COMMENTARY: Ruminants, climate change and climate policy

Abstract

Greenhouse gas emissions from ruminant meat production are significant. Reductions in global ruminant numbers could make a substantial contribution to climate change mitigation goals and yield important social and environmental co-benefits. A lthough a main focus of climate policy has been to reduce fossil fuel consumption, large cuts in CO 2 emissions alone will not abate climate change. At present non-CO 2 greenhouse gases contribute about a third of total anthropogenic CO 2 equivalent (CO 2 e) emissions and 35–45% of climate forcing (the change in radiant energy retained by Earth owing to emissions of long-lived greenhouse gases) resulting from those emissions 1 (Fig.1a). Only with large simultaneous reductions in CO 2 and non-CO 2 emissions will direct radiative forcing be reduced during this century (Fig.1b). Methane (CH 4) is the most abundant non-CO 2 greenhouse gas and because it has a much shorter atmospheric lifetime (~9years) than CO 2 it holds the potential for more rapid reductions in radiative forcing than would be possible by controlling emissions of CO 2 alone. There are several important anthropogenic sources of CH 4 : ruminants, the fossil fuel industry, landfills, biomass burning and rice production (Fig.1c). We focus on ruminants for four reasons. First, ruminant production is the largest source of anthropogenic CH 4 emissions (Fig.1c) and globally occupies more area than any other land use. Second, the relative neglect of this greenhouse gas source suggests that awareness of its importance is inappropriately low. Third, reductions in ruminant numbers and ruminant meat production would simultaneously benefit global food security, human health and environmental conservation. Finally, with political will, decreases in worldwide ruminant populations could potentially be accomplished quickly and relatively inexpensively. Ruminant animals consist of both native and domesticated herbivores that consume plants and digest them through the process of enteric fermentation in a multichambered stomach. Methane is produced as a by-product of microbial digestive processes in the rumen. Non-ruminants or 'monogastric' animals such as pigs and poultry have a single-chambered stomach to digest food, and their methane emissions are negligible in comparison. There are no available estimates of the number of wild ruminants, but it is likely that domestic ruminants greatly outnumber the wild population, with a reported 3.6billion domestic ruminants on Earth in 2011 (1.4 billon cattle, 1.1 billion sheep, 0.9 billion goats and 0.2 billon buffalo) 2 . On average, 25million domestic ruminants have been added to the planet each year (2million per month) 2 over the past 50years (Fig.1d). Worldwide, the livestock sector is responsible for approximately 14.5% of all anthropogenic greenhouse gas emissions 3 (7.1of 49GtCO 2 eyr –1). Approximately 44% (3.1GtCO 2 eyr –1) of the livestock sector's emissions are in the form of CH 4 from enteric fermentation, manure and rice feed, with the remaining portions almost equally shared between CO 2 (27%, 2GtCO 2 eyr –1) from land-use change and fossil fuel use, and nitrous oxide (N 2 O) (29%, 2GtCO 2 eyr –1) from fertilizer applied to feed-crop fields and manure 3 . Ruminants contribute significantly more (5.7GtCO 2 eyr –1) to greenhouse gas emissions than monogastric livestock (1.4GtCO 2 eyr –1), and emissions due to cattle (4.6GtCO 2 eyr –1) are substantially higher than those from buffalo (0.6GtCO 2 eyr –1) or sheep and goats (0.5GtCO 2 eyr –1) 3 . Globally, ruminants contribute 11.6% and cattle 9.4% of all greenhouse gas emissions from anthropogenic sources. The total area dedicated to grazing encompasses 26% of the terrestrial surface of the planet 4 . Livestock production accounts for 70% of global agricultural land and the area dedicated to feed-crop production represents 33% of total arable land 4 . The feeding of crops to livestock is in direct competition with producing crops for human consumption (food security) and climate mitigation (bioenergy production or carbon sequestration) 5 . Deforestation has been responsible for a significant proportion of global greenhouse gas emissions from the livestock sector and takes place mostly in tropical areas, where expansion of pasture and arable land for animal feed crops occurs primarily at the expense of native forests 4,6 . Lower demand for ruminant meat would therefore reduce a significant driver of tropical deforestation and associated burning and black carbon emissions. The accompanying reduction in grazing intensity could also allow regrowth of forests and other natural vegetation, resulting in additional carbon sequestration in both biomass and soils with beneficial climate feedbacks 5,6 . Lower global ruminant numbers would have simultaneous benefits for other systems and processes. For example, in some grassland and savannah ecosystems, domestic ruminant grazing contributes to land degradation through desertification and reduced soil organic carbon 5 . Ruminant agriculture can also have negative impacts on water quality and availability, hydrology and riparian ecosystems 4,7 . Ruminant production can erode biodiversity through a wide range of processes such as forest loss and degradation, land-use intensification, exotic plant invasions, soil erosion, persecution of large predators and competition with wildlife for resources 4–7 . Ruminant production also has implications for food security and human

Full text

2 NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange
opinion & comment
COMMENTARY:
Ruminants, climate change and
climate policy
William J. Ripple, Pete Smith, Helmut Haberl, Stephen A. Montzka, Clive McAlpine and Douglas H. Boucher
Greenhouse gas emissions from ruminant meat production are significant. Reductions in global ruminant
numbers could make a substantial contribution to climate change mitigation goals and yield important
social and environmental co-benefits.
