The World Bank’s approach to increasing the
vulnerability of small coffee producers
Sasha C. Breger Bush1
Colorado College, Colorado
Springs, CO, USA
February 2010
BWPI Working Paper 113
Brooks World Poverty Institute
ISBN : 978-1-907247-12-5
Creating and sharing knowledge to help end poverty
This essay critically engages the World Bank’s recent experiments in providing marketbased price risk management for coffee farmers. Using the case of Mexico and the
recent 1998–2002 coffee crisis, I argue that such advocacy of farm-level use of
derivatives markets entails large direct and indirect costs for coffee farmer wellbeing.
This is especially so for smallholders. Not only might hedging with derivatives further
destabilise and reduce producer incomes, but the opportunity cost of the Bank’s
advocacy, in terms of foregone risk management alternatives, is also problematic. I
conclude with a discussion of several risk management alternatives that may better
support small coffee producers facing volatile commodity prices.
Keywords: coffee, agriculture, World Bank, development, derivatives, risk
Sasha C. Breger Bush holds a PhD in International Studies from the University of
Denver, and is currently a Visiting Assistant Professor in the Department of Economics
and Business at the Colorado College in Colorado Springs, CO, USA.
Coffee prices are falling. Again. Coffee farmers in over 50 developing countries are
currently faced with the prospect of financial ruin in what may be the third coffee price
crisis in 20 years. Coffee future prices are down by almost 30 percent from their March
2008 highs. The International Coffee Organization (ICO) reports that cash prices in June
2009 were down almost 15 percent over the previous month as the prospect of a
Brazilian frost appeared more remote. There are continued worries that the ongoing
financial crisis will stymie coffee demand in the rich countries, putting further downward
pressure on prices.
In the meantime, the World Bank continues to offer up its ill-conceived set of
recommendations regarding ‘market-based price risk management’. This euphemistic
phrase refers to the Bank’s efforts to link developing country farmers to derivatives
markets, so that they might hedge their output and secure their incomes in the face of
commodity price volatility and decline. By hedging with futures and options, coffee
farmers can procure some much-needed price insurance, all the while avoiding the
pitfalls of government intervention in commodity markets, or so the Bank has argued.
In fact, the Bank’s recommendations risk further destabilising the incomes and
livelihoods of coffee producers, especially small ones, and increasing their vulnerability
to market shocks. Futures hedging can make coffee farmer incomes wildly unstable,
undermining the purported stabilising benefits of such strategies. Potentially greater
financial insecurity is thus one of the direct costs to farmers of the Bank’s orientation to
the commodity price problem. Other direct costs may include those created via the
perpetuation of chronic oversupply in coffee markets. Conditions of market oversupply
contribute to low and volatile coffee prices and may be aggravated by incentives
generated in the hedging process. Small producers, least able to manage such
fluctuations given their meagre resources, are usually those most negatively affected by
Perhaps outweighing such direct costs are the opportunity costs of the Bank’s work in
this area. Continued, misguided recommendations of derivatives-based schemes have
directed donor time, energy and resources away from alternative schemes and policies
that actually work. What opportunities has the Bank foreclosed in its pursuit of marketbased price risk management? One of the more important consequences of this pursuit
is that today, with coffee prices in decline, there is no working system of social support in
place that can help producers successfully weather this shock. Here again, it tends to be
small producers that suffer disproportionately in the absence of effective social
protection policies.
In this paper I present empirical evidence regarding both the direct and opportunity costs
of this World Bank’s experimentation with the wellbeing of small coffee farmers. I begin
in the first section with a discussion of the World Bank’s commodity price risk
management work to the present. In the second section, I discuss price risk and small
farmer vulnerability thereto. In the third section, I present evidence on the destabilising
impact of futures hedging on farmer incomes. In the fourth section, I present data that
suggest several alternative arrangements that have not been pursued by the Bank but
that may effectively offer real support to coffee farmers facing market shocks. In the final
section, I conclude with a few summary remarks. Given space constraints, I use the
example of Mexico and the 1998–2002 coffee crisis to make various empirical and policy
points throughout.
World Bank derivatives market advocacy
Liberalisation of the global coffee market began with the collapse of the International
Coffee Agreement (ICA). The ICA, with its import and export quotas, had managed
global coffee supplies and prices between 1962 and 1989. From 1989 onward, and for a
variety of reasons, coffee-producing counties began to dismantle the domestic
institutions that had come to be legitimised on the basis of the Agreement. In Brazil, El
Salvador and Mexico, governments recognised in 1989 that ‘the suspension of
international quotas effectively ended the need to control the coffee market’ (Akiyama, in
Akiyama et al., 2001: 87). Marketing boards began to be eyed with suspicion by
producers in India, with the Coffee Board’s complete dismantlement achieved by 1996
as a result of this domestic opposition (Akiyama, in Akiyama et al., 2001). In Uganda,
Cameroon, Cote D’Ivoire and Madagascar, the price slump that followed the ICA’s
collapse made domestic commodity price stabilisation arrangements fiscally untenable.
As prices fell, demand for more stabilisation funding rose at the same time that
governments were collecting less revenue from coffee taxes. The early 1980s debt crisis
in the developing world, resulting in the so-called ‘lost decade’, had further cast shadows
upon such financially unsustainable programmes. Making matters worse, corruption
among marketing board officials was rampant. By the 1980s and early 1990s, producers
in some countries, like India, had begun to respond, demanding the dismantlement of
stabilisation arrangements. Snyder notes that in Mexico, ‘a powerful grassroots
movement of small coffee producers …mobilized against’ the state-owned Mexican
Coffee Institute (INMECAFE), hastening neoliberal reform (Snyder, 2001: 21).
