Change Font: A A A A Contact Us What's New FAQs Subscribe home
HomePublicationsDynamics of Development in Rural CommunitiesInfrastructure and Rural Development in Agrarian Reform Communities

Infrastructure and Rural Development in Agrarian Reform Communities

The community level models shall be able to characterize the multiplier effect that any targeted intervention is expected to generate, not only among the direct beneficiaries, but among the indirect beneficiaries as well. Even if only a segment of the community directly benefited from an intervention, if there is a viable sustainability facility for that intervention, then the whole community can benefit as well. This analysis is expected to provide evidence on how the benefits received by direct beneficiaries are spread into the whole community. This evidence will further support the idea that resources should not be spread too thin, that a few bundles should be implemented instead of plenty single/stand alone interventions. This strategy will provide optimal gains and better chances of observing multiplier effects and subsequent sustainability.

Panel models are used to generate evidence on the accumulation effect of infrastructure and other development interventions at the community level. This will provide simple evidence of sustainability of development gains of the current strategies.

The Rural Communities

The data used here came from the Agrarian Reform Communities Level of Development Assessment (ALDA) conducted by the Department of Agrarian Reform. The yearly activity is aimed to generate information about the different aspects of development among the agrarian reform communities supported by the department. It collects basic indicators at the community level and aggregates them into an index for each of the key result areas (described in the next paragraph) and finally into an overall index. The index is a measure of a community's development, so if interventions among key result areas are accomplished, rural development will be feasible.

The department's major goal is land distribution. Since land distribution alone will not necessarily result in tenure improvement and eventually in rural development, other support services are also provided either directly or through facilitation of access to such. The key result areas of the department include land tenure improvement (LTI), economic and physical infrastructure support services (ECOPISS), farm productivity and income (FPI), organizational maturity (OM), basic social services (BSS), and gender and development (GAD). An index or a score (range of 0100) is computed for each of the key result areas before the overall index is completed. The indicators are immediate outputs (either directly provided or facilitated to be attained), intended to meet the key result. A community's score depends on the amount of output resulted from the interventions delivered.

The department does not necessarily deliver all those services, but facilitates for other departments to deliver such in the ARCs. An ARC is comprised not only of agrarian reform beneficiaries (ARB), but of all households in the community regardless of whether they are beneficiaries. The idea is to develop the entire community for sustainability of the development gains. At present, the ARCs account for approximately 20% of all rural communities in the country.

To be able to deliver all those results, regular appropriation from the government as well as development assistance from multilateral agencies (grants or loans) are used. The development assistance called foreign-assisted projects (FAP) usually includes the following menu: physical infrastructure (construction and/or rehabilitation), community and institutional development support, agricultural productivity and rural enterprise development, basic education support (construction/rehabilitation of school buildings), and primary health care (construction of health centers, provision of medicines, etc.). Some projects include all these support services, while others include a subset and or only one of them.

As of 2005, the households in the ARCs have an average total income of PhP 92,773 per household, of which an average of PhP 53,802 comes from agricultural sources and PhP 39,420 comes from non-agricultural sources. These figures are along the same lines as those presented earlier, considering that the coverage is 20% of all communities only (total income is PhP 110,482). The ARCs have been in progress for about 8 years as of 2005.

The average overall index for all communities is 71.52, indicating that there is still ample room for the department to provide rural development inducing interventions. Among the key result areas, LTI is the only one approaching completion (100) with the average index of 92; the target is to complete land distribution by 2008. The BSS also yield a higher average score of 87, followed by ECOPISS with 68, OM with 67, FPI with 66, and GAD with 51. The communities are better served in terms of basic social services but work needs to continue, and even grow in intensity when it comes to economic and physical infrastructure, farm productivity improvement and gender and development.

The total land area distributed relative to scope averages 92% per ARC, consistent with the actual ARBs relative to the ARB scope of 93%. One negative effect of agrarian reform is that some beneficiaries did not appreciate the purpose and were easily lured by the quick cash value of the land. Only 77% of the ARBs are still cultivating the land as of 2005.

