Mozambique : Political Choices and Rural Development  









Paper presented to the World Bank Seminar on Rural and Agricultural Growth and Poverty Reduction in Mozambique



January 2004




Steven Kyle

Cornell University

Mozambique : Political Choices and Rural Development  


January 2004

Steven Kyle

Cornell University



I.  Introduction


            This paper examines the political choices confronting the government as it tries to promote rural development in Mozambique.  As such, it focuses on issues of growth, but more particularly on who benefits from that growth.  It is therefore necessarily provocative in that these issues relate directly to who gets the money and who makes the decisions which result in some people getting money but not others.


            This is what politics everywhere revolves around.  Indeed, if there were easy ways for everyone to win and nobody to lose then politics wouldn’t be contentious at all.  The key fact for development is that there are some activities which will promote growth faster than others and this necessarily means that the benefits cannot be spread evenly over the population.  To try to spread the benefits evenly is to forego growth potential.


            This is not to say that it is not possible to have growth with equity.  Quite the contrary.  What it says is that in an activity like agriculture which relies directly on the natural resource base, there will be some locations more favorable and others less so to growth of particular kinds.  It is also true that as the economy as a whole grows, there will be structural changes which have implications for agriculture in terms of where resources flow, and for who even remains a farmer at all.


            This paper does not attempt to be exhaustive.  Rather, the object is to try to provoke discussion of issues which will  arise as Mozambique contemplates the next phase of its rural development.  These issues are roughly divided into those which relate directly to the functions of MADER and the decisions that it must make on the one hand, and more general issues of rural development on the other.





II.  MADER - Inside and Outside


            A.  Proagri I and the Renovation of MADER


            There is no doubt that MADER has benefited enormously from the first phase of Proagri and that it is now capable of operating effectively in a way that was not imaginable previously.  The improvements in bureaucratic process are important, but perhaps the most important is the visible effect on the morale and overall work atmosphere that is evident in even a short visit.  All evaluations of Proagri have noted the inward focus of this first phase but have emphasized the need to reorient the focus of the next phase so that there is more impact “in the field” and less focus on core functions in the capital city.


            What exactly does this mean?  Clearly, there is a link between the often touted concept of “decentralization” and the need to have a “field impact”.  However, it is extremely important to understand that these two terms are not synonymous and in fact may well translate into very different targets for the second phase of Proagri.  It is essential that these distinctions not be blurred and that the goals of the next phase be made explicit.  Several points can be made which relate to this:


1.  Decentralization can be taken by many to mean that funding should shift away from the central ministerial functions and flow toward provincial and district bureaucracies instead.  This is already happening to some degree, but we should realize that a focus on provincial or district level bureaucracies still amounts to an inward looking bureaucratic focus and is not the same as a focus on farmers themselves.


            While there is no doubt that lower levels need more money to be effective, and need to be strengthened and reorganized, we should not confuse ability to administer budgets at the district level with ability to affect farm level outcomes.  The bureaucracy is a means to an end, not an end in itself.


2.  It is essential that a reorientation of Proagri not result in a loss of the gains already made in core functions in the Ministry.  These gains need to be consolidated and made permanent.  They are a necessary support to what comes after, not one side of an either/or choice.  This means that adequate funding to support these functions must still be guaranteed in the next phase.


3.  A corollary of the second point above is that some activities are best pursued at a decentralized level but others are not.  That is to say, there are some things which ought not be decentralized.  To take an example, nobody would recommend that, e.g. construction or rehabilitation of a national highway be directed by district level officials.  Similarly, some agriculture sector functions need support/direction from the national level as well.  One example is research, where the public good attributes of the activity together with the low critical mass of researchers dictate a national level process of funding allocation, research prioritization and goal setting even though the process of research itself must necessarily involve direct farm level contact.  Another example is the national price information system, which requires central gathering and dissemination of information.  Other examples exist, but the point is that some evaluation of the proper location of different ministerial functions needs to occur, and the appropriate funding should flow accordingly.


4.  If farm level impact is the goal of the second phase of Proagri, then targets and goals need to be set accordingly.  This means that while bureaucratic efficiency may be an intermediate goal to be achieved, the outcome on the ground will be the test of true impact at the end of the day.  The overall guide to direct goal setting is, of course the PARPA, which supports the idea that indicators of success should/must be far broader than targets for production or marketed surplus, important as these may be. 


            To focus attention on results rather than process, some of the indicators might include measures of household welfare (income, nutrition, etc.), production related measures such as  yield increases, or technology related indicators such as adoption of new varieties, crops or agricultural practices.  Given the broader mandate of rural development as opposed to agriculture per se, measures of household welfare can include such things as education and health, while at the same time it must also be acknowledged that off-farm activities such as marketing are the key to growth and commercialization of the agricultural sector.