World Bank structural adjustment programmes, particularly in Sub-Saharan African
countries (e.g. Uganda and Togo), also demanded the liberalisation of agricultural
markets. Structural adjustment places conditions on development loans from the World
Bank. ‘Conditions commonly applied to agricultural commodity subsectors included
eliminating or significantly reducing government intervention in determining prices,
dissolving parastatals [e.g. marketing boards] or stripping them of some powers, and
eliminating some agricultural import subsidies’ (Akiyama et al., 2001: 12). The World
Bank, as well as other international financial institutions like the International Monetary
Fund (IMF), during the 1980s and 1990s was abandoning its prior focus on public
interventionist economics, turning instead to policy that emphasised the central role of
unfettered markets. The incredible growth of the ‘Asian Tiger’ developing economies
(like South Korea, Taiwan, and Singapore) – growth allegedly based on a marketoriented set of reforms – legitimised this policy shift at the Bank and the IMF.
While commodity market liberalisation was ongoing, researchers, particularly at the
World Bank, began to recommend commodity price risk management policies that were
consistent with it. Identifying the types of price risk management arrangements that were
compatible with commodity market liberalisation began to occupy significant amounts of
research time. The World Bank’s International Task Force on Commodity Price Risk
Management (ITF), frequently working in conjunction with the Bank’s Commodity Risk
Management Group (CRMG), is now devoted to supporting the development of marketbased price and weather risk management arrangements and focuses largely on
commodity and weather derivatives markets. The ITF explains its mission: ‘With the
liberalization of many developing country economies, farmers are not insulated from
price risks…Markets that can enable producers and organizations to reduce price risk
exist…The ITF aims to increase the capacity of local organizations to access commodity
risk management markets’ (ITF, 2006b: 1–2). In this manner, the ITF recognises the
absence of effective price risk management for farmers in the liberalisation era, and
recommends that derivatives markets fill this outstanding gap.
To further their goals, the ITF and CRMG have been financing, administering and
monitoring pilot programmes to deliver market-based price risk management to small
farmers. From cooperative intermediation in Mexico and Uganda to local banking
products in El Salvador and Tanzania, the ITF has been steadily tinkering for about a
decade with the mechanisms for hooking small producers up to futures and options
markets. I have no figures regarding the expense on these efforts over the past ten
In the winter of 2008, the World Bank offered its ‘introductory course to the Market based
Risk Management for Development Distance Learning Program (MRMD)’ that sought to
‘introduce participants to a market-based risk management approach for correlated risks
affecting the agricultural sector, such as weather risk and commodity price risk’.
Designed for policymakers (and looking an awful lot like a sales pitch),
The course [sought] to address a number of questions. Why are traditional coping
mechanisms inadequate in addressing the costs and consequences of extreme
weather events and price shocks in lower income countries? What are the
advantages of market-based risk transfer mechanisms in comparison of traditional
shock mitigation measures? What are the appropriate market-based risk
management instruments and how do they work? How should these market
instruments be utilized? What is the appropriate division of public and private roles
in a country’s proactive risk management agenda? 1
It is not only the World Bank, however, that is actively recommending that policymakers
focus their efforts on derivatives as a means of managing farm-level risks. UNCTAD
recently noted that, ‘As early as the 1970s, UNCTAD studies recognized the potential for
market-based price risk management instruments in delivering welfare gains to
commodity sector participants’ (UNCTAD, 2006: ix). The FAO’s Committee on
Commodity Problems writes,
As a proactive approach for minimizing uncertainty, risk management tools are
increasingly replacing government support programs as an alternative for raising
price predictability and enhancing producer income stability…On a national front,
institutional development and legislative measures permitting derivatives trading
have paved the way for striking success in managing risks and reducing volatility in
many developing countries (FAO, 2007: 1).
The Common Fund for Commodities, an organisation within the framework of the UN,
has been funding projects in Eastern and Southern Africa to develop market-based
commodity price risk management arrangements.
Further, in the developing countries themselves, market-based price risk management
has also received growing attention. Brazil and India have established their own
commodity derivatives exchanges that offer coffee contracts. The governments of
Mexico, El Salvador and Brazil all operate programmes to provide futures and/or put
option contracts to farmers. In Uganda, Union Export Services, Ltd. (an umbrella
cooperative) has been negotiating over-the-counter (OTC) put options with international
dealers on behalf of their member cooperatives. And, in Tanzania, there is hope that
futures contracts might be provided to farmers via popular rural savings and credit
To reiterate, the aim of this paper is to question this policy gamble that is being made by
larger and larger segments of the development community, and to explain how it is that
such policies and programmes actually make more vulnerable those they purport to help.
Small coffee farmer vulnerability to price risk
Even if they have not always been the focus of the World Bank’s commodity price risk
work, small developing country farmers are increasingly taking centre stage in
discussions of market-based price risk management. As they should: small, poor farmers
are among those most vulnerable to market shocks and they tend to suffer
disproportionately when a shock occurs. Indeed, the International Labour Organization
(ILO) notes that, ‘In general, a poor person cannot insure to anything like the extent that
~pagePK:2524753~piPK:51421526~theSitePK:4607,00.html Accessed 5 August 2009.
a rich person can, and yet is almost certainly exposed to more risks, more uncertainty
and more adverse outcomes’ (ILO, 2004: 11).
With few or no alternatives available, poor farmers (who tend to be small) and their
families often resort to reducing consumption of basic goods and expenditures on health
and education in order to cope with shocks. Holzmann and Jorgenson (2000: 7) note
The absence of efficient market-based or government provided consumptionsmoothing instruments often results in the use of costly informal coping
mechanisms once the adverse income shock hits, such as pulling children out of
school, reducing nutritional intake, selling productive assets, or neglecting human
capital accumulation.
Sayer (2002: 10) relates the situation of many Ugandan coffee farmers during the 1998–
2002 coffee crisis who struggled ‘to buy essentials like sugar, soap, salt, kerosene, tools
and clothes … Secondary school enrollment has declined as tens of thousands of
children are sent home for lack of fees’. (Sayer. Such income shocks can also
undermine the ability to repay debts.