In terms of physical infrastructure, accomplishment rates are similar. Accomplishment of rural roads (length) averages 60%, 65% for bridges (length), 69% for the number of irrigation systems required, and 62% for area coverage of irrigation (relative to total irrigable area); 67% of ARBs needing irrigation are actually served, and 73% requiring post-harvest facilities (warehouse, dryers, etc.) are served.

For the economic infrastructure, accomplishment in credit provision is at 54%. This can be attributed to the low repayment rate at 44%. Some credit facilities are not sustainable because funds are easily drained due to low repayment rates. The adoption rate of modern agricultural technologies is low at 38%.

Among the foreign-assisted projects, the best profile of supported communities can be observed from the Agrarian Reform Communities Development Program by the World Bank (ARCDPWB), with an average index of 75.96, the highest among the projects as of 2005. This is supported with the highest income of PhP 102,300 per annum per household. One important feature of the project is that aside from the bundles of intervention provided, it implements a cost-sharing arrangement among stakeholders (beneficiaries, local government, and national government).

The Microfinance project has an average index of 75.29, also among the highest, but the average income is lower at PhP 90,915. This is one highly specialized project without diversity in menu, focusing only in credit, with minimal institutional capacity development. Similar is true for Agrarian Reform Infrastructure Support Project (ARISP) focusing on rural infrastructure with some institutional development, usually related to the infrastructure (user's group).

The Belgian Integrated Agrarian Reform Support Project (BIARSP) has a wide menu of interventions, thus the average index is high at 73.02, but income is only PhP 83,940. Although the menu is diverse, provision to a community is not necessarily bundled, spreading resources to several communities.

The Western Mindanao Community Initiatives Project (WMCIP) yields the lowest index (64.03) and income (PhP 72,471). Aside from a limited project menu, it was also implemented by government line agencies, local government units, and non-government organizations. The usual implementers of such projects are development consulting groups with established track records. The package of implementers could be detrimental to the project because of the complex political dynamics they engaged in.

The importance of bundling and diversity of interventions in rural development has been initially illustrated in the household data, and now in community-level data.

Rationale for the Models

The community (ARC) level data (ALDA) is a panel data of four years from over 1,000 communities. Models for overall index of rural development, indices for each of the key result areas, farm income, non-farm income, and total income are fitted using mixed panel models with and without autoregressive errors. Estimation was also done using the hybrid backfitting algorithm.

The empirical models therefore assess the dynamics between rural infrastructure and rural development indexed by a typical household in the community, based on short- and long- term outcome indicators. The outcome indicators are perception (summarized into an index), actual income measurements, or community level aggregate index representing various targets of the development interventions of the government.

The Dynamics of Rural Community Development

The availability of panel data (ARC level over a 4-year period, most recent is 2005) enabled the analysis to be conducted over time and space. The panel models with and without autocorrelation are fitted to investigate the dynamics of community level rural development across the ARCs and over time. The scores of each community for overall index, for each of the key results areas, farm, off-farm, non-farm, and total income were regressed on some infrastructure indicators provided to the community.

Two panel data models were fitted: a ordinary random effects model (results in Appendix 1), and a random effects model with autoregressive errors (results in Appendix 2). Models for income data have poor performance while those for indices yield adequate fit. The generation of income data that is based on some ad hoc procedure could have tampered with data quality, thus affecting model fit. Some statistics on model fit and model performance are summarized in Table 4.1 [ PDF 111.2KB | 1 page ].

The random variation of the indices and income data across the communities are clear in Table 4.1. This could mean that even if similar interventions are implemented, there is still an ARC-specific source of variation that will make the actual outcome different from the one expected from that intervention. The significance of the autoregressive parameter is evidence of the accumulation of outcomes in the key result areas as exhibited by the ARCs over the years. This could mean that the kinds of interventions currently implemented in these ARCs are leading towards sustainability. The basic strategy used by the Department of Agrarian Reform in the delivery of development intervention is a mixture of direct provision and facilitation. While facilitation is used for budget coming from the general appropriations, direct provision and facilitation are used for foreign-assisted projects. The benefit of the facilitation effort is not only limited to cost minimization, but also the building up of the sustainability infrastructure, where the communities are gradually being empowered in the process.