Although a main focus of climate
policy has been to reduce fossil fuel
consumption, large cuts in CO2
emissions alone will not abate climate
change. At present non-CO2 greenhouse
gases contribute about a third of total
anthropogenic CO2 equivalent (CO2e)
emissions and 35–45% of climate forcing
(the change in radiant energy retained by
Earth owing to emissions of long-lived
greenhouse gases) resulting from those
emissions1 (Fig.1a). Only with large
simultaneous reductions in CO2 and
non-CO2 emissions will direct radiative
forcing be reduced during this century
(Fig.1b). Methane (CH4) is the most
abundant non-CO2 greenhouse gas and
because it has a much shorter atmospheric
lifetime (~9years) than CO2 it holds the
potential for more rapid reductions in
radiative forcing than would be possible by
controlling emissions of CO2 alone.
ere are several important anthropogenic
sources of CH4: ruminants, the fossil
fuel industry, landlls, biomass burning
and rice production (Fig.1c). We focus
on ruminants for four reasons. First,
ruminant production is the largest source
of anthropogenic CH4 emissions (Fig.1c)
and globally occupies more area than any
other land use. Second, the relative neglect
of this greenhouse gas source suggests that
awareness of its importance is inappropriately
low. ird, reductions in ruminant
numbers and ruminant meat production
would simultaneously benet global food
security, human health and environmental
conservation. Finally, with political will,
decreases in worldwide ruminant populations
could potentially be accomplished quickly
and relatively inexpensively.
Ruminant animals consist of both
native and domesticated herbivores that
consume plants and digest them through
the process of enteric fermentation in a
multichambered stomach. Methane is
produced as a by-product of microbial
digestive processes in the rumen.
Non-ruminants or ‘monogastric’
animals such as pigs and poultry have a
single-chambered stomach to digest food,
and their methane emissions are negligible
in comparison. ere are no available
estimates of the number of wild ruminants,
but it is likely that domestic ruminants
greatly outnumber the wild population,
with a reported 3.6billion domestic
ruminants on Earth in 2011 (1.4 billon
cattle, 1.1 billion sheep, 0.9 billion goats and
0.2 billon bualo)2. On average, 25million
domestic ruminants have been added to the
planet each year (2million per month)2 over
the past 50years (Fig.1d).
Worldwide, the livestock sector is
responsible for approximately 14.5% of all
anthropogenic greenhouse gas emissions3
(7.1of 49GtCO2eyr–1). Approximately
44% (3.1GtCO2eyr–1) of the livestock
sector’s emissions are in the form of CH4
from enteric fermentation, manure and
rice feed, with the remaining portions
almost equally shared between CO2 (27%,
2GtCO2eyr–1) from land-use change and
fossil fuel use, and nitrous oxide (N2O)
(29%, 2GtCO2eyr–1) from fertilizer
applied to feed-crop elds and manure3.
Ruminants contribute signicantly more
(5.7GtCO2eyr–1) to greenhouse gas
emissions than monogastric livestock
(1.4GtCO2eyr–1), and emissions
due to cattle (4.6GtCO2eyr–1) are
substantially higher than those from
bualo (0.6GtCO2eyr–1) or sheep
and goats (0.5GtCO2eyr–1)3. Globally,
ruminants contribute 11.6% and cattle
9.4% of all greenhouse gas emissions
from anthropogenic sources. e total
area dedicated to grazing encompasses
26% of the terrestrial surface of the
planet4. Livestock production accounts for
70% of global agricultural land and the
area dedicated to feed-crop production
represents 33% of total arable land4. e
feeding of crops to livestock is in direct
competition with producing crops for
human consumption (food security) and
climate mitigation (bioenergy production or
carbon sequestration)5.
