A paper based on the Instructor's Resource Kit, Module 1, Managing the Implementation of Development Projects, by Jerry Brown under the direction of John Didier, World Bank Institute, Washington D.C., 1998. Robert Youker may be reached at bobyouker@worldnet.att.net and John Didier at jdidier@worldbank.org. Published here April 2003.

## Horizontal Logic

So far, we have talked about the vertical, up and down, logic of the hierarchy of objectives and established a Why-How, Ends-Means, relationship between its levels. However, there is also a horizontal, left-right, logic that specifies what "outcomes" the project is to achieve at each level and this makes clear what assumptions are being made at each level. These assumptions can be thought of as "if-then" relationships. For example: "If the new farming practices are effective, then the yields on rice production should increase enough to cover the added cost of the inputs and also provide an increased profit to the farmer."

This horizontal logic is made explicit by adding two columns to the hierarchy of objectives table. These are: "Measurable Indicators of Results" and "Assumptions". Items in the measures column make it easier to evaluate progress during project implementation as well as the impact after implementation i.e. during the following operations. Items in the assumptions column help people to understand the "conditions", i.e. the if-then relationships that must exist for the project to achieve the higher-level objectives.

The horizontal logic for the example of the rice project is shown in Figure 2.

##### Figure 2

By making explicit their thinking about measures and assumptions, the project teams are giving reviewers, including the potential beneficiaries, a better opportunity to understand their thinking and to help them avoid mistakes. For example, the planners are assuming that the price of rice will not fall as production, i.e. supply, increases. How realistic is this assumption? If there is considerable reason to doubt it, then the proposed project must be considered highly risky. Perhaps a less risky alternative should be found?

Or perhaps the objective of increasing farmer income by \$x,xxx is too optimistic if the assumption is faulty? Maybe a \$yyy increase is more realistic? Further, related activities and conditions must also be taken into account. For example, will storage and milling capacity have to be increased to accommodate larger yields and will rice marketing need to be expanded to prevent a supply glut?

The big benefit of the hierarchy of objectives is that it makes this type of discussion possible. So long as the hierarchy is not too complicated with too many levels and columns, it provides a basis for analysis that policy makers, beneficiaries, and other stakeholders can discuss and jointly shape a realistic project. Perhaps that is why the technique, which was originally developed in the 1970s, is again becoming popular as planners seek better ways of involving stakeholders and developing practical solutions that meet real needs.