Two sets of indicators at the community level are analyzed: the indices for each of the key result areas and the overall index representing the output level measures as a result of the intervention. On the other hand, average household income at the community level can be considered as the outcome (not necessarily immediate result) of various efforts/interventions including infrastructure and other development assistance.

The constant estimates for each of the models are summarized in Table 4.2 [ PDF 89.4KB | 1 page ] along with the average values of the indicators as of 2005. Considering that panel data is used, the constant can be used as a benchmark or a recurring level over the years. This is the value of the indicator cumulated over the years. The difference between the current level and this benchmark is a result of the efforts specific to the current period, the model specifying the details on who contributed what. Stability/robustness of the indicator to various recurring factors is said to have been established if the benchmark is equal to the current level. The proportion or percentage of the benchmark relative to the current level will give an indication of how stable the indicator is to the fluctuations of various determinants.

Consider farm income where the benchmark is 62% of the 2005 average. This means that new efforts in 2005 aimed to increase farm income resulted in one third more than what was expected (as a result of the cumulative efforts over the years), which is PhP 33,403 among typical rural households. The off-farm income, which can still be agriculture-based but was gained from different land cultivated by the household, has a lower benchmark at 42%. The non-farm income is worst at only 25%. This means that the activities in the past intended to generate non-farm income among rural households in the ARC have not been sustainable enough to expect more recurring income. The non-farm income level is determined mostly (75%) by the current interventions. This high percentage may also reflect problems of design specifics of an intervention that is not necessarily sustainable. Whatever gains were generated for the current year cannot be expected to happen again in the future. For the total income, which is like the weighted average of farm, off-farm and non-farm income, the benchmark is close to half of the current level. This means that as a whole, the income- generating interventions provided among the ARCs still require further tightening to make sure that gains they generate can be expected to recur in the future.

The proportion of the benchmark of overall index is similar to the one on farm income. The fact that farm productivity and income, which have been in the mandate of many departments, have very high proportion (84%) of benchmark can be interpreted as meaning that the interventions among the ARCs are focused mostly on farm income generation. There are less intensive, perhaps not sustainable interventions intended to generate non- farm income, confirmed by the low percentage of the benchmark of economic and physical infrastructure support services with only 21%. Although in terms of welfare (basic social services and gender and development), farm productivity and institutional development (organizational maturity) interventions towards the ARCs have been relatively successful, two key areas still need further improvement. Although land tenure improvement has been the flagship goal of the Agrarian Reform Program, and the Department is indeed relatively successful in land distribution, the problem occurs after land distribution. Land tenure improvements do not recur because there are beneficiaries who are easily tempted by easy cash and often exchange tenure for some money, either for a long period or permanently. Even if more land is distributed, the tenurial profile among the ARCs is not really improving; intensifying efforts towards income diversification would be more effective. However, while there is massive provision of rural infrastructure, maintenance and sustainability is not properly institutionalized before the project is turned over to the beneficiaries or the local government. Furthermore, economic infrastructure activities are scanty (or nonexistent) or incorrectly designed. Even if roads are constructed to a reasonable length, they will not be optimally used without marketing support for both the agricultural produce and the products of microenterprises.

The non-recurring part of the indicators is explained in terms of age of the ARC, tenurial conditions, rural infrastructure, credit, technology adoption, and benefits from official development assistance.

