Guest blogpost: Andrew Barnett on low carbon development trade offs

Solar charger, Kenya_Solio_Flickr Creative Commons

Andrew Barnett, Director of The Policy Practice,  inspired by the discussion at The Royal Society on Monday 31 March 2014 to discuss the policy implications of research on low carbon development and sustainable energy access in low-income countries, writes for the STEPS Centre blog.

I very much enjoyed meeting the teams from Science and Technology Policy Research (SPRU) at the University of Sussex, the STEPS Centre and the African Technology Policy Studies Network (ATPS) group from Kenya. Their recent working paper 61, Sustainable energy for whom? Governing pro-poor, low-carbon pathways to development: Lessons from solar PV in Kenya, provides a useful contribution to the field, not least by providing a detailed history of Photo Voltaics development in Kenya. The paper draws strongly on the huge intellectual legacy from many years of research at SPRU on the limits to simplistic views of technology transfer and the need for an innovation systems approach to the management of technical change.

In this blogpost I would like to see if I can help to build on this work.  The paper makes the point that there is a perception of a  “tension… between increasing energy access and pursuing low carbon development” (p Xii and page 2).  But it seems to me that it is a great deal more than a perception and this trade-off may well be one of the key intellectual issues in this area of research.  There are now a number of authoritative sources to suggest that there can be a trade-off between these two objectives and they cannot be so easily dismissed.  No less an authority is the current chief economist at DFID, Stefan Dercon, who wrote a paper for the World Bank on this topic, see Policy Research Working Paper 6231,  He concludes:

“Green growth is in no way necessarily bad for the poor. But the key message of this paper is that promises that green growth will offer a rapid route out of poverty are not very plausible; there may well be less rapid an exit than with more conventional growth strategies. To sustain growth, green growth also needs to be weighed in terms of its ability to reduce poverty. To sustain poverty reduction, green growth may involve giving up some possible environmental benefits, to keep the growth-poverty elasticity high. Since poverty reduction remains at the top of the agenda, different shades of green may be needed. In particular, poverty reduction is a powerful force for giving the poor more resilience to the increasing risks of climate change; they should not be asked to pay the price for greening the planet.”

For me understanding that a trade-off CAN exist is at the heart of the “political economy” that is increasingly seen as being of critical importance to understanding apparently illogical human behaviour in this area (see  Of course if there is to be lots of genuinely ‘new money’ for low carbon growth, it is a laudable objective for academics to see how they can help to capture as much of it as possible for the poorer people of Africa.  But for me the SPRU analysis would be strengthened if it accepted and dealt with the possibility of a trade-off.  Indeed if the possibility of a trade-off were put at the centre of the work it would lead nicely onto the other issue at the heart of the STEPS programme namely the existence of “diverse pathways“.  The issue is NOT (as suggested at the meeting) that new technologies always start off less competitive than established technologies, it is that from among all the possible future pathways some will contribute more to energy access and others will contribute more to low carbon “development” for a given level of resource input (by the way there does not appear to be as much effort into unpacking the concept of “energy access”, or even development, as are applied to other terms in the analysis – the reduction of energy poverty must surely involve USE of modern energy services and a view as to what is considered to be an adequate or minimum level of consumption).

A related, but different, point is raised when the report says in passing that “many low carbon alternatives are not yet competitive with carbon intensive technology options” (page 23). This is likely to be true, but it also seems to need more analysis in any discussion of modern energy services and poor people. Not least of the reasons being that it goes a long way to explain the behaviour of governments that act on the belief that at least FOR THE MOMENT PV are too expensive for many of their poor people.  A number of high officials in Tanzania, and researchers in Kenya (such as Stephen Karekezi, Director of AFREPREN/FWD) explicitly explained this to me in the 1980’s.

The other key issue that you have to confront when talking of the energy needs of poor people is that most poor people do not currently have sufficient “access” to any modern energy services, whether low or high carbon, primarily because they are poor.  If our analyses are to be useful they must address this issue head on and centrally.  One of the report’s team said at the Royal Society meeting that “subsidies are essential”  if poor people are to have access to low carbon energy options. I am happy to support this view, but it is almost trivial as a conclusion. The report’s main conclusion states that market forces alone will not drive the widespread uptake of low carbon energy technologies in low-income developing countries”.  But surely the interesting point is what is said next.  The team needs to say more about what “market making subsidies” would look like as there are already well documented cases of market destroying subsidies.  But more importantly the team could make much of their SPRU colleague’s work on the need for the state to undertake risky equity investments if innovation is to be dynamic, and to demand an equity return on their successes (see Prof Mariana Mazzucato, Professor in the Economics of Innovation at SPRU).