Expectation of a market shock, in addition to its actual occurrence, can lead to similar
welfare reductions. Morduch (1999: 187) explains: ‘Fear of risk can lead poor
households to forgo potentially valuable new technologies and profitable production
choices’. A study by Rosenzweig and Binswanger (in Morduch, 1999: 187–88) revealed
that the farm profits of poor Indian households exposed to the risk of monsoon dropped
by roughly 35 percent, but that wealthier farmers’ profits were not affected at all.
UNCTAD (2002: 6) agrees: ‘Uncertainty about future incomes makes it difficult for
farmers to make commitments with respect to future payment obligations (e.g. debt
repayment, investments in land, machines, equipment and school fees for children).’
Such ‘capital rationing’ involves a trade-off between lower, certain and more stable shortterm incomes on the one hand, and potentially higher but more uncertain incomes in the
long term on the other. The World Bank (2005: 2) puts it as follows: ‘[T]he poor are
forced to make production decisions using the objective of minimizing risk, rather than
maximizing profits, and thus they must forego more remunerative activities that could
provide means of escape from their poverty.’
Among other factors, farmers with alternative income sources, savings, sufficient
productive assets (especially land and livestock) and access to credit are not negatively
impacted to the same extent as are lower-income, undiversified and asset-poor farmers
with bad or no credit. This is why, UNCTAD (2002: 6) notes, ‘Smaller, poorer farmers are
more risk-averse than larger, richer ones’. The greater welfare costs of instability and
uncertainty for the poor resulted in the World Bank (2001) citing ‘vulnerability to risk’ as
one of the four dimensions of poverty in the 2000–01 World Development Report.
Looking specifically to Mexico, Davis’ study of ejido producers in Mexico since
liberalisation began in the 1980s reveals that wealthier producers were able to respond
to increased risk and uncertainty more effectively than poorer producers. Specifically,
producers with larger plots of land, greater levels of assets (like livestock) and more
‘human capital assets’ have been better able to adjust to the risks associated with
economic liberalisation than producers with less land, less human capital and few to no
assets. While the former group has invested in more cattle, as well as fruit and vegetable
production (which is less prone to low prices than corn, which was farmed almost
exclusively before liberalisation), poorer producers continue to struggle:
These households, with less land, livestock, migration and human capital assets,
have significantly lower household incomes than the households referred to above
that have been able to adjust successfully. Without access to adequate levels of
assets, these households will continue to struggle in the current austere
macroeconomic environment (Davis, 2000: 103).
Ninety percent of Mexico’s coffee is grown in four southern states of Veracruz, Chiapas,
Oaxaca and Puebla by small producers, many of whom are indigenous peoples. The ITF
(2002) reports that some 180,000 of Mexico’s 282,000 producers are indigenous, with
coffee being their main or only source of cash income. And, the small producers that
dominate the south tend to be very poor relative to the rest of the country’s inhabitants:
In 1996, the four main coffee producing states Chiapas, Oaxaca, Puebla and
Veracruz had a per capita gross internal product that was around 43 percent of
Mexico’s average for Chiapas and Oaxaca, and around 64 and 67 percent for
Veracruz and Puebla, respectively. (ITF, 2002: 6).
It should come as no surprise that it was Mexico’s smaller and poorer coffee farmers
that experienced severe hardship during the recent coffee crisis, while wealthier, larger
farmers fared much better. A recent study by Eakin and colleagues of small coffee
producers in Veracruz found that during the crisis almost 72 percent of producers
reduced their harvest, 62 percent ‘reduced use of purchased inputs’, almost 97 percent
were ‘negatively affected’ in their ‘ability to acquire basic goods’, and 80 percent were
less able to pay medical expenses (2006: 164). The study crucially finds that households
with ‘access to market and technical information, finance and having sufficient land with
which to diversify into alternative crops’ were better able to deal with the crisis (Eakin et
al., 2006: 168–89).
To sum up, it seems that the last thing that small, poor Mexican coffee producers may
want is to participate in yet another market that wreaks havoc on their lives and
Direct costs of the World Bank approach
The World Bank’s advocacy of market-based price risk management risks further
destabilising the incomes and livelihoods of small coffee producers, for at least two
reasons. First, in the short run, futures hedging can generate wild fluctuations in
producer incomes for those who are not sufficiently well capitalised to meet margin calls
with reserves. Second, in the medium to long run, futures hedging could in aggregate
put downward pressure on coffee prices if hedging provides incentives for increasing
individual production. Both of these costs are potentially more devastating for small, poor
producers, for reasons discussed above. Below, I elaborate on these direct costs.
It should be noted that there are many more direct costs to farmers that could be
mentioned, many of which I have discussed elsewhere.2 What I intend here is to briefly
discuss a couple of direct costs, so that it becomes clear that the World Bank’s approach
to commodity price risk management may very well undermine the wellbeing of those it
purports to help.
Margin calls and future price volatility
Prior to commencing trading, every investor in exchange-traded futures markets must
post an ‘initial margin’ that serves as a good faith deposit to cover any trading losses
sustained. After trading begins, all investors must maintain a minimum balance in their
trading accounts, called the ‘maintenance margin’. Market gains and losses are tallied
daily, i.e. ‘marked-to-market’, meaning that farmers will be required to post additional
margins daily if the market moves against them. Investors must maintain their margins in
order to keep their positions open.
Anderson and Danthine (1983: 249) remark upon the relationship between margin calls
and future price volatility:
the day to day fluctuations of the futures price will generate a series of random
cash flows over the period during which a futures position is held. This is due to the
requirement that futures accounts ‛mark to market’ daily.
Indeed, even if the general trend in future prices is downward (as would be the case
during a crisis), short hedgers (like coffee farmers) could still face serious, temporary
income losses, due to the day-to-day volatility of future prices. Put differently, a hedge
that is profitable over a period of a year or two may be very unprofitable from day to day
or month to month depending on market behaviour.
Breger-Bush (2009) considers many additional economic, political and social costs associated
with futures hedging and World Bank advocacy thereof for farmers, such as basis risk and other
contributors to farmer income insecurity, risks of political backlash to policy endeavours that fail to
help those most in need, and the poor track record of derivatives market intermediaries in the
developing world.