The age of the ARC is the number of years since the community was launched. The mechanism of identifying communities where the agrarian reform beneficiaries belong and helping not just the beneficiaries but the whole community was started in the early 90's. There are ARCs that have been launched only recently. Since intervention is provided for the whole community, it is expected that gains or status of the ARCs will be proportional to the age (on the assumption that the design of the intervention is correct). Farm income (p<0.000), off-farm income (0.001), total income (p<0.000), organizational maturity (p<0.005), and economic and physical infrastructure support services (p<0.000) are all increasing over the years. For every year added to the age of the ARC, these indicators are expected to grow as well. However, land tenure improvement (p<0.000) is declining over the years. This is consistent with the discussion above, explained by the difficulty in ensuring that the beneficiaries will continue cultivating the land and do not forego their tenurial right over it.

There are three tenure indicators used in the analysis. Percentage of agrarian reform beneficiary to potential beneficiaries will indicate accomplishment level of the scope of the program. The percentage of distributed area to total area covered by the program will complement the counts of the number of beneficiaries. Then the percentage of beneficiaries who are still cultivating the land is a measure of maintenance of tenure after land distribution. Expectedly, all the three indicators contributed significantly to the land tenure improvement index. The total income is not affected significantly by any of these three tenure indicators because farm income, which comprises a big part of total income, is also not directly affected by these tenure indicators. Non-farm income, however, is affected by percentage of area distributed, while off-farm income is affected by percentage of beneficiaries to the potential number. Through land distribution alone, it is not really expected that income will increase instantaneously, until other dimensions of agrarian life also improve. In addition to land tenure improvement, indices for other key result areas will also improve, including the overall index. The proportion of accomplishment in area or beneficiary coverage will affect organizational maturity and gender and development indices because the institutional strengthening efforts are intensified in areas where there are more beneficiaries. Economic and physical infrastructure support and correspondingly farm productivity indices are also affected by indicators of accomplishment in land tenure improvement.

Among the types of rural infrastructure usually provided or facilitated for the agrarian reform communities, rural roads exhibited effects on the largest number of indicators. Income from all sources as well as the total income increases for every accomplishment in roads constructed or rehabilitated relative to the demand identified by the communities. Rural roads facilitate accessibility by bringing the community out of isolation and exposing them to development in other areas. This will lead to growth in demand for other infrastructure and other development interventions. The expected mode of delivery of such intervention is not necessarily direct provision, but beneficiaries are willing to cooperate/contribute for as long as access to such needs will be facilitated. Among the indicators, only land tenure improvement and farm productivity did not exhibit direct benefits from meeting the needs for rural roads. Understandably, accomplishment in land distribution is dominated mostly by socio-political and legal factors, while farm productivity is dependent on soil fertility as well as on farming systems. These factors, however, can be enhanced with improved accessibility, thus resulting in indirect benefits from rural roads. Along with roads, bridges are needed to complete the accessibility network in rural areas, so similar benefits from bridges also manifest.

The irrigation indicators used include proportion of the number of irrigations to the identified needs, proportion of the serviced area to the total irrigable area, and proportion of agrarian reform beneficiaries with access to total beneficiaries within the service area. As expected, the indicators contribute to the index on economic and physical infrastructure support. This support also appeared to be an important determinant of the overall index, illustrating its role in the development of the community. However, the indicators do not contribute significantly to any of the income indicators or to indices of attainment of other key result areas. Irrigation will definitely have an indirect effect but the absence of a direct effect on other indicators may be explained by some details of the irrigation system. First of all, many canals are built without using concrete. Earthen canals have short lives and holes can be easily created by rodents and from cracks during dry spells in summer months. Maintenance of the canals would often be a major stimulus of conflicts among users, eventually threatening the sustainability of the infrastructure. For better understanding of the impact of irrigation, some indicators on how many of these irrigation systems are properly maintained may help. Similar results for post-harvest facilities hold.

Credit is as important as any other intervention. All the income indicators and indices except one index are affected positively by the proportion of credit needs met in the ARC. Farm productivity is not directly affected by credit because the loan proceeds must be spent on input procurement or planting before they result in increased productivity.