Deforestation has been responsible for a
signicant proportion of global greenhouse
gas emissions from the livestock sector and
takes place mostly in tropical areas, where
expansion of pasture and arable land for
animal feed crops occurs primarily at the
expense of native forests4,6. Lower demand
for ruminant meat would therefore reduce
a signicant driver of tropical deforestation
and associated burning and black carbon
emissions. e accompanying reduction in
grazing intensity could also allow regrowth
of forests and other natural vegetation,
resulting in additional carbon sequestration
in both biomass and soils with benecial
climate feedbacks5,6.
Lower global ruminant numbers would
have simultaneous benets for other
systems and processes. For example, in
some grassland and savannah ecosystems,
domestic ruminant grazing contributes to
land degradation through desertication
and reduced soil organic carbon5. Ruminant
agriculture can also have negative impacts
on water quality and availability, hydrology
and riparian ecosystems4,7. Ruminant
production can erode biodiversity
through a wide range of processes such
as forest loss and degradation, land-use
intensication, exotic plant invasions, soil
erosion, persecution of large predators and
competition with wildlife for resources4–7.
Ruminant production also has
implications for food security and human
© 2014 Macmillan Publishers Limited. All rights reserved

NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange 3
opinion & comment
health. Roughly one in eight people in the
world are severely malnourished or lack
access to food owing to poverty and high
food prices2. With over 800 million people
chronically hungry, we argue that the use
of highly productive croplands to produce
animal feed is questionable on moral grounds
because this contributes to exhausting the
world’s food supply. Conversely, ruminant
agriculture will remain important in pastoral
or subsistence situations where ruminants
can provide a source of food from landscapes
that cannot be used to practicably sustain
crops (for example, grasslands). For these
regions, particularly in developing countries,
ruminants represent a stock that can buer
against times of bad harvest or other
detrimental uctuations.
In developed countries, high levels
of meat consumption rates are strongly
correlated with rates of diseases such as
obesity, diabetes, some common cancers and
heart disease8,9. Moreover, reducing meat
consumption and increasing the proportion
of dietary protein obtained from high-
protein plant foods — such as soy, pulses,
cereals and tubers — is associated with
signicant human health benets8,9.
Although policymakers strive to reduce
fossil fuel emissions, the livestock sector has
generally been exempt from climate policies
and little is being done to alter patterns of
production and consumption of ruminant
meat products5,10. Annual meat production
worldwide is growing rapidly, and without
policy changes is projected to more than
double from 229milliontonnes in 2000 to
465milliontonnes in 20504. e greenhouse
gas footprint of consuming ruminant
meat is, on average, 19–48 times higher
than that of high-protein foods obtained
from plants (Fig.2), when full life cycle
analysis including both direct and indirect
environmental eects from ‘farm to fork’ for
enteric fermentation, manure, feed, fertilizer,
processing, transportation and land-use
change are considered. Non-ruminant
meats such as those from pigs and poultry
(and marine sheries) have a lower carbon
equivalent footprint, although they still
average 3–10 times greater than high-protein
plant foods (Fig.2). Pigs and poultry also
consume feed that could otherwise be more
eciently consumed directly by humans.
Moving forwards, there are steps
that governments and international
climate negotiators can take to curb
global ruminant increases and reduce
emissions from the agricultural sector.
Reducing meat consumption as a demand-
side mitigation action oers greater
greenhouse gas reduction potential
(0.7–7.3GtCO2eyr–1) than the supply-
side measures of increased crop yields
(0.2–1.9GtCO2eyr–1) or livestock feeding
eciency (0.2–1.6GtCO2eyr–1) (Table2
in ref.5). In terms of short-term climate
change mitigation during the next few
decades, if all the land used for ruminant
livestock production were instead converted
to grow natural vegetation, increased CO2
sequestration on the order of 30–470% of
the greenhouse gas emissions associated
with food production could be expected5,11.
Nonetheless, policies targeting both supply-
side measures to improve agricultural
production eciencies and demand-side
mitigation for encouraging behavioural
changes to reduce meat consumption
(particularly ruminant meat) and waste
have the best chance of providing rapid and
lasting climate benets5. Inuencing human
behaviour is one of the most challenging
aspects of any large-scale policy, and it is
unlikely that a large-scale dietary change
will happen voluntarily without incentives12.
