsibility for Complexity”
 
The author makes an apt analysis of some problems that result from treating complex situations as simple ones, using assumptions, log-frames and policy cycles.  He hits the mark when he notes that many reporting is quite separated from daily work and that plans, reports and tables tend to be used only for reporting purposes.  Complex situations usually only trigger more in-depth analysis, more elaborate reporting requirements and a tighter watch on indicators.  Local staff flights in a compliance strategy, characterised by risk aversion, instrumental use of tools and a focus on ‘low hanging fruits’.
Planning, monitoring and evaluation (PME) is necessarily a tick-box exercise (to fit in with unrealistic assumptions embedded in the tools) drawing efforts away from the ‘real work,’ to justify projects ex post and explain how everything went according to the plan initially set out (whether or not this was in fact the case). (p.13)
Studies have shown that, in this context, M&E is often carried out to ‘prove not improve’: for example, monitoring activities frequently revolve around reporting on expected indicators as predefined in a log frame, rather than providing real space to look at the unfolding effects and side-effects of an intervention (Bakewell and Garbutt, 2004).
Complex situations are often encountered in education and development contexts.  In Cambodia we link high repetition and drop-out rates to the quality of teaching.  Increasing teaching quality through the introduction of student-centred methodologies will reduce drop-out rates.  Many other factors that affect drop-out rates are not forgotten, but considered as assumptions outside the scope of the programme.
The more difficult the problem, the greater the perceived need for careful planning, intricate assessment and consultation and negotiation with partners and interest groups before anything is done.  Implementation is firmly fixed in advance, with programmes and projects tied to specific activities and outputs that result from extensive, even multi-year  negotiations. Efforts during implementation are then restricted to following a rigid pre-set schedule and plan of activities. (p.12)
But how to deal effectively with complexity?  Assumptions are often outside the scope of small development organisations such as VVOB, even if we would take complexity into account.  Complexity might be an excuse to dodge responsibilities for not reaching goals.  One strategy is to free up resources and time for a wider range of activities that may affect the assumptions such as advocacy.  For example, in Cambodia we have spent considerable time helping the Ministry of Education updating the teacher training curriculum.   Another strategy is to move away from simple, SMART indicators that give the illusion of measuring outcomes and impact of the programme, and move towards ‘principle-based’ and ‘mission based’ monitoring and evaluation (p.27).
The text doesn’t provide a simple recipe to deal with complexity, but rather a set of principles, concepts and case studies that may be useful in certain contexts and are loosely based on concepts of complexity theory.  It doesn’t advocate sweeping away traditional tools and instruments, but recognizes that a combination of tools is likely.
What is clear, however, is that complexity can no longer be swept under the carpet. While there is not yet one comprehensive framework, there is a growing collection of models, tools, and approaches to effectively develop interventions in the face of these multifaceted problems.(…)  Shaping policy will always be a matter or degrees, and a negotiation between bottom-up and top-down structures, between planned and emergent responses and between technical and participatory guidance. (p.21)
Decentralisation is a central ingredient of a strategy in complex situations, but may involve a trade-off:
One aspect is that decentralising tasks within government will often require building capacities at lower levels of organisation – in local government bodies and elsewhere. There may be a ‘chicken and egg problem,’ whereby there is reluctance to decentralise tasks to lower-level units until they have proved their capacity to carry them out, even though it is impossible to do this until decentralisation has actually occurred. One solution is to begin by decentralising simpler tasks for which lower-level capacity is clearly evident or for which the costs of failure would not be severe (Marshall, 2008). (p.25)
However, one major criticism of pilots is that too often they are not allowed to ‘fail,’ and hence they provide less opportunity for learning. The importance of ensuring that you can learn from an intervention is emphasised in Snowden’s concept of ‘safe-fail experiments’ (2010): these are small interventions designed to test ideas for dealing with a problem where it is acceptable for them to fail critically.
A way of handling uncertainty is to make use of collective or distributed cognition. Complex spaces need experts to compete/disagree with each other to increase diversity, rather than a consensus based approach. For emergence we need to force conflict by bringing in different people with different backgrounds. In complex systems we should also bring in safe-to-fail experiments and prevent premature convergence by moving people around into different groups.
Questioning one’s strategy is crucial in achieving long term impacts.  This paper inspires to distance oneself from the strategy we may take for granted too often and open up to alternative approaches. 