Margin requirements represent an important mechanism for the pass-through of future
price volatility to farmer incomes, consumption and wellbeing. Margin calls can be
devastating, even for large traders. A US farmer recently interviewed in The New York
Times stated:
The nightmare scenario is when you have to make margin and you’re looking out
your back door and seeing, maybe, a crop problem. Everybody has a story about a
guy they know getting blown out of his hedge by unmet margin calls (Henriques,
Small producers are more likely to be ‘blown out of a hedge’ even without crop failure,
given their low income levels and few to no cash reserves or assets. For small farmers,
margin calls may eat away at income previously designated for consumption, education,
medical expenses and other necessary items. They may also result in the accumulation
of significant debts if producers can access credit markets.
Figure 1 presents a graph of daily settlement prices for Arabica coffee on the New York
Board of Trade (NYBOT, now merged with the Intercontinental Exchange, ICE) during
the first crop season of the 1998–2002 coffee crisis (that is, the 1998–99 crop season).
As is clear from the graph, there is significant inter-day price volatility on this exchange,
despite the general, downward trend.
Figure 1. Daily settlement prices for Arabica coffee, New York Board of Trade,
1998–99 crop season.
A short hedge maintained over the course of this crop season would have been
profitable. But it is unlikely that a small, poorly capitalised farmer could have lasted long
enough to realise these profits. The volatility of future prices over this period means that
a short hedger would have had to make frequent margin calls, with often substantial
sums of cash.3
To illustrate, Figure 2 presents a chart that details daily gains and losses for a one yearlong hedge. As Mexico is being used as an example throughout, the data presented are
based upon future prices from NYBOT, that exchange on which a Mexican coffee farmer
would most likely trade.4
Let us assume a hypothetical Mexican coffee farmer who grows enough coffee to sell
one futures contract on the New York Board of Trade (37,500 lbs; in reality, a Mexican
coffee farm of average size and average yields would produce only roughly seven
percent of the coffee required to sell one contract), has certain output levels, can afford
the initial margin and other trading fees, and has sufficient information and knowledge to
hedge effectively. Let us also assume that this farmer decided to begin hedging in 1998,
as coffee prices began their precipitous fall to their lowest real levels in over 100 years.
Last, let us assume that the farmer undertakes the most simple of hedging strategies –
what I call the plain annual hedge. The plain annual hedge involves the farmer selling
her crop forward one season, using only one contract, i.e. in April 1998 the farmer sells
one March 1999 Arabica coffee futures contract. In theory, a short sale such as this
effectively locks in prices for the season for a crop that has yet to be harvested. The
daily gains and losses for this strategy for the 1998–99 crop season appear in Figure 2.
‘Now, wait just one moment!’, the reader may object. ‘Is it not possible for the farmer to save
some hedging profits from the earlier part of the hedge to use to post margins in later segments?’
Of course this is possible, if sufficient profits are made early on to cover later losses. This is
sometimes, but not always the case. Even if there were sufficient early profits to make saving for
later margin calls workable, I think it is unlikely in practice for two reasons.
First, predicting the path of future prices is close to impossible and, therefore, deciding how much
of the gains from the hedge to save, is fraught with difficulty. As will be seen in the data presented
below, the daily gains and losses from a hedge over this crop season are highly irregular.
Teweles and Jones (1999: 107) note that future prices move very erratically, if not randomly,
making prediction a rather futile effort.
Second, it strikes me as unreasonable to entertain at much length the idea that small, poor
producers would save such substantial portions of their hedging profits in order to post future
margins. I do not mean to imply that the poor cannot save. The poor can and do save and invest.
What I do mean to say is that few small coffee farmers could be reasonably expected by
policymakers to save enough of these gains to cover future margins (which are substantial). By
my calculations, a Mexican coffee farmer of average size, realising average yields, with the
average number of family members for rural communities in Mexico, would have earned $1.31
(2007 international dollars) per family member per day over the course of the 1998–2002 coffee
It stands to reason that farmers in such a position would probably spend a large portion of the
gains from hedging immediately upon receipt or close thereto.
There is a domestic derivatives exchange in Mexico; however, it does not offer coffee contracts.
Figure 2. Daily gains and losses for a plain annual hedge, crop season 1998–99.
Looking at Figure 2, the gains from the hedge peak at US$3656 on 11 August 1998
(following a fall in prices of almost 10 US cents from the previous day) and the losses
reach a trough at US$2,625 on 3 November 1998 (following a price rise of seven US
cents from the previous day), with this range a reflection of the volatility of future prices.
Further, in this case the farmer could not have realistically saved her profits earlier on to
pay later margin calls. Between 9 and 16 April 1998 (early in the hedge), losses
amounting to over US$2,900 would have been accumulated, with no hedging profits
remaining to cover any portion of this amount, even if all previously earned profits had
been saved and dedicated to making margin calls (i.e. this is a net loss).
The losses sustained periodically over the course of the hedge are of such a magnitude,
relative to the incomes of small producers, that it seems unreasonable to think that
margin calls could be met without serious hardship. Even if we scale down the losses to
the size of the average Mexican coffee farm,5 the accumulated losses in the second and
third weeks of April 1998 (see above) amount to over US$200. This remains a very large
sum to produce in a few hours’ time for a family whose members live at the extreme
poverty line.6
To sum up, margin requirements are an important mechanism for the transmission of
future price volatility to farmer and family wellbeing. Future prices are sufficiently volatile
so as to generate hedging losses of a magnitude that is likely to be seriously disruptive
of the farmer’s financial position. Not only are the losses large, but they are also
Recall that the hedging calculations assumed the sale of one contract, specifying 37, 500lbs of
coffee. The average Mexican coffee farm is 2.67 acres in size, which, holding average yields
constant, is about seven percent of the size of the farm that would be required to grow 37,500 lbs
of coffee. In other words, I am proceeding for the moment as if a small coffee farmer sold a share
of a futures contract proportional to her output.