Implementing a tax or emission trading
scheme on livestock’s greenhouse gas
emissions could be an economically sound
policy that would modify consumer prices
and aect consumption patterns12. A tax has
recently been successfully modelled for the
European Union with tax rates proportional
to the average greenhouse gas emissions
per unit of food sold10, although social
justice, equity and food access issues need
to be carefully considered. Such demand-
side mitigation measures have more
social and environmental co-benets than
supply-side measures5.
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
Ruminants
Natural gas,
oil, industry
Landfills
and waste
Biomass
burning
Coal
Rice
Emissions (Gt CO
2
e yr
–1
)
4.0
3.8
3.6
3.4
3.2
3.0
2.8
2.6
2.4
2.2
2.0
Year
Billions of ruminants
N2O
0.17
Ozone-
depleting
substances
0.32
CH4
0.50
Other
0.01
CO2
1.7
1961197119811991 2001 2011
a
cd
Year
1980 2000 2020 2040 2060 2080 2100
4.0
1.5
2.0
2.5
3.0
3.5
Direct radiative forcing (W m
–2
)
bContinued
2008
emissions
80% cut in
non-CO2
80% cut in
CO2
80% cut in all
greenhouse gases
Figure 1 | Compound- and sector-specific emissions of greenhouse gases, associated radiative forcing
and global ruminant numbers over the past 50 years. a, Estimates of direct radiative forcing in 2008 for
CO2 and non-CO2 greenhouse gases from anthropogenic sources. b, Projections of radiative forcing in
four dierent scenarios: constant future emissions at 2008 levels (red); 80% reduction in only non-CO2
emissions (orange), 80% reduction in only CO2 emissions (blue), and 80% reductions in both non-CO2
and CO2 emissions (green). c, Estimated annual anthropogenic emissions from major sources of methane
in recent years. Error bars represent 1 standard deviation. d, Global ruminant numbers from 1961 to 2011.
Data for a–c from ref.1, d from ref.2.
© 2014 Macmillan Publishers Limited. All rights reserved

4 NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange
opinion & comment
International climate negotiators
can take steps to reduce greenhouse gas
emissions from livestock as well as from the
burning of fossil fuels. So far, global climate
policy instruments have mainly focused on
engineering improved industrial processes,
energy eciency and investments in
alternative energy generation technologies,
because sustainability has been
predominantly interpreted as technological
progress rather than changed patterns of
human behaviour6. Continued growth of
ruminant meat consumption will represent
a major obstacle for reaching ambitious
climate change targets. e substantial
environmental and climate costs of
increased meat consumption have been
recognized by the United Nations Food
and Agriculture Organization4. However,
mitigation of greenhouse gas emissions
from ruminants has not received adequate
attention in negotiations under the United
Nations Framework Convention on
Climate Change13. Meeting documents
show that activities to reduce emissions
from ruminants and agriculture in general,
and in negotiations on land use, land-use
change and forestry and reducing emission
from deforestation and forest degradation
have been disproportionately slow13. e
land-use accounting under the Kyoto
Protocol provides insucient coverage
of land-based emissions considering
their large contributions to greenhouse
gas uxes. e Kyoto Protocol also only
covers industrialized countries, so it misses
some of the largest emerging ruminant
producers. Further, under Articles 3.3
and 3.4 of the Kyoto Protocol, emission
reduction commitments for cropland and
grazing land management are optional in
many situations14.
e above-presented evidence calls
for a more comprehensive approach to
accounting in the Agriculture, Forestry
and Other Land Use sector, following
the lead of those countries requesting
mandatory accounting for land-based
emissions, including cropland and grazing
land sectors14. Progress would be facilitated
if emissions resulting from ruminant
livestock production are placed on the
agenda of forthcoming global climate
meetings such as the annual sessions of the
Conference of the Parties. Current national
policies on mitigating climate change could
also be revised to curtail emissions from
ruminant livestock in both developed and
developing countries.
Because the Earth’s climate may be
near tipping points to major change, the
need to act is increasingly pressing15,16.
