Chapter 5


2) The value of the works performed or assets created compared to the costs of the
 scheme (e.g. wages, materials, etc.).

There is often a trade-off between both objectives to reach the vulnerable groups and to create
investment in an efficient way through public works programmes. Although productivity and
output under public works is often lower than that achieved under alternative arrangements
(e.g. private contracting), public work schemes can still be cost-effective in the sense that
they provide targeted assistance in cash or kind to vulnerable groups. Often the cost
effectiveness according to criteria 2) could be increased if public work programmes were
equipped with appropriate material inputs and complementary technical assistance.

Only projects which can absorb a large amount of unskilled labour are suitable for
implementation under public works arrangements. Typical examples are projects in the fields
of infrastructure and natural resource conservation, e.g. rural roads and feeder roads
construction, dam construction, afforestation and erosion control. Even if cost-effective, such
projects usually demand large fiscal expenditures, and the assets created are, in most cases,
public goods which do not generate revenue for the government. To reduce government
expense, the costs may be fully or partly borne by foreign assistance, either in form of
budgetary support or, for example, in kind, as food aid deliveries.


Box 5.4: Food wages in Ethiopia

The standard food wage recommended by the Ministry of Agriculture was 3 kilogrammes of grain and 120
grammes of oil per day. This ration covered the daily subsistence requirements of six people, offering some
1800 kilocalories per head. However, there were many deviations from the standard because of constraints on
availability and difficulties in estimating numbers of potential participants.(...)

Food wages were mostly consumed within the household. In Wollo, Kholin (1987) found that 71 percent of
FFW payments were consumed. Admassie and Gebre (1985) derived an average of 85 percent from their
surveys. A comparable figure of 72 percent was found in Gara Godo. The remainder was sold to purchase
clothing and other durables (24 percent) or to pay taxes (4 percent).(...) In Adele Keke 97 percent of the wage
was consumed, with 3 percent sold to purchase salt and sugar. Sales of the food wage during 1988 averaged 137
kilogrammes per household in Gara Godo despite an average price of only 0.2 birr per kilogramme received for
wheat, compared with an average 1987 price of 0.5 birr per kilogramme.

Which brings us to the food wage versus cash wage debate. The evidence that participants would support a shift
in labour-intensive works away from food wages to cash wages (currently much discussed in Ethiopia) is not
strong. Kohlin's (1987) study found that 70 percent of respondents preferred a food wage. Admassie and Gebre
(1985