See footnote 3.
frequent. In this context, hedging appears to be a destabilising economic force at the
farm level.
It is important to note that margin calls are more destabilising for small producers than
for other actors in the global coffee economy. While large coffee conglomerates, like
Kraft and Nestlé, have substantial cash reserves (or other liquid assets) to finance
margin calls that can be drawn upon without impacting day-to-day operations (see
Newman, 2009; Ponte, 2002), small producers do not. To be fair, the World Bank has to
some extent recognised that futures trading requires significant capitalisation at the
individual level, in order for it not to be disruptive. Recent efforts to hook farmers up to
futures markets via intermediary institutions are motivated in part by the desire to
overcome such obstacles to individual trading.
However, some concern is warranted here, even though intermediation efforts have
been largely unsuccessful at this juncture (Breger-Bush, 2009). The intermediary
solution that the Bank has most favoured is a model that portends even more
vulnerability for small producers. In various documents, World Bank researchers note the
benefits of using local banks to pass on or otherwise sell price insurance to farmers, so
that farmers avoid having to trade directly (CRMG, 2005a and b; ITF, 2006b; Tiffen and
Fernandez, 2005). Local banks, it is argued, have a vested interest in passing on price
insurance, in order to avoid farmer loan defaults during price crises. Yet, these are forprofit models in which farmers would have to pay a hefty fee to obtain this service. In a
similar vein, Tiffen and Fernandez (2005) speak favourably about a programme in El
Salvador, where local banks have extended ‘dedicated lines of credit’ to producers in
order to help them finance margin calls. I tend to be sceptical of recommendations that
small farmers take on more debt. Indeed, the World Bank has not been sufficiently wary
of the fact that various models of futures market intermediation may also serve to
increase farmer vulnerability to price shocks.
Hedging and incentives for overproduction
Moving away from short-run fluctuations, the World Bank’s policy agenda also risks
making small farmers more vulnerable in the long term. Ironically, the more headway the
Bank makes in promoting derivatives, the greater is the likelihood of this particular brand
of long-run vulnerability. Put simply, if lots of producers start hedging, and if the price
protection afforded by hedging boosts farmer incentives to increase output, global
supplies will rise and prices will fall. Articulated back in the 1930s by Keynes in the
commodity price context, this idea has been recently reworked into a theoretical model
of the costs of aggregate futures hedging of agricultural commodity price risk. I explain in
more detail below.
Writing in the wake of the Great Depression, Keynes recognised that falling demand for
agricultural commodities (like that which had occurred earlier in the decade) spelled
trouble for commodity prices. He argued that world commodity markets were like a
‘Tragedy of the Commons’, in which rational individuals pursuing their own self-interest
collectively effaced or destroyed something valuable to the collective. Keynes noted that
under such conditions, individually, there was a disincentive for primary commodity
producers to withhold any of their stocks from the world market. Prices were falling,
storage was costly, and so was postponing repayment of agricultural loans. But, as each
individual producer released stocks onto the market, the price of a given commodity
would decline, generating a suboptimal outcome for producers as a whole. He writes: ‘If
demand fluctuates, a divergence immediately ensues between the general interest and
the course of action in respect of stocks which is most advantageous for each
competitive enterprise acting independently’ (Keynes, 1938: 449).
The idea of prices as some kind of collective good (a ‘common resource’ in economics
textbooks) also informs Lence’s recent (2009) study of the costs of futures hedging.
Taking on economic theories of hedging that offer models based on a single
representative agent, Lence instead assumes two such agents, one who hedges and
one who does not, in order to get at the effects on welfare of futures hedging in
aggregate. The results of his simulation, which uses data on rubber production, indicates
that ‘nonadopters’ (those who do not hedge) sustain welfare losses, irrespective of how
large the portion of ‘adopters’ (those who do hedge) is among all farmers. Even the
adopters sustain welfare losses when they represent a large enough share of total
Overall, our findings suggest that there is no basis for advocating the use of futures
as a means to improve commodity producers’ well-being. The reasons for this
assertion are that producers as a whole lose from making futures available, and
more important, that futures availability reduces welfare even for adopters, provided
the market share of the latter is sufficiently large (Lence, 2009: 167).
In fact, the only group in the model that systematically gains from hedging in aggregate
is consumers. In the coffee context this is very disturbing. Wholesale coffee consumers
are largely big multinational firms from the global North and retail consumers are
generally located in the rich countries (Ponte, 2002). Lence (2009, 167) ultimately
concludes as follows about this distributive issue:
the push by international organizations (e.g. ITF) to improve the availability of
futures among producers may ultimately enhance the lot of consumers in
developed countries, while reducing the welfare of producers in developing
countries. To many, this regressive redistributional outcome is likely to be both
surprising and undesirable.
Opportunity costs of the World Bank approach
Moving on to what may be the saddest part of this story, the World Bank’s continued
advocacy and administration of derivatives-based price risk management schemes
means that limited time, energy and resources are being directed away from solutions
that may actually help small coffee farmers. (This particular opportunity cost seems
especially noteworthy today, as coffee prices continue to slide.) Using Mexico as an
example, below I outline and provide empirical support for two particular alternatives that
in this case offer substantive alternatives to futures hedging: supply management and
Fairtrade. The point I wish to make with this example is that there are probably many
viable alternatives to futures in each country case, substantive alternatives that provide
real support for farmers.
Beginning with supply management, I note, firstly, that in no way do I intend here to
overly romanticise the International Coffee Agreement or those marketing boards that
complemented it. The ICA was fraught with problems and there is a large literature that
has examined its various provisions, successes and failings (e.g. Bates, 1997; Talbot,
2004; Bohman, Jarvis and Barichello, 1996; Gilbert, 2005). Marketing boards were often
instruments of political oppression and corruption, and a means of taxing agriculture to
feed industry or pet political projects (e.g. the Ugandan marketing board under Idi Amin).