Politics and Priority Setting in Proagri II


            The first phase of Proagri focused on improving ministerial performance across the board.  Difficult political decisions which would tend to favor some stakeholders over others were avoided to a great extent, but this will not be possible in the next phase if the desired results are to be achieved.  Simply put, Proagri II is intended as an investment program and as such must focus on costs and benefits in its allocation of resources.  Field level impacts that are sustainable are the ultimate goal and this necessarily implies that activities which are capable of making money should be emphasized since those that don’t will fade away when donor funding ends.


            This immediately becomes a political issue since it is obvious that there will be winners and losers.  It is the role of the ministerial leadership to promote allocations which maximize returns even though political pressure to spread the money around to all will be strong.  Some of the dimensions along which choices will have to be made are:


1.  Spatial - Agriculture depends on the natural resource base and some provinces are better endowed than others.  This means that more resources should go to those areas with greater agroclimatic potential since there will be a greater return on the investment in those areas.  This implies that provinces with better rainfall and soils will be favored.  Other considerations include proximity to markets or roads, but the political process must be guided by a rational assessment of investment returns.  A cursory look at Mozambique’s resource base indicates that some of the central and northern provinces have greater agricultural potential than the drier south, where crop potential is more limited and cattle raising more prevalent.


2.  Ministerial function - If the goal is to generate a field impact then those parts of the ministry most directly related to this should be favored above others.  As a general statement, this means that the research institutes and the extension functions should NOT be treated merely as yet another national directorate competing for funds.  They need to be prioritized and given a larger share of the pie since it is these functions that will actually work to change on-farm conditoins most directly.  It should be noted that this does not prejudge the question of whether extension should be done in-house or outsourced to NGO’s or other entities.  Rather, it indicates that extension services (however delivered - see below) and the research results they are intended to convey need to be invested in as priority items.


3.  Crop - While it is not necessarily the job of the government to promote particular crop choices, it IS the job of the government to set priorities for national agricultural research.  While various assessments have correctly emphasized the need for a demand driven approach to agricultural research, this begs the question of exactly which of all the competing demands will be met first and which will not.  Here, the government needs to be clear in setting priorities according to the goals of the PARPA and the potential for research to generate increases in marketable crops on a large scale.  It is likely that staple grains (e.g. rice, maize)  which are widely grown by smallholders will have the potential for the greatest impact though other horticultural or cash crops may also be important.  It should be noted that export potential is an important part of the calculation for any decisions and that food crops are high on this list. 


            Given the low yields obtained in Mozambique compared even to neighboring countries with similar agroclimatic conditions, together with the lack of even basic screening and adaptation of existing varieties of many important crops, there is are several areas where relatively quick and large gains can be made.  Focusing the limited capacity of the research system on these will be important even though this means that some constituencies will not be happy.


            Tables 1, 2 and 3 show recent estimates of potential returns to yield improvements in major crops. It is immediately obvious that the potential is larger in some cases than others.  Relating to point 1 above, it is also obvious that some provinces weigh more heavily than others.  Table 4 shows average yields for important crops in Mozambique compared to African averages and also those for Zimbabwe (prior to the onset of the current difficulties).



4.  Farm type - In colonial times, an explicit distinction was made between commercial and family farms, and little effort was expended on improving conditions on the latter.  To recreate or perpetuate such a bimodal division in the MADER bureaucracy would be a mistake given the problems inherent with bimodal systems everywhere else in the world through history.  Even so, everyone is aware that there do exist alternative farm types and there will be continuing choices to be made regarding such issues.  It is clear, however, that the low level of technology used on small farms means that the potential gains are all that much greater while at the same time, the distributional consequences of improving welfare of smallholders is in keeping with the goals of the PARPA.  The tradeoff is often one of time: Quick results in terms of output can often be achieved by importing large scale commercial growers even though the results achieved may be less widespread across the population. 


            The role of the government is not necessarily to favor one type and discourage another.  Rather, given the likelihood that any probable future scenario will contain both large farms and small, the questions for the government are two:


-   Which government priorities in funding research or investing in improved infrastructure will generate the highest returns over the long run?


-   Which among these will not/cannot be done by the private sector on its own?



5.  Location in production/marketing chain -   A more expansive view of agriculture and rural development requires a vision which includes off-farm activities broadly conceived.  Accordingly, the question then must be asked, where do the most important constraints to progress occur?  While many of these are undoubtedly production-oriented, it is clear to most observers that marketing is the most important bottleneck in many cases.  This means a broadened focus not just toward activities such as formation of producer associations, but also toward agribusiness more generally.





III. Structural Transformation and the Politics of Agricultural Investment and Taxation


            Let us presume that Mozambique is successful in promoting economic development over the coming years.  If we take this to mean that per capita income grows at sustained rates of 5% then the average Mozambican will be twice as rich in 14 years as he or she is today.  At a growth rate of 7% this doubling of income will occur in a little more than 10 years.  Such changes are entirely possible, but have profound implications for rural development and rural society.