If the main problem of “energy access” for poor people is poverty then the team also could say useful things about the need for energy end-use technologies that provide the energy services that help people to increase their incomes.  This is of course a complex area, but then SPRU has never been frightened by complexity!
While the PV industry provides a most useful case study (and clearly the team know a great deal about PV), the policy analysis would be strengthened by locating this work within the broader focus (framing?) of other energy delivery systems (both for electricity, and for the other energy services that poor people need/want, such as heat for cooking). This would open up a number of other “diverse pathways”, some of which are also low carbon (such as co-generation from agricultural waste, hydro and geothermal electricity to name but three).  This is a big ask of course, but it would also underline the many trade-offs within the low carbon scenario.

Photo Voltaics are already the least cost source of electricity in more remote areas, and for sparsely populated areas (particularly for LED lighting, phone charging and radio).  But for me the relative success of PV in densely populated parts of the world is an index of the failure of power sector reform in that particular country.  Such reform is where agencies such as DFID would get the biggest bang for their aid buck.  Focussing solely on PV diverts their attention from these more critical concerns. It is noteworthy in this context to remember that while PV is flourishing in Kenya so are the companies supplying electricity to Kenya from fossil fuels at huge cost per kilowatt hour in name of emergency power!  We must guard against continuing to play our favourite violins when Rome is clearly burning!

Useful links


  1. Posted for Andrew Scott, Research Fellow, Climate and Environment Programme, ODI:
    As you’d expect Andrew packs a lot of points into a short piece. There is nothing particularly controversial in the idea that there are potentially trade-offs between pursuing an energy access objective and pursuing a low-carbon energy objective. Whether the two objectives are compatible depends on location and application. Increasingly renewable options are competitive, and in remote and off-grid systems, renewables are already often the most cost-effective option, as Andrew notes. So, the trade-off is not always there. (Andrew’s reference to discussions about PV in the 1980s doesn’t help his case. Costs have come down dramatically – see useful note
    on this published this week

    I would draw a distinction between Dercon’s macroeconomic argument about green growth and poverty reduction and the microeconomic or sectoral nature of the SPRU/IDS research. At the macroeconomic policy level there is a balancing of objectives which may lead to trade-offs between them. In the energy sector this is a balance between energy security, energy access andlow-carbon energy. At the microeconomic level, for the energy consumer, there’s a similar balancing but the factors that influence the decision are different. The research team do recognise that they have examined only a piece of the energy picture in Kenya, and are aware that the government there is far more interested in developing geothermal resources than it is in extending electricity access through solar PV. Andrew’s political economy point is particularly relevant to why this is the case.

    Andrew’s subsidies point is an area we have been discussing in ODI. If subsidies are necessary to enable the poorest households to access adequate energy services, what is the best way to provide them? Subsidising electricity tariffs tends to benefit all consumers, not just the poorest. Subsidising the technologies (energy or end-use) may have the same effect. Cash transfers may not be used for energy consumption. Our sense is that this is an area for further research.

    Andrew B suggests that the spread of PV is as much the result of failure in the wider energy system as it is a success of technology transfer and diffusion. I think there’s some truth in this. However, just as green growth doesn’t necessarily reduce poverty, power sector reform doesn’t necessarily extend energy access to the poor. Where there has been success in extending access to 90% or more of the population, the capacity to generate and distribute electricity has been accompanied by policies to ensure the poor are included. Which takes us back to the subsidies question.

  2. Posted for Rob Byrne, STEPS Centre and SPRU:
    Thanks Andrew (Barnett) for stimulating this discussion. You’ve made some important points on energy access, as I would fully expect you to do! And thanks also to Andrew Scott for your thoughts, some of which I would have tried to articulate myself but you have done so much more clearly than I am likely to have achieved.

    Before I come to some specific points in the discussion, I think it is worth clarifying what the research was actually trying to do. We were interested in explaining the success of the Kenyan PV market and seeing if there were any lessons that could be learned for the promotion of PV elsewhere, as well as for other low carbon technologies. PV in Kenya is an interesting case for several reasons. First, it is said to be the most successful per-capita off-grid PV market in the world. Second, the PV market in Kenya is usually described as ‘unsubsidised’. Third, there has been a lot of rhetoric around solar home systems (SHSs) providing a solution to the issue of off-grid energy access (however defined). Fourth, Kenya was one of the pilot countries of the Lighting Africa programme, which sought to promote PV lanterns for the so-called bottom of the income pyramid ‘market’ in order to sustainably replace kerosene-based lighting. And this programme appears to have been very successful in Kenya.