Lowering peak climate forcing quickly
with ruminant and CH4 reductions would
lessen the likelihood of irreversibly
crossing such tipping points into a new
climatic state1. Reducing the numbers of
ruminants will be a dicult and complex
task, both politically and socially. However,
decreasing ruminants should be considered
alongside our grand challenge of
signicantly reducing the world’s reliance
on fossil fuel combustion. Only with the
recognition of the urgency of this issue
and the political will to commit resources
to comprehensively mitigate both CO2 and
non-CO2 greenhouse gas emissions will
meaningful progress be made on climate
change. For an eective and rapid response,
we need to increase awareness among the
public and policymakers that what we
choose to eat has important consequences
for climate change. ❐
William J.Ripple1*, Pete Smith2, Helmut Haberl3,4,
Stephen A.Montzka5, Clive McAlpine6 and
Douglas H.Boucher7 are at 1Department of Forest
Ecosystems and Society, Oregon State University,
Corvallis, Oregon, USA, 2Scottish Food Security
Alliance-Crops and Institute of Biological and
Environmental Sciences, University of Aberdeen,
Aberdeen AB24 3UU, UK, 3Institute of Social
Ecology Vienna, Alpen-Adria Universität
Klagenfurt, Wien, Graz, Schottenfeldgasse 29,
1070 Vienna, Austria, 4Humboldt-Universität
zu Berlin, Integrative Research Institute on
Transformations of Human-Environment Systems
(IRI THESys), Friedrichstraβe 191,D-10117 Berlin,
Ge rmany, 5National Oceanic and Atmospheric
Administration, Earth System Research Laboratory,
Boulder, Colorado 80305, USA, 6e University of
Queensland, School of Geography, Planning and
Environmental Management, Brisbane, Queensland
4072, Australia, 7Tropical Forest and Climate
Initiative, Union of Concerned Scientists,1825K
Street, NW Suite 800, Washington DC 20006, USA.
*e-mail: bill.ripple@oregonstate.edu
References
1. Montzka, S.A., Dlugokencky, E.J. & Butler, J.H. Nature
476, 43–50 (2011).
2. FAO ST AT (FAO, accessed 12 August 2013);
http://go.nature.com/Z23f7E
3. Gerber, P.J. et al. Tackling Climate Ch ange rough Livestock —
A Global Assessment of Emissions and Mitigation Oppor tunities
(FAO, 2013).
4. Steinfeld, H. et al. Livestock’s Long Shadow: Environmental Issues
and Options (FAO, 2006).
90
80
70
60
50
40
30
20
10
0
Beef (extensive)
Sheep
Beef meadow systems
Beef (intensive)
Seafood (fisheries)
Pork
Seafood (aquaculture)
Poultry
Eggs
Meat substitutes
(vegetal)
Pulses (beans, peas, soy)
Ruminant
Other
Carbon equivalent footprint
(kg CO2e per kg product)
Figure 2 | Average carbon equivalent footprint of protein-rich solid foods per kilogram of product from a
global meta-analysis of life-cycle assessment studies. Extensive beef involves cattle grazing across large
pastoral systems, whereas intensive beef typically involves feedlots. Meat substitutes are also known
as meat analogues, which are high-protein plant products that have aesthetic qualities (such as flavour,
texture, appearance) of specific types of meat. Error bars represent standard errors. Data from ref.17.
© 2014 Macmillan Publishers Limited. All rights reserved

NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange 5
opinion & comment
5. Smith, P. et al. Glob. Change Biol. 19, 2285–2302 (2013).
6. McAlpine, C.A., Etter, A., Fearnside, P.M., Seabrook, L. &
Lawrence, W.F. Glob. Environ. Change. 19, 21–33 (2009).
7. Beschta, R.L. et al. Environ. Manage. 51, 474–491(2012).
8. American Dietetic Association J. Am. Dietetic Assoc.
109, 1266–1282 (2009).
9. Fraser, G.E. Am. J. Clin. Nutr. 89(supplement), 1607S–1612S (2009).
10. Wirsenius, S., Hedenus, F. & Mohlin, K. Climatic Change
108, 159–184 (2011).
11. Schmidinger, K. & Stehfest, E. Int. J. Life Cycle Assess.
7, 962–972 (2012).
12. Popp, A. et al. Glob. Environ. Change 20, 451–462 (2010).
13. http://unfccc.int
14. Views on Land Use, Land-use Change and Forestry Issues Referred
to in Decision 2/CMP.7, Paragraphs 5-7. Submissions from Parties
and Admitted Observer Organizations 12–18 (SBSTA, UNFCCC,
2013); http://go.nature.com/hLAtTN
15. Lenton, T.M. Ambi o 41, 10–22 (2012).
16. Whiteman, G., Hope, C. & Wadhams, P. Nature
499, 401–403 (2013).