            First, such growth will not occur if there is not a sustained level of investment in the agricultural sector.  This investment MUST come from both the government and the private sector.  While experience both within Mozambique and elsewhere in the world demonstrates repeatedly that agriculture per se is an activity best left to the private sector, there are nevertheless crucial investments that must be made by the government since the private sector simply cannot make them.  Among these investments are basic infrastructure such as roads, health and education, and research and dissemination of new technologies.  The returns on these investments are extremely high (see below) and the desired increases in rural (and urban) incomes will not occur without them.[1]


            If the average farm inhabitant today feeds himself and perhaps one other person, in the future each farm resident will have to feed an additional two, three, or more people.  This fact stems from the often observed and never violated patterns of development seen in every country in history as per capita income increases:


1.  Agricultural output grows in absolute terms


2.  Agricultural output as a share of total GDP declines


3.  The agricultural labor force as a share of the total goes down


4.  Rural-to-urban migration raises the share of population in cities while the share in rural areas goes down.


            There are several important considerations which follow directly from the points above:


-           Success in promoting development means that not all farmers today will remain in the farm sector in the future.  Many will migrate to urban areas while others will engage in non-farm activities in rural areas.  This means that there will be some “winners” and some “losers” as more successful farmers expand and continue producing while less successful farmers leave agriculture in search of other pursuits.  In other words, a lot of the farmers we see today won’t be farming in twenty years, or at least their children won’t.


            One implication of this is that if not all farmers are created equal, then not all farmers will provide the same return on investment.  It is to be expected that more successful farmers will (and should) have access to more investment funds than ones who are not able to remain profitable.  Such changes may well be because of inherent differences in abilities which exist in any group of farmers (or anyone else) here or anywhere else in the world.  However, even if all farmers really were equal in terms of endowments and abilities, the realities of structural transformation would result in a significant change in the labor force away from agriculture and toward other sectors.


-           If most of the population of the country is non-farm, then they will need to buy their basic food requirements rather than growing them.  This means that ALL crops will be cash crops in the future.  There will be no artificial division between those grown “for export” and those grown for own-consumption.  Indeed, in developed countries farmers themselves buy most of their basic food requirement, and don’t know or even care if the crop they grow is consumed domestically or exported.  All that matters is that they are paid for it and that the local market functions well enough for them to buy what they need.


-           If it is to be expected that a large part of the current rural population will leave agriculture, then it is all the more important to make sure that they have the education needed to succeed in other kinds of work.  Let us be clear: Farmers are more productive when they are educated and more educated parents have been shown to have better nourished and healthier children, so there is already reason to favor a policy of educating rural populations.  However, the adaptability and skills needed to change occupations reinforces this policy, and underscores the need to promote education as a basic part of the rural investment program.


-           We have noted above that both public and private investment are needed in agriculture and rural areas generally if development is to succeed.  However, it is equally true that agriculture must necessarily be taxed in some manner if the government is to have the wherewithal to make the investments required.  That is, the first decision is where to invest - and the answer for any economist is to put the money where it will earn the highest return.  The second decision is what to tax - and here the answer is to tax those activities which will least distort the decisions of producers and consumers. 


            This often leads economists to recommend lump-sum taxes of some description, but regardless of the particular form of the tax, it is virtually inevitable that at early stages of development such as that in which Mozambique presently finds itself, it is as a practical matter unavoidable that agriculture will be taxed in some manner simply because that is where most of the people are.  The most important point to remember here is that it is essential that the utmost care be taken to avoid discouraging those activities we are trying to promote.  It is entirely possible that the optimal course would be to invest in an activity since it generates a high return and then to tax that same activity once the money is made.  The key is to keep the taxation at a level which does not discourage continuation of or reinvestment in moneymaking sectors.



IV. Land Tenure


            Any observer of Mozambican agriculture over the period since independence is aware of the debate over revision of the Land Law and the intensity of feeling that this debate can arouse in many participants.  The author of this paper is an economist and therefore has fairly predictable opinions as to what a “good” land law should look like.  In a nutshell, such a law would be one that would best support a well functioning land market.


            However, there is much to recommend that all stakeholders should accept the current Land Law as it now stands, at least for the short term.  There are several reasons for this:


1.  The political capital needed to effect further change in the land law would be huge and could well be wasted anyway.  This is particularly true given the fact that the next elections will result in a new President who will likely not want to re-engage in this debate, at least not as a first order of business.  In addition, many stakeholders like the current law and would strongly oppose change.


2.  One of the reasons often advanced for a revised law is the need for smallholders to use land as collateral for credit.  However, lack of a particular form of land tenure is not the only or even the biggest barrier to smallholder rural financial services.  Smallholders could have officially recognized tenure of any form at all and would still not get credit from the formal financial system. 


3.  Unofficial land markets operate anyway.  This is known to be true in areas with substantial land pressure (e.g. the environs of Maputo) and will occur elsewhere as incentives to do so increase.  While experience in other countries indicates that unofficial markets are really not a substitute for official ones, the fact that such transactions are taking place will help create pressure for ratification of what exists in reality and for legal amendments to make it possible to have official recognition.