    In explaining how the PV market has grown so much in Kenya, we found that it has actually benefited significantly from subsidies of various kinds that have helped to build a partial innovation system around PV, SHSs have not been reaching the poor (in line with the findings of others – see, for example, Jacobson’s 2004 PhD) but have been adopted by a rural middle class, and the Lighting Africa programme has also helped to build an innovation system of provision for solar portable lamps (SPLs) that do appear to be reaching poorer people.

    What emerges from this is that the rhetoric of a private unsubsidised market is not borne out in this case and so, to respond to one of Andrew B’s points, it is worth noting that market forces alone do not explain the success of PV in Kenya. Further, it is also not simply a case of creating the right enabling environment (i.e. conducive policy, regulation, etc.) in which entrepreneurs and market forces can then act. In line with Mariana Mazzucato’s ideas of the entrepreneurial state, public money (mainly from donors) has been creative and constitutive in relation to developing technologies but, more, has also been critical to developing local capabilities. (I should say that we use the definition of Martin Bell and Keith Pavitt for capabilities: “skills, knowledge and experience, and institutional structures and linkages”.) In other words, we do go beyond noting that market forces are insufficient (our first conclusion rather than our main one), making something of the next step in the discussion that Andrew B rightly says is important.

    What we have not done is to draw out from the evidence what ‘market-making subsidies’ would look like. They are described in the case study but not highlighted in the conclusions. Instead, we have discussed what these subsidies need to support – meaningful network-building, rich context-specific learning, publicly-available market and technological research, awareness-raising, experimentation. All these activities have been supported – to different extents and at different times – in the Kenyan PV market, both in its earlier SHS form and in its more recent SPL pathway. It is impossible to know how much subsidy has been provided to the Kenyan market but the Lighting Africa intervention cost about USD 5 million. Many other donor-funded projects have been implemented since the early 1980s, ranging in size from perhaps a few tens of thousands of dollars to much larger. In some ways, the precise figure is unimportant. The interesting point is that the case provides at least some clues as to what market-making subsidies could look like and what they could help to achieve. In our view, this is an as-yet under-researched issue, possibly because of what we would argue is the dominant narrative of private sector led development (and the Kenyan PV market has often been described as an exemplar of this), which closes down discussion of alternative development strategies to some notion of free markets delivering services. If I can stylise for a moment, this closing-down leads to an expectation that properly-financed entrepreneurs, working within suitably enabling policy environments where technology-costs include all externalities, will realise our sustainable development goals. Hopefully, our analysis of the evolution of the Kenyan PV market goes some way to demonstrating that this is not sufficient. Entrepreneurs, finance, ‘real’ costs and enabling environments are all necessary but there are numerous other issues to do with the systemic nature of development that need addressing through active, risk-bearing and creative public support.

    In order to conduct this analysis, we have taken as read some of the points that Andrew B has – again, rightly – argued need unpacking. I agree it is useful to try to understand better what trade-offs there might be and what their nature is. Some of these are discussed in Andrew B’s post and further discussed in Andrew S’s response. One that does not seem to get much attention is the notion of a trade-off between high-carbon development now and the risk that this will lock energy systems into unsustainable futures, thereby storing up huge problems for later. To some extent, this is another political economy question and one, I think, that needs serious analysis. At the heart of why I think this needs serious analysis is the implicit assumption – it seems to me – that switching from high to low-carbon energy services is unproblematic. Political economy would tell us that this is not at all unproblematic: there are high-carbon political and economic interests at play, sunk investments, hard infrastructure, etc., and so they will not be ‘switched off’ easily. But there are also social and cultural expectations, norms, practices, etc., that are similarly resilient – perhaps even resistant – to change.

    Andrew B also makes the point that we have left unpacked the notions of ‘energy access’ and ‘development’. This is certainly fair. And, on the issue of energy access, Andrew rightly distinguishes that it is not just about access but about use of energy services. Which technologies facilitate ‘access’ (or energy-service use) and development (human and economic) is indeed an important question. We have, as Andrew S notes, acknowledged that our study is on just one technology (but several electrical services) in one country. We have also not considered the use of energy services for raising incomes. But we do have aspirations to conduct research on other technologies, services and contexts, and it is clear that more work is needed to understand how to exploit these technologies, etc., to best help poor people raise their incomes or improve their livelihoods. But, by steering away from a sole focus on entrepreneurs providing end-use technologies through private markets to a focus on innovation system building, we believe there is the beginning of an approach to ways of better understanding these complex challenges. We don’t have all the answers but we do, thanks to this and other discussions, have some important questions.