17. Nijdam, D., Rood, T. & Westhoek, H. Food Policy 37, 760–770 (2012).
Acknowledgements
We thank R. Lamplugh, B. Kauman, E. Stehfest and
R. Comforto for comments on an early dra of this
paper. W.R. was an Oregon State University L.L. Stewart
faculty scholar during this project. P.S. is a Royal
Society-Wolfson Research Merit Award holder. H.H.
gratefully acknowledges research funding from EU-FP7
(Volante, grant no. 265104) and the Austrian Science
Funds (project no. P20812-G11). S.A.M. acknowledges
the support of the NOAA’s Climate Program Oce and
its Atmospheric Chemistry, Carbon Cycle and Climate
Program. C.M. is supported by the Australian Research
Council (FT100100338). D.B. thanks the Climate and Land
Use Alliance for its support of the Union of Concerned
Scientists’ Tropical Forest and Climate Initiative.
COMMENTARY:
Social learning and sustainable
development
Patti Kristjanson, Blane Harvey, Marissa Van Epp and Philip K. Thornton
To understand what social learning approaches can oer the sciences of adaptation and mitigation, we
need to assemble an appropriate evidence base.
Research-for-development institutions
such as the Food and Agriculture
Organization (FAO) of the UN,
CGIAR and their partners are under
mounting external pressure from donors
to link knowledge to actions that achieve
substantive, long-lasting and demonstrable
development outcomes1. If research is
genuinely to result in benecial changes in
behaviour, policies and institutions, research
outputs need to be much better informed
by and engaged with the processes through
which individuals, communities and
societies learn and adapt their behaviour
in the face of change2,3. Social learning
approaches may be able to contribute
substantially to this aim4. Denitions vary,
but in a nutshell social learning approaches
facilitate knowledge sharing, joint learning
and knowledge co-creation between diverse
stakeholders around a shared purpose,
taking learning and behavioural change
beyond the individual to networks and
systems. rough an iterative process of
working together — engaged in interactive
dialogue, exchange, learning, action,
reection and continuing partnership — new
shared ways of gaining knowledge emerge
that lead to changes in practice5. As such,
social learning builds on well-established
traditions from participatory development,
but puts learning and collective change at
the centre of engagement. Social learning
can provide a way to address complex
socio-ecological (so-called wicked)
problems by integrating diverse knowledge
and value systems at many dierent levels
and through dierent learning cycles.
From theory to practice
As a concept, social learning is appealing.
But how can we implement it as eectively
and eciently as possible? In practice, it
takes many dierent forms and can be
used to eect dierent types of change.
Some examples of innovative sustainable
agricultural development projects and
programmes that are taking this approach
are shown in Table1. ese illustrate a
range of scales at which social learning
and change are happening, from the
individual to the community to networks
and systems. e range of outcomes from
these projects is equally wide, from changes
in the way farmers go about their business
to new agricultural input distribution
systems to the creation of new institutions
and the empowerment of national
agricultural planners.
On the face of it, social learning
approaches should be able to
contribute to smarter, more eective
research-for-development institutions in
terms of performance and governance, and
also help them to achieve more sustainable
results, measured as development
outcomes6. We also know that iterative
learning processes are perceived to be
a critical component of adapting to
environmental change, and that there is
an absence of learning tools that can be
applied in contexts where uncertainty is
high7. But at the moment, we have only
limited evidence on the impact of social
learning approaches on tangible development
outcomes, and not much is known about
the costs of social learning approaches in
comparison with more traditional, linear
practices8. ere has been only limited
eort put into evaluating social learning
methods beyond one-o case studies and
post hoc or appreciative reections9,10.
Larger-scale reviews of social learning
have thus far focused on its framings and
methodologies more than on its ultimate
impacts. Scientists are particularly concerned
with the transaction costs that they perceive
to be high (for example, the amount of time
spent dealing with ‘messy partnerships’) and
a limited ability to replicate and scale up
results more broadly.
A framework for gathering evidence
In view of the limitations of the current
evidence base and calls for greater empirical
rigour in evaluating social learning11, we
are embarking on a systematic evidence-
gathering eort, using a common evaluative
framework to track new initiatives from
© 2014 Macmillan Publishers Limited. All rights reserved