            Rather than expend political capital in a probably fruitless effort to change the law it would be far more productive to focus on getting smallholders officially recognized tenure under the current legal regime.  Some form of recognition is far superior to no recognition at all.  Once small farmers have any kind official recognition, they will become a political force in their own right (or at least a more effective one than at present) and this will do more to protect their rights than the difference between one form of tenure and another.


Small Farms, Large Farms, and the Rural Economy and Society


            Perhaps one of the biggest issues facing authorities now and into the future is the question of to what extent large commercial farms should be allowed or encouraged as opposed to small farms.  There is a very important point to be made at the outset.  To encourage a dualistic farming system where there is one favored sector of large farms coexisting with a much poorer and less favored sector of family farmers is a very bad idea.


            Who among us would want to replicate the disaster that is happening across the border in Zimbabwe?  This is the basic problem with such dualistic systems.  They have some attractive characteristics in that they can generate faster increases in cash crop production based on rapid adoption of modern technologies but even if these seeming economic benefits outweigh the costs (which is debatable) the existence of a bimodal system puts into place long term social and economic inequities that sooner or later will cause major dislocations as has happened in Zimbabwe.


            Having said this, there is definitely a role for large farms in the country’s rural development strategy.  Large farms can serve as sources of wage income for surrounding areas, and in some cases are the scale most appropriate to the crop grown.  These cases, though, are rare, which is one of the main reasons contract farming with smallholders has become a popular mode of production.


            What, then, to do with proposals for large scale commercial operations?  First, it should be noted that the eventual goal for small farmers is to become larger commercial operations themselves, since that is what success will inevitably mean. (Success certainly can’t mean that they will stay small and poor) The real test for the ability of a large operation to contribute to rural development is the extent to which it links with other domestic entities both on the input side and on the output side.


            So, operations which demand raw materials from smallholders (as contract farming operations do) or which feed into local processing operations (mills, etc.)  are increasing the degree of integration of the rural economy while raising incomes.  Enclave production units which have little or no interconnection to any local entity will not help promote development.  They will simply compete for resources (land in particular) while doing little else to raise welfare.


            It should be noted that the observations in the paragraphs above say nothing whatsoever about the nationality of entrepreneurs who engage in large scale contract farming (or other large) operations.  This is because from a purely economic point of view it matters little what passport a capitalist has.  A capitalist will put his or her money wherever it is safe and can earn the highest return, and this is true for citizens as well as foreigners.  While there may be valid reasons for wanting local participation in large projects, from the point of view of rural development it is less important who the entrepreneur is, than what type of project they propose and how it will benefit the country.  A foreigner, no less than a citizen, will keep any profits in the country if there are good investment opportunities creating the incentive to do so.  Foreigners too are subject to taxation on income, thus providing the government with any needed leverage to change behavior or increase revenue.



V. Cash Crops, Food Crops and the Role of MADER in Promoting Them





            It was noted above that it is NOT the role of the government to decide who should grow what crops and where.  However, it IS the role of the government to decide what national research policy is.  There is no substitute for a rational process of economic analysis to decide which crops can generate the highest levels of returns (i.e. profits) for the broadest segments of the population.  Here, it is important to remind ourselves of one of the most basic lessons of the section on structural transformation above: ALL crops become cash crops in a successful strategy of rural development.


            In the short run, however, there is much to recommend a public research strategy emphasizing basic staple food crops.  Among the reasons for this are:


- Staple grains such as maize and rice are already widely grown by smallholders so that even if off-farm markets are adverse they can still eat their output.


- Given that they are widely grown, their culture is well understood and technological advances can be widely disseminated, spreading the benefits over a very large population


- International research efforts have generated an enormous backlog of advances in new varieties as well as other areas which have yet to be adopted or adapted to Mozambican conditions.  Given the extremely low levels of Mozambican yields a relatively small investment in research and extension can double or even triple yields of these crops.


- Crops such as maize and rice are already exported, and so production increases have a positive macroeconomic impact as well as a local one.  Malawi, Zambia (and currently Zimbabwe) are food deficit countries where long term trends indicate good export opportunities for Mozambique, while the Zambezi delta is the only major area of rice production in the entire SADDC region which is a net rice importer.


            Perhaps the biggest reason to concentrate on a few crops with the biggest potential impact is the fact that the national research capacity is still extremely limited.  There simply isn’t enough of a critical mass of trained researchers to spread their efforts across all of the research goals which appear viable.  It is essential to pick those which are most important and to do this as a matter of national priority setting rather than leaving it to negotiation with donors.


            In fact, there is a real danger in inviting cooperation from the entire range of international agricultural research centers or even the entire range of departments within bilateral partners such as EMBRAPA.  To do so would spread resources so thin that there would be only scant progress in any one area.  The country cannot afford this, and equally cannot afford to take a passive role in the decision process.  Neither the IARCs nor bilateral partners will make the decision to promote some part of their range of expertise at the expense of others since they too have their own internal political imperatives.  Rather than agreeing to “Christmas tree” projects with every conceivable ornament hung on them, the Mozambican government must itself prioritize activities and then seek international assistance in accord with those priorities.