    1. Thanks to all for this stimulating discussion.

      I think this discussion on trade-offs is an important one, but Rob makes an important point regarding the aim and scope of our research which wasn’t designed to address the trade-offs issue (although this is clearly an interesting and important area for empirical research as Andrew B noted in his comments at the event that sparked this discussion).

      I thought it worth flagging that Peter Newell (with Jon Phillips and Ana Pueyo) has led some research that was designed to complement our own project (also funded by CDKN) and which explicitly addresses the issue of trade-offs via political economy analysis of solar in Kenya.

      You can read the policy briefing note here

      The full report will follow soon.

  3. Posted for Russell J. deLucia, PhD, President, The Small-Scale Sustainable Infrastructure Development Fund, Inc.

    In response to Andrew’s blog, I would like to make the following comments:

    1. The matter of use of the modern energy or what my colleagues describe as the productive and life improving energy-dependent uses – be they water pumping, or distant learning and medicine – are critical. These uses and the necessary investments still get too little attention in “energy access” programs.

    2. PV [and improved biomass stoves] continues to get RELATIVELY too much focus – and many of these other “green” options (or at lease much less carbon intensive options) are relatively neglected such as:
    • pico & small hydro, various viable biomass conversion and not just biogas, but also new pyrolysis technolgies and of course LPG don’t get enough focus. Remember Kirk Smith’s recent presentations.
    • And the very challenging problem of addressing Energy Efficiency and improved modern energy access for the millions of micro enterprises – usually in the informal sector. S3IDF experience to share in puffed rice production in India and the problems associated with the use of the dirtiest fuel ever -old tires – which is not just local and global pollution but a cancer vector from what is in the combustion gases.

    3. Failure of the grid players remains and a massive problem, something DFID and others focused on in the past: what has happened to these programs?

    No more for now.

  4. Posted for Stephen Karekezi, Director, AFREPREN/FWD, Kenya

    I read through Andrew’s blogpost which I concur with. Russell’s comment that there is excessive emphasis on clean energy options (such as PVs with a large import content) that generate little income for the poor is spot on.

    I would actually go further and argue that the excessive emphasis on these marginal clean options such as PV has diverted research capabilities, expertise, funding and policy attention from the more important CLEANER options with a proven track record in enhancing access and reducing poverty. As Russel correctly notes, there is very limited work on lowering the cost and increasing the effectiveness of grid-based electrification of the poor, particularly in sub-Saharan Africa where, too often, population growth rates are higher than the growth in connection rates.

    Other proven cleaner energy options that receive limited attention include agro-industry based small/pico hydro or cogen plants that lower the energy cost of processing crops from poor small holders and thus ensure higher revenue flows to low income small-scale farmers.

    I would also add another category of cleaner energy options that deserves more attention namely: proven non-electric cleaner options that directly increase incomes of the poor and expand access. Examples include wood-fired boilers used in both small and medium-scale agro-industries, windpumps for irrigation, biomass-fired or solar driers for agro-processing etc. Many of these non-electric options can provide the basis for flourishing clusters of SMEs.

  5. Posted for Rob Byrne, STEPS Centre and SPRU

    The posts by Russell and Stephen highlight the issues around technology choices, particularly in terms of which technologies could be more beneficial to the poor. This is clearly an important discussion but it would be good to hear if there are any thoughts on what can be learned from our research findings, which have tried to explain how PV has become so successful. That is, to what extent could the approaches used to promote PV in Kenya be used to promote other low carbon technologies more generally? And, crucially, to what extent and in what ways could these other technologies (and PV, wherever it is likely to be appropriate) be used to increase the use of energy services for productive purposes?

  6. I feel sure that one of the lessons is to establish precisely under which circumstances each low carbon option best meets the need.

    Certainly the issue of MES for income generating services is critically important. I have met a lot of people recently in the PV business who say that the key area for expansion is in more powerful PV systems probably based on conventional lead acid batteries. There is said to be “power gap” between the milliamp outputs of conventional SHS and the need to meet a demand for power to run small income generating motors, such as fridges, blowers and cutters, and TV for which it appears there is a considerable demand. I know the Shell Foundation and DFID are interested in this area. But it is proving difficult (because it is difficult to make a sealed unit and theft of batteries is a problem?). Options not using lead-acid batteries is also apparently difficult.
    what do others think?

Comments are closed.