Thinking positively

In Cambodia VVOB focuses on student-centered approaches in science education in teacher training institutes.  Teacher trainers are guided towards adopting approaches to teaching that involve students, stimulate them to think and engage instead of passively noting down and regurgitating what the teacher declaims.

Some student-centred approaches focus on stimulating students’ writing skills.  Writing and the skills that accompany it (collecting thoughts, filtering, structuring, creativity, conciseness…) are arguably important 21st century skills.  To extend writing in science lessons beyond copying teachers’ notes we introduced techniques such as 3-2-1 sheets, 2-minute papers and creative writing (link to my presentation on these).  These techniques aim at stimulating students to write about what they have learned in their own words.

There may be unintended benefits as well

Interestingly, New Scientist refers to a study from David Creswell from Carnegie Mellon University that revealed that letting students write about what they have learned, stimulating qualities such as creativity and independence, gives them ‘self-affirmation’ that enables them to perform better.  

‘Compared with a control group, students who ‘self-affirmed’ in this way had lower levels of adrenaline and other fight-or-flight hormones in their urine on exam day.’  (Health Psychology, vol 28, p 554, quoted in New Scientist, 27 August 2011)

Apart from letting students re-interpret what they have learned, these writing exercises may also – if well designed – contribute to improving students’ sense of self-worth.

Limits to Performance Pay

Time in between two OU modules, H800 and H807, leaves me some time to catch up on interesting articles, such as one in New Scientist (9 April 2011) on the alleged effects of performance-based payment schedules, commonly applied in the finance sector.

Making teachers’ pay dependent on reaching certain outcomes is frequently hailed as a way to increase the quality of teaching and education.  However, as New Scientist reported, the scientific evidence for the effectiveness of performance pay is surprisingly weak.  Even more, performance pay may prove counteractive and actually reduce teaching quality.



How does performance pay sound?

A central argument for the use of performance pay is that it increases external motivation, the motivation to do something because of the attached reward.  Conversely, intrinsic motivation refers to behaviour based on what we like or are interested in, without an obvious reward attached.  It appears however, that these two forms of motivation are not independent of each other, but inversely related.  This means that introducing external motivators such as performance pay actually reduces the internal motivation, a phenomenon known as the ‘overjustification effect’ (Deci, 1995).

There are several reasons why this overjustification effect may kick into place.  Performance is notoriously difficult to measure, and performance-related systems often end up measuring the wrong things, such as short-term results and outcomes instead of long-term vision and behaviour.  A teacher’s job is more than only achieving satisfactory test results.  Ignoring this multidimensional character of teaching may lead to  “teaching to the test”, a catchphrase used to describe narrowing of curriculum in an effort to elevate student test scores. Moreover, teachers work as members of a team. Introducing performance-related rewards at the individual teacher level might reduce incentives for teachers to cooperate.

Second, gaming behaviour is rife and there may be unintended consequences, like doctors becoming reluctant to accept the most serious cases (study in Annals of Internal Medicine, quoted by New Scientist).  Podgursky and Springer (2007) provide an overview of some documented ‘opportunistic’ strategies, like assisting students with test questions and ensuring that low-performing students are absent at tests.

A possible explanation for the overjustification effect is that performance-pay related systems come across as coercive and controlling (cognitive evaluation theory).  Teachers may encounter so many rules and administration that their motivation and creativity are being sapped.


Finally, performance pay does appears to be useful in some cases, like for repetitive tasks, but it’s unfit for improving learning in complex environments like schools and classrooms. Teachers should be attracted to and retained in the teaching profession with a decent and competitive salary, but maintaining and fuelling their intrinsic motivation seems the best way to optimize their performance.