            None of the above should be taken to mean that horticultural or other high-value crops should be ignored.  They shouldn’t.  However, the government needs to take care that it does not dissipate its efforts in areas where the private sector is active.  Among these areas are obviously those where large scale operations (e.g. sugar) or contract farming are the dominant forms (as in cotton or tobacco).  Even in some smallholder horticultural crops there is sufficient NGO activity that the government need not duplicate efforts.


            A final caveat is the observation that it is not always the case that farm level technology is the most important constraint.  Marketing is probably the most important constraint to progress in many areas, and is discussed in more detail below.




Seed Supply Systems


                        At the present time, Mozambique is host to a variety of seed supply systems, ranging from informal, based on farmer retention of own-seed, to formal, the most important of which is SEMOC, the national seed company, which is a joint venture between the government and the private sector.  In addition, there are direct private sector seed supply operations in such areas as cotton, where JVCs provide seed to those farmers growing in their areas.  Overall, it is likely that Mozambique will continue to have a variety of seed sources and supply mechanisms, and this is beneficial, given the diverse needs and conditions of production in different sectors and areas of the country.  As has been noted elsewhere[2] widespread adoption of modern varieties can often depend upon informal as well as formal seed supply systems, given the cost effectiveness of informal methods in getting seed to dispersed low income populations.  This is particularly true in Mozambique, where low population density and poor transport conditions increase costs for centrally directed operations.[3]



Fertilizers and Other Purchased Inputs


            In large areas of Mozambique, as in other parts of Africa, much of the value added in smallholder agriculture comes from production technologies where the only other major input apart from labor is the hoe (enxada) and billhook (catana).  Fertilizer use per hectare is far lower than in any other area of the world, and is below world average application levels by a factor of approximately 4.   This disparity is even more striking in the case of maize where average fertilizer use in Africa is less than one eighth that of developing countries as a whole.[4]


            Evidence from Mozambique shows that fertilizer use there is low even by African standards.  Only 7% of all farming households use fertilizers at all, though as expected, this type of technological intensification is more prevalent in the more densely populated central provinces.  According to surveys, land tenure has no correlation with fertilizer use, but use is greater among households with more than 3 hectares of land.2


            Accordingly, there is ample scope for increasing output through increased applications of nutrients.  This is particularly true in areas where population growth is creating pressures for intensification, as is the case in the South and in peri-urban zones.  Eventually it will be true throughout the country. In fact, one recent study suggests that 50-75% of the maize yield increases in developing countries outside of Africa  from the mid 1960's to the 1980's can be attributed to fertilizer use. That such a strategy could be viable in Mozambique is supported by the fact that in three of Mozambique’s neighboring countries, Malawi, Zambia, and Zimbabwe, half or more of the maize area was fertilized in 1990.   Maize yields in Mozambique averaged 0.5 T/ha. (1991-95) which is half or less of those in these three neighboring countries, demonstrating that it is likely that there is considerable untapped potential.  Even so,  there are several inherent problems with increasing fertilizer use, some of which are especially pertinent in the Mozambican context.    Among these are:


- Fertilizer represents the largest cash outlay for those who use it.  This is a prohibitive problem for many smallholders, who can only use purchased inputs if rural traders are willing to extend credit.


- Given the erratic nature of rainfall in some areas, risk increasing inputs such as fertilizer will be applied in sub-optimal amounts by risk averse smallholders


- Fertilizer is bulky, and both transport and storage costs are high.


            Government can usefully assist by promoting infrastructure development to reduce transport and storage costs, promote group purchases where possible at local and higher levels through cooperatives or other mechanisms as a means to exploit economies of scale, and sponsor research into appropriate use of high analysis fertilizers such as urea, diammonium phosphate and triple superphosphate as a way to bring prices down.  Fertilizers are more likely to be cost effective on horticultural crops at the present time; indeed, examples of this can be found in some areas.  In the long run field crops such as maize, which can be very responsive to fertilizer application, are also likely to provide additional demand.


            A note on producer associations is appropriate.  There is a history of government-created and supported producer associations with a distinct political purpose.  The history of Mozambique and other countries shows that there are many pitfalls that afflict such government sponsored entities and that they are probably more often counterproductive than they are useful.  In addition, once created  they are difficult to back away from .  On the other hand, associations which are truly voluntary and created from the grass roots do have the potential to be a positive force for smallholders.  There are some NGO’s which have experience in assisting this type of association and which have been successful in providing benefits for farmers.  To the extent that these groups can remain truly independent, they are worth encouraging.






            Cashew is a crop of such historical and widespread importance to the smallholder sector and to the country’s balance of payments that it would be a mistake not to mention it in this paper.  Cashew has been the subject of intense political debate in the past over government policy toward processing and export taxes.  It good that these debates are largely a thing of the past - it would be a mistake to reopen these issues in light of the fact that there are several important initiatives that ought to be supported in the cashew subsector regardless of government policy toward taxes or exports.