But, there is evidence that supports the conclusion that supply management, of some
kind, would probably serve to raise coffee prices and support and insulate farmer
incomes in the face of market shocks. While a new ICA-like quota system seems
politically unlikely in today’s political climate,7 new programmes that perhaps combine
initiatives to reduce global coffee supplies with funding and technical support for on- and
off-farm diversification may be viable. Part of the reason that supply management and/or
reduction may have some merit in the coffee context is that coffee has consistently been
marked by ‘chronic oversupply’, with global production well outpacing global demand
over time. And, like most commodities, the long-term trend in coffee prices is downward,
suggesting that global supply reductions could have beneficial short- and long-run
Indeed, there is little disagreement on the relative success of the ICA in stabilising and
augmenting world prices. Akiyama, Tamassa and Varangis find that, ‘Overall, the quota
system had a stabilizing effect on world coffee prices’ (1990 in Bates, 1997: 18). Jorge
Cardenas, Chairman of the World Coffee Conference, states in a 2001 speech:
‘International coffee prices have shown a fluctuation of than 50 percent annually in
recent years whereas during periods when the market was regulated prices fluctuated
between 10-15 percent around their medium-term trend’ (Cardenas, 2001: 2). Gilbert, a
notable exception in this context, contends that price volatility did not increase following
Especially because the original ICA had the support of big coffee consumers like the US and
Western Europe, with these countries opposing quotas since 1989. Indeed, every new ICA since
1989 has been stripped of its economic management clauses and has set forth an agenda of
mere promotion and research. As a side note, it seems that the US is the real obstacle here, as it
was the US who spoke out most adamantly against the quotas as the 1989 Agreement was being
negotiated (Wild, 2004; Talbot, 2004).
the collapse of the ICA. However, he does note that while, by his measure, world price
volatility stayed more or less constant, the elimination of state stabilisation arrangements
has meant that volatility is passed through to producers, rather than being absorbed by
the state (Gilbert, 2005: 6). From the perspective of the farmer, then, price volatility has
increased since the fall of the ICA.
Likewise, studies by Hermann et al. (1990) and Franz and Vogelvang (1991) (in Bates,
1997: 18–19) found that prices during the pre-1989 ICA period were higher than they
would have been without quotas. A 1993 study by Herman, Burger and Smit argues that
coffee prices under the ICA regime were 24–30 percent higher than the estimated
market-clearing price without the ICA (in Gibbon, 2005: 19). Gilbert calculates average
coffee prices during the interventionist and post-ICA eras, excluding the ‘anomalous’
years of high prices between 1976 and 1979 when ICA quotas were suspended (2005:
3). He finds that the average price of coffee, in terms of 2000 prices, between 1960 and
1988 was 178.3 cents per pound (in USD); the average price between 1990 and 2005,
again in terms of 2000 prices, was a mere 87.5 cents per pound (USD), about 49
percent of prices in the prior period.
Crucially, none of these studies looked at the prices that growers actually received
during the ICA era, instead opting to study world coffee prices. This is an important
oversight, especially considering that farmgate prices tended to be lower than world
prices under the ICA. Marketing boards and like institutions that managed exports,
production and prices domestically often drove a wedge between export and farmgate
prices, diverting these revenues into government coffers (Thiele et al., 2006). In order to
understand precisely the conditions facing farmers under the ICA, I have performed a
few calculations that allow for a comparison of the income of farmers under the ICA to
that obtained via futures hedging. I present my methods and results below.
To figure the real annual income earned by Mexican coffee farmers under the ICA, I
used historical ‘prices paid to growers’ data from the International Coffee Organization
(ICO) from 1976 to 1989 (these are the years for which data is available), and then
converted the figures into 2007 Mexican pesos. I then devised six different hedging
strategies to see how each would have performed for a Mexican coffee farmer during the
recent 1998–2002 coffee crisis, and converted my results into 2007 Mexican pesos.8
Future price data was obtained from NYBOT. The hedging strategy that resulted in the
largest gains was a multi-year rollover strategy that was ongoing for four crop seasons,
1998–99 to 2001–02.9 I then compared the real 2007 income earned via this hedge to
the real 2007 income earned under the ICA.
Please note that this only serves to bias my results in favour of futures hedging. Indeed,
hedging can be wildly profitable during periods of falling prices.
Rollover hedging involves the simultaneous purchase and sale of futures contracts such that
investors are always trading in the nearby contract (the market for which is the most liquid). A
multi-year rollover involves hedging multiple years of output all at once. For a four-year multiyear
By my estimation, real income for a Mexican Arabica producer with a farm of average
size would have seen his income range from a peak of just over $115,000 pesos in 1977
to a trough of just over $53,500 pesos in 1989. The average, real annual income over
the 1976–89 ICA period is $75,420 pesos.
Relative to the income earned from hedging and coffee sales during the 1998–2002
coffee crisis by a Mexican Arabica farmer of average size, the ICA performed rather
well.10 The average, real annual income for the average sized Mexican farmer who used
the most profitable strategy above was only $51,757 pesos – this is over $23,000 pesos
less than the average, real annual income of this same sized farmer during the 1976–89
period. The ICA managed to maintain an average income over 45 percent higher than
did coffee sales combined with the most profitable hedging strategy that I devised during
the coffee crisis.
Fairtrade too seems a vibrant alternative to futures hedging in Mexico, with the country
already producing a large share of global Fairtrade coffee supplies. Fairtrade11 is an
independent, private, non-profit certification and labelling initiative that incorporates
principles of economic, social and environmental justice. As such, scholars have noted
that Fairtrade networks are ‘counter-hegemonic’ and represent ‘globalization from below’
(Raynolds et al., 2006: 180). The Fairtrade Labelling Organizations International (FLO)
notes that Fairtrade allows farmers ‘to escape from poverty and provide themselves and
their families with a decent standard of living’.