Adoption of farming technologies also affected almost all indicators, except off-farm income. Trainings on different farming technologies, although they may not individually benefit some households, can generally produce a positive effect for the whole community.

The Effect of Official Development Assistance

The effect of the different projects funded through official development assistance (ODA) could have been accounted for in the previous section since most of the projects included rural infrastructure in their menus. The effects presented here will be isolated as additional contributions of the project. Those presented above can be considered as the pooled effect of various efforts (including ODA) at the ARC level intended to push rural development. Organizational maturity can simplify the facilitation of access to development interventions.

The effects of some ODA-funded projects on target output or outcome (income and indices on key result areas) are explored below. ARISP is a project that was funded through a loan from the Japan Bank for International Cooperation (JBIC) and was followed by a second phase that is still on-going. The project menu of Phase 1 was dominated by rural infrastructure that includes irrigation, rural roads, and post-harvest facilities. There was an institutional development component, but the purpose was mainly the development of viable organizations or cooperatives of users that will ensure sustainability of the infrastructure. Still including the menu under Phase 1, Phase 2 includes potable water system, agricultural development support, and livelihood projects. Total income, farm income, and non-farm income of ARCs covered by ARISP are higher than those not covered. In addition, the overall index among beneficiary ARCs are also higher. This is an example of a comprehensive project that started with infrastructure and expanded to include other support services, resulting in actual income increase among beneficiaries.

The ARCDP project funded through a loan from ADB is also a comprehensive package of development interventions, but with a different implementation strategy, following a demand- driven approach. The project menu also includes rural infrastructure (roads, irrigation, and potable water), land surveying and titling (in support of land tenure improvement), agricultural development, rural enterprise development, community and institutional development, and credit. Total income and farm income of households in the ARC beneficiaries are higher. Although the project menu includes certain elements of non-farm income generation, no significant effect was established, probably because the menu is too diverse. In addition to income, farm productivity, economic and physical infrastructure support, and overall indices are also better among ARC beneficiaries.

Another ARCDP project, funded this time through a loan from the World Bank, started with a menu of rural infrastructure (roads, irrigation, and potable water), community development, technical assistance, and agriculture and enterprise development and in Phase 2, credit was included. The implementation strategy is innovative: cost-sharing between the beneficiaries, the local government, and the national government was required. Because of the arrangement, more ARCs were covered to be part of the project. Substantial dilution in the benefits could have occurred, judging by the fact that none of the income indicators among the ARC beneficiaries appeared to be different from those of the non-beneficiaries. The overall index, though, and those of some key result areas are better among beneficiary ARCs. This is an example of a project where, although the menu is good, spreading it too wide may have watered down the effect.

Finally, the BIARSP project with funding from the Belgian government has a wide range of interventions in the menu, including basic education, agricultural production, primary health care, water and sanitation. The menu excludes rural infrastructure. None of the income indicators are different among the ARC beneficiaries compared to the non-beneficiaries. Although farm productivity and the overall index of ARC beneficiaries are better off, all other indicators show no difference. This project illustrates the importance of rural infrastructure in targeting income increase as part of the development goals.

Some of the key development learning here highlights the importance of rural infrastructure if it is intended to target income increases. With rural infrastructure alone, income will not change, but with substantial support services, infrastructure will be used optimally. Given a good project menu, potential benefits will not be optimized if interventions are spread among too many areas. Although project diversity is good, making project menus exceedingly diverse is too much of a good thing.

Download this Discussion Paper [ PDF 471.2KB| 49 pages ].

[previous chapter] [next chapter]

Post a Comment

We welcome your feedback on this publication. Post a comment. ADBI is not obliged to acknowledge or publish comments and may abridge or edit them before web posting.


There are [0] comment(s) for this entry. Post a comment.

    The views expressed in this paper are the views of the authors and do not necessarily reflect the views or policies of the Asian Development Bank Institute (ADBI), the Asian Development Bank (ADB), its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms.

    Back to Top 
    © 2015 Asian Development Bank Institute.