            The fact that cashew is a widespread smallholder crop that can contribute significantly to the balance of payments is sufficient reason for the government to take a particular interest in its cultivation.   Relatively small gains translate into large ones when spread over hundreds of thousands of producers.  Given the very low incomes of most of these producers, it is clear that private sector initiative cannot be relied on to provide what is needed to poverty stricken smallholders.


            The appearance of a devastating disease, oidium, makes development of resistant varieties an urgent task.  There are apparently resistant strains that can be used as input to such a research program, which would feed into a much needed program of seedling multiplication and replanting.  The growth of manual technology processing plants, when combined with rejuvenated plantings, has the potential to provide significant cash income for households while at the same time providing an important export.




            Mozambique lacks a tradition of integration of research, extension, and teaching functions such as is found at land grant universities in the US or their peers elsewhere.  The lack of communication between these various entities reduces the potential for their work to contribute to the goal of increasing rural welfare and farm output.  However, it is clear that both by strengthening the relevant institutions and by improving the linkages between them, it is possible to make a substantial impact on farm welfare and marketed surplus.


            Research and extension are two functions which subsistence level smallholders cannot be expected to be able to pay for on their own.  While privatization of these functions for commercial cash crop growers is likely to provide good results, there is no feasible way for a private sector entity to generate a reasonable rate of return if catering to the smallholder sector.  However, it is exactly in this sector that the greatest potential long run gains are to be found, and from a national policy point of view, improvement of the national food balance via increases in marketed surpluses of food crops is a goal of primary importance


            One result of the current importance of NGO’s in extension is a wide variety of approaches in different districts.  This is beneficial in that there is an opportunity to compare and evaluate the different programs and methods used, in addition to the fact that different locations have different problems and constraints to be addressed.  However, there is a very real issue in deciding the overall balance between NGO’s and national systems in providing needed services in rural areas.  The Ministry of Agriculture and Rural Development is currently undertaking a pilot project in outsourcing extension services to non-governmental providers.


            Outsourcing has been debated in Mozambique, with some insisting that the best path is to strengthen the government extension system, others promoting the virtues of NGO’s, while still others promote the use of government contracts to get the needed services provided.  Several observations spring to mind in connection with these debates:


1.  NGO’s often have more experience in providing extension services than does the government system in a particular area, and may well be able to do a particular task faster and cheaper than could otherwise be achieved.


2.  While it may be preferable to have a strong government extension system, building it up from its present state is a long term task.  Doing this is important since providing a two-way channel of communication between the farm level and the ministry, but this should not get in the way of other arrangements in the short term if they work and are cost effective.


3.  No matter which of the bureaucratic forms through which extension is provided, it must be remembered that there are only a certain number of extensionists available, and they will work for whoever provides them with the best salary and other work conditions.  What is important is that they keep on working as extension agents (regardless of who actually sends them their paycheck) and that they have useful information for small farmers. 


            It is likely that a variety of extension provision mechanisms will be useful over the next years.  It would be a mistake to insist on particular format or to forego the experimental approach which seems to have been adopted by MADER as the public extension system is gradually strengthened.  What is important to remember is that small family farmers need extension, and that the returns to such services can be very large.  At the same time, there is little reason to think that the private sector could provide these services on its own, given the limited capacity of smallholders to pay.




VI. Markets and Roads      




            It would be difficult to overemphasize the importance of  improving transportation and road quality.  Given the extremely poor state of the transport network in many areas, particularly in the North and in Zambezia, transportation is the single most important factor that can contribute to the long run sustainable growth of agriculture.  The donor community’s country assistance strategy has rightly targeted this problem with its successive Roads and Coastal Shipping credits.


            From the farm point of view, lowered transport costs (which should be understood to include improved security along transport routes) provides benefits on both the input and output side as it simultaneously allows lower input costs together with higher farmgate prices for output.  Given the desire to increase the extent of integration of smallholders into the market economy, these links are an indispensable necessary condition for this to occur.


            From the point of view of the marketing system, it is clear that lowered costs in terms of travel time and depreciation of trucks will have a profound effect both on the marketing margins required for services to remote areas, and on the ability of traders to compete in any given area.  Investments in improved roads can do more to improve the financial calculations of marketing agents than can any other intervention.  As noted at various points throughout the paper, there are few public activities more important than ensuring adequate transport systems throughout the country.  Private sector recapitalization and accumulation can only take place if there is sufficient public capital investment in complementary sectors. 


            It is at this point nearly impossible to prejudge, based on current traffic flows, the optimal road configuration for the future, since current traffic flow data are themselves heavily influenced by the poor state of the roads.  However, it is clear from maps of production zones and current roads that there is a great need for rehabilitation, particularly in the most populous and the most potentially productive zones.  As argued above, this will not only allow lower prices for inputs and higher farmgate prices for outputs, but will also facilitate growth of the marketing network and will lower costs for all who need to use the transport system.