Fairtrade organisations link farmers directly with roasters and other consumers, who pay
a fixed price per pound of coffee. The reduction in the number of middlemen along the
path from field to cup is one way in which Fairtrade coffee networks are able to maintain
higher prices for farmers. As of June 2008, the Fairtrade price was US$1.26/lb for
washed Arabica beans,12 and US$1.01/lb for washed Robusta beans.13 By comparison,
the reader may want to note, that in the recent coffee crisis Arabica prices paid to
growers in Mexico reached about US$0.25/lb, suggesting the significant level of income
rollover, a trader would hedge four seasons worth of output the firstt year, three seasons worth
the second year, two in the third, and one in the last year. In theory, the gains from such a
strategy are pretty close to equal to those that would have been earned if futures markets
extended four years into the future (which they do not: most commodity markets go about 18
months forward, with metals markets going somewhat further).
The ICA was complemented domestically in Mexico by INMECAFE, the Mexican Coffee
Fairtrade should not be confused with ‘Fair Trade’, with the latter representing a broader
movement for social, economic and environmental justice that does not always conform to the
strict standards of Fairtrade. Further, many corporations, like Starbucks, have implemented
corporate social responsibility (CSR) strategies, the standards of which are lower than those
utilised by Fairtrade.
ws[tt_news]=32 Accessed 26 May 2008.
Accessed 26 May 2008.
support that the Fairtrade premium provides producers during crises. Calo and Wise’s
calculations about the sufficiency of the Fairtrade premium to cover farmer costs suggest
something similar: ‘Fair Trade’s fixed and higher premiums for both conventional and
organic coffee could bring all producers to profitability’ (2005: 26). In addition to the
Fairtrade price, producers receive a US$0.10/lb premium dedicated to social and
community investment. Such social investments help producers gain access to better
financing, community-owned and operated transportation and communications systems,
technical assistance, and health and education services.
Currently, the FLO has 20 member labelling organisations in 21 countries that certify the
following commodities: coffee, tea, rice, bananas, mangoes, cocoa, cotton, sugar,
honey, fruit juice, nuts, fresh fruit, quinoa, herbs, spices and wine.14 Almost one-and-ahalf million farmers, and workers, represented by over 600 producer organisations in 51
countries currently reap the benefits of Fairtrade.15 In the coffee context, however,
Fairtrade sales are only a small proportion of total global sales. In 2005, Fairtrade sales
were less than one percent of total global coffee production, by volume (although by
value Fairtrade’s share would be somewhat higher). Baffes and colleagues note that
‘differentiated’ coffee sales, i.e. trade outside of ‘traditional channels’ including fairly
traded, organic and other ‘niche’ coffees, accounted for roughly six to eight percent of
global consumption. They also report that, of the 240,000 60 kg bags of Fairtrade coffee
consumed in 2001, over 40 percent of this amount was consumed in Germany and the
Netherlands (Baffes, Lewin and Varangis, 2005: 207).
Mexico is the world leader in Fairtrade and Fairtrade-organic coffee exports. By 2005,
roughly 15 to 20 percent of Mexican coffee farmers were involved in Fairtrade
(representing about ten percent of Mexican coffee exports), Fairtrade-organic or were
transitioning over from traditional production techniques (Calo and Wise, 2005: 16). High
levels of coffee farmer ‘social organisation’ (i.e. well-organised, long-standing
cooperatives) as well as traditional farming practices that are already ‘passively organic’
are among those factors that have enabled Mexico’s relative success in penetrating
these niche markets (Calo and Wise, 2005: 16–18). This indicates that there is real
potential in Mexico to incorporate greater numbers of producers into Fairtrade
production. Small producers might especially benefit from public and donor support in
meeting the FLO’s certification requirements, given that most Mexican Fairtrade
production is already done by small farmer cooperatives. Raynolds, Murray and Taylor
(2006) do urge some caution, however, about overestimating the benefits of Fairtrade.
They note that small producers often have difficulties meeting the technical and quality
requirements set by the FLO and emphasise that significant support may be needed to
overcome such obstacles. There is further concern about the sustainability of the
movement in the long term if increasing numbers of producers move to Fairtrade
15 Accessed 26 May 2008. Accessed 26 May 2008.
production. At least in the short- to medium-term, however, it appears that Mexican
farmers could profitably avail themselves of this opportunity.
All this is to say that there are alternatives to futures hedging that hold much promise.
While only a first pass, the facts and figures presented above illustrate the income
benefits for farmers of supply management and Fairtrade. At the very least, there is
reason for policymakers, development institutions and researchers to cease their
derivatives dogmatism and take a long look at other viable programmes to insulate small
coffee farmers from market shocks.
In essence, the World Bank is gambling with the lives and livelihoods of small coffee
farmers in its efforts to incorporate derivatives markets into agricultural development
agendas. And this gamble is a costly one. Futures hedging can not only destabilise
farmer incomes via margin calls and aggregate price effects, but the Bank’s persistence
in pushing this solution has misdirected precious resources away from solutions that
may actually work.
Kamal Kabra, an influential Indian economist, submitted the following as part of his
statement to the Standing Committee on Food, Consumer Affairs and Distribution, which
was at the time considering ‘a decision to intensify and expand the futures markets’ in
The farmers, already exposed to the vicissitudes of input prices and quality, along
with the usual gamble in monsoon, would be exposed to yet another source of
uncertainty and trouble over which and against which they would hardly have any
defence (Kabra, 2006: 16).
Precisely right.
Oh, and one last thing… Did I mention that coffee prices are falling again?
Akiyama, T., Baffes, J., Larson, D. F., and Varangis, P. (2001). Commodity Market
Reforms: Lessons of Two Decades. Washington DC: World Bank.
Anderson, R. and Danthine, J.-P. (1983). ‘The time pattern of hedging and the volatility
of future prices’. The Review of Economic Studies 50(2), 249–66.
Baffes, J., Lewin, B., and Varangis, P. (2005). ‘Coffee: market settings and policies’. In
World Bank, Global Agricultural Trade and Developing Countries. Washington, DC: The
World Bank.
Bates, R. H. (1997). Open Economy Politics: The Political Economy of the World Coffee
Trade. Princeton, NJ: Princeton University Press.
Bohman, M., Jarvis, L., and Barichello, R. (1996). ‘Rent seeking and international
commodity agreements: the case of coffee’. Economic Development and Cultural
Change 44(2), 379–404.
Breger-Bush, S. C. (2009). ‘Hedging to safety: A political economy of derivatives, risk
and agriculture in the developing world’. Doctoral dissertation, Korbel School of
International Studies. University of Denver.
Calo, M. and Wise, T. (2005). ‘Revaluing peasant coffee production: organic and fair
trade markets in Mexico’. Working Paper. Medford, MA: Tufts University Global
Development and Environment Institute.
Cardenas, J. (2001). ‘The world coffee crisis’. Paper presented at World Coffee
Conference, London.
CRMG (2005a). Lessons Learned – Price Risk Management. Washington, DC: The
World Bank’s Commodity Risk Management Group.
CRMG (2005b). Price Risk Management Work Summary. Washington, DC: The World
Bank's Commodity Risk Management Group.
Davis, B. (2000). ‘The adjustment strategies of Mexican ejidatarios in the face of
neoliberal reform’. CEPAL Review 72, December.
Eakin, H., Tucker, C. and Castellanos, E. (2006). ‘Responding to the coffee crisis: a pilot
study of farmers' adaptations to the coffee crisis’. The Geographic Journal 172(2), 156–
FAO (2007). Managing Supplies to Raise International Agricultural Commodity Prices.
Rome: FAO Committee on Commodity Problems.
Gibbon, P. (2005). The Commodity Question: New Thinking on Old Problems. New York:
United Nations Development Program.
Gilbert, C. L. (2005). ‘The long run impact of the ending of coffee control’. Paper
presented at Second World Coffee Conference: International Coffee Organization.
Henriques, D. B. (2008). ‘Price volatility adds worry on US farms’. The New York Times,
22 April.
Holzmann, R. and Jorgensen, S. (2000). Social Risk Management: A New Conceptual
Framework for Social Protection and Beyond. Washington, DC: The World Bank.
ILO (2004). Economic Security for a Better World. Geneva: International Labour
ITF (2002). Mexico: Price Risk Management, Phase 2 Report. Washington, DC:
International Task Force on Commodity Price Risk Management.
ITF (2006a). Delivery Models. Washington, DC: International Task Force on Commodity
Price Risk Management.
ITF (2006b). Introduction. Washington, DC: International Task Force on Commodity
Price Risk Management.
ITF (2006c). Lessons Learned. Washington, DC: International Task Force on Commodity
Price Risk Management.
Kabra, K. (2006). ‘Comments furnished to the Standing Committee on Food, Consumer
Affairs and Distribution’. Hearings on the Forward Contracts Amendment Bill (2006).
New Delhi: Government of India.
Keynes, J. M. (1938). ‘The policy of government storage of foodstuffs and raw materials’.
The Economic Journal 48, September.
Lence, S. H. (2009). ‘Do futures benefit farmers?’ American Journal of Agricultural
Economics 91(1), 154–167.
Morduch, J. (1999). ‘Between the state and the market: can informal insurance patch the
safety net?’ The World Bank Research Observer 14(2), 187–207.
Newman, S. (2009). ‘Financialization and changes in the social relations along
commodity chains: the case of coffee’. Review of Radical Political Economics 41(4).
Ponte, S. (2002). ‘The “latte revolution”? Regulation, markets and consumption in the
global coffee chain.’ World Development 30(7).
Raynolds, L. T., Murray, D. L., and Taylor, P. L. (2006). ‘The future of Fair Trade coffee:
dilemmas facing Latin America's small-scale producers’. Development in Practice 16(2).
Sayer, G. (2002). Coffee Futures: The Impact of Falling World Prices on Livelihoods in
Uganda. Oxford, UK: Oxfam.
Snyder, R. (2001). Politics After Neoliberalism. Cambridge, UK: University of Cambridge
Talbot, J. M. (2004). Grounds for Agreement. Lanham, MD: Rowman and Littlefield.
Teweles, R. J. and Jones, F. J. (1999). The Futures Game: Who Wins? Who Loses?
And Why? Third Edition. New York: McGraw-Hill.
The World Bank (2005). Managing Agricultural Production Risk. Washington, DC: The
World Bank, Agricultural and Rural Development Department.
Thiele, R., Lay, J., Godart, O., and Bussolo, M. (2006). The Impact of Commodity Price
Changes on Rural Households: The Case of Coffee in Uganda. Washington, DC: The
World Bank.
Tiffen, P. and Fernadez, G. (2005). Report on the BMI-FEDA Hedging Program in El
Salvador. Washington, DC: The World Bank.
UNCTAD (2002). Farmers and Farmers' Associations in Developing Countries and Their
Use of Modern Financial Instruments. Geneva: United Nations.
UNCTAD (2006). Enabling Small Commodity Producers and Processors in Developing
Countries to Reach Global Markets. Geneva: UNCTAD.
Wild, A. (2004). Coffee: A Dark History. London: Fourth Estate.
World Bank (2001). World Development Report 2000-01: Attacking Poverty. New York:
Oxford University Press. Available online at:,,contentMDK:2
ml (accessed 18 February 2010).
The Brooks World Poverty Institute (BWPI) creates and
shares knowledge to help end global poverty.
Executive Director
Professor David Hulme
BWPI is multidisciplinary, researching poverty in both
Research Director
Professor Armando Barrientos
the rich and poor worlds.
Our aim is to better understand why people are poor,
what keeps them trapped in poverty and how they can
be helped - drawing upon the very best international
Brooks World Poverty Institute
The University of Manchester
Humanities Bridgeford Street Building
Oxford Road
M13 9PL
United Kingdom
practice in research and policy making.
The Brooks World Poverty Institute is chaired by Nobel
Laureate, Professor Joseph E. Stiglitz.