            A perennial problem in Mozambique and elsewhere is that of the recurrent cost of maintenance.  Here there is a tradeoff.  While the maintenance need of dirt roads are far greater than paved ones, they can be addressed in a much more labor intensive manner.  This approach has the benefit of injecting additional cash into rural areas and is one to be encouraged but only if the maintenance does in fact get done.  Given the numerous examples of roads which are repaired only to fall back into a more deteriorated state, there is much to recommend paving the most important roads, and certainly the primary roads, many of which have yet to be improved.  Regardless of whether roads are paved or not, it is extremely important to ensure adequate maintenance, in order to avoid the need to rebuild roads from scratch every few years.


            From a political point of view it is worth noting that road building is a politically popular activity in all countries - but road maintenance is not.  Given that agriculture is not able to insist on the resources needed to maintain roads, this fact must be factored into any decisions regarding extension and/or rehabilitation of the road network.





            This is perhaps the biggest single bottleneck constraining rural development in Mozambique at the present time.  One of the biggest underlying problems is the state of roads as noted above, but equally important is the lingering attitudes of many that marketing is a somehow parasitical activity which rather than providing needed services, is merely exploiting the labor of the peasants.


            There is no doubt that traders can under some circumstances exploit monopoly positions, but the solution to this problem is not to try to control them but rather to encourage other marketers to enter and compete with them.  On the producer side, formation of producer associations can reduce the ability of any individual trader to exploit individual farmers.  This is one important reason to promote the continued formation of producer associations.


            There are indeed large profits to be made in agricultural marketing.  This is normal in an activity as underprovided as marketing services are in many areas in Mozambique, and serves as a signal to the private sector to expand activities.  While some may resent the notion that marketers can make a lot of money, it should be noted that their costs are extremely high in Mozambique, given the poor state of roads, and also that they sometimes lose money, though this is seldom viewed as a problem by outsiders.


            Apart from the abysmal state of infrastructure and the high costs that result, one of the biggest obstacles to growth in marketing is in fact the system of government permits and licenses that must be negotiated in order to be in compliance with the law.  While small traders can avoid much of this, they also cannot grow into large operations unless they comply.  Elimination of much of this regulatory apparatus would be beneficial to rural development and to the country in general.




Rural Finance


            Rural financial development is a slow moving variable.  That is to say, growth of formal financial systems which extend credit to farmers on a voluntary and competitive basis is something that occurs late in the development of rural and agricultural economies.  This was true in the US and Europe and will be equally true in Mozambique and elsewhere in Africa.  Nevertheless, there are various ways in which financing can be provided to smallholders, though not necessarily in a traditional bank.


            Probably the most important of these is the bundling of financial services with provision of other inputs such as seed and fertilizer or with consumer goods.  In fact, this type of informal financial system was the predominant form prior to independence when small rural traders provided the bulk of services to smallholders.  While there is little reason to expect a return of the small cantineiros that were prevalent in colonial times, the use of ad hoc credit arrangements by input suppliers will likely remain an important source for smallholders, especially as markets in general develop and grow in rural areas.


            Other types of credit provision are the object of numerous projects by NGO’s and others in recent years.  Grameen Bank style micro-credit schemes have been tried as well as various village level savings and loan operations.  Continued experimentation along these lines together with replication of successful ventures will help to provide the basis for financial sector growth.  In the meantime, it is likely that large farmers will continue to have better access than will smallholders, and that rural people with relatives in urban areas will also have more options than will others.


            What is very clear from the experience of numerous countries around the world is that attempting to force rural credit market development by government fiat can lead to disastrous results.  Whether done by subsidy or by government requirements on the allocation of bank loan portfolios, such rural credit “development” not only stifles real rural financial market growth, but often leads to worsening income distribution and the weakening of formal sector financial institutions.






VII . HIV and Rural Development


            Mozambique has high levels of HIV infection, but they are nevertheless below those of many of its neighboring countries.  As the process of rehabilitation and economic reactivation and growth continues, it is likely that these rates of infection will tend to rise.  This has several implications for agricultural development and rural policy:


1.  Labor scarcity - Massive mortality of prime-age rural inhabitants implies that there will be more demand for labor-saving technological change than would otherwise occur.  It also implies that rural labor markets will have higher wages than they otherwise would.  While this may well translate into demand for labor saving agricultural technologies as traditionally understood, smallholders will look at it from a household perspective.  That is to say, if labor time is scarce then smallholders will seek to save time however possible across the whole range of household activities.  This includes farming operations but is not limited to them.  It is entirely possible that some extremely time consuming activities such as cooking, wood or water supply would generate the most attractive opportunities for labor saving change.


2.  Roads and HIV - The epidemiology of HIV in Africa makes clear the crucial role of roads and transport routes in communicating the disease.  Given the massive road building/rehabilitation work ongoing in Mozambique, it would be extremely negligent not to recognize the role this will play in helping speed the spread of HIV.  Accordingly, HIV education and prevention efforts should go hand-in-hand with road work.


3.  Extension and HIV - There are two main points to be made regarding extension and HIV:


- First, extension agents are the main source of information on the topic for many people, regardless of whether they are officially mandated to do HIV education efforts.  This means that they must have correct information, and should not promote false notions such as the AIDS preventing qualities of new crop varities (as was witnessed by this author in some locations)


- Second, extension agents are themselves highly at risk to be carriers of the virus themselves.  They are largely in the most at-risk age group, have transport, and have contact with many different people in the course of their work.  This is another powerful reason to require AIDS education for all extension workers, but is also a reason to expect higher than normal mortality rates among extension workers in the future.













Table 1:  Estimated Additional Annual Net Benefit from Distribution of Improved Seeds Alone Based on Purchasing Seed Only Once Every Three Years in Mozambique*



Estimated Area Harvested (1000 ha)

Estimated Annual Benefit from Improved Seed

Percent Contribution to Total Increase

Three Most Important Provinces for Improved Material

109 Meticais

109 USD






Nampula, Zambezia, Cabo Delgado

Open Pollinated Maize





Manica, Zambezia Tete







Nampula, Tete






Nampula, Zambezia, Tete






Cabo Delgado

Tete,  Nampula






Zambezia, Sofala, Gaza

Sweet Potatoes





Zambezia, Maputo, Manica/Sofal






Tete, Zambezia, Manica







Source:   Rohrbach et. al. 2001







Table 2:  Annual Additional Net Benefit by Province and by Household in Each Province Accruing from the Distribution of Improved Seed in Mozambique



Number of Rural Households*

Percent of Total Annual Additional Benefits

Net Annual Additional Benefit Per  Household


US Dollars






Cabo Delgado























































Source: Rohrbach et. al. 2001

Table 3: Additional Net Benefit (1,000,000 MT) from Distributing Improved Seed Alone By Province








S Potato













Cabo Delgado


























































































Total 106 MT










Total 106 USD










Total 1017











Percent of Total Kcals










Source:  Rohrbach et.  al. 2001

*Conversions from Metric Tons to Kilocalories obtained from Annex to FAO Food Balance Sheet Handbook 2001









Table 4: Estimates of Average Yields in Mozambique and Comparisons (t/ha.)



Average Yielda


Yield b

Comparative Zimbabwec


Yields Africac


5.0 -7.0

10.0 - 15.0




0.3 - 0.6

0.5 - 1.5




0.3 - 1.3

1.0 - 3.0




0.5 - 1.0

1.0 - 2.5




0.2 - 0.4

0.4 - 0.6




0.3 - 0.6

0.6 - 1.2




0.1 - 0.2




a  Varies according to agroecological conditions

b  Based on yields achieved with intermediate technology

c  FAO Production Year Book.

Source: ASM



            [1] In fact, there is a historical example of what happens when these investments are not made: the case of the Soviet Union, where the government neglected investment in agriculture, with a consequent decline in the ability of the country to feed itself and a decline in the real value of urban wages as food became more expensive in relative terms.  Eventually, these tensions played a major role in the eventual slowing of the pace of growth and development.

            [2]See, for example, V. Venkasetan, Seed Systems in Sub-Saharan Africa, World Bank Africa Technical Dept. Series No. 266, 1994.

            [3]This implies that the role of the government may not necessarily extend beyond varietal development, depending upon the public good properties of the seed in question.  As discussed in Jaffee and Srivastava Seed System Development - The Appropriate Roles of the Private and Public Sectors World Bank Discussion Paper 167, 1992, and in M. Morris ed. Maize Seed Industries in Developing Countries Lynne Rienner, 1998, the private sector cannot be expected to actively engage in production where benefits accrue to a broad audience and cannot be recouped by any individual or company.  This characteristic often applies to breeding and research, but is much less often true of multiplication or marketing & distribution.  Government support, where necessary, can be effectively provided through such mechanisms as the joint venture approach now existing in Mozambique.  In this way, the benefits of quality control and marketing support can be provided where needed without sacrificing the financial discipline which the private sector can provide.  Nevertheless, it is still the case that many farmers in Mozambique will continue to rely on more informal sources and that permitting and encouraging these can help promote efficiency and growth.


[4] See P. Heisey and W. Mwangi, “Fertilizer Use and Maize Production” Chapter 13 in Africa’s Emerging Maize Revolution, Byerlee & Eicher eds. Lynne Reinner Publishers, 1997.


* Based on yield increases of 25% for grains and legumes; 35% for roots and tubers and 1998/99 prices and estimated yields and hectares harvested.  Additional  Net Benefit is the average of three years of additional value (output price X higher yield minus current value) minus the cost of the seed purchased in year one.  Except for sweet potato, estimates are based on provincial level area and yield estimates from the Early Warning System in MADER for 1998/99.  Output price data are actual figures from MADER or the national accounts estimates and seed price data are from SEMOC.  See Table 3 below for further detail.

* Based on 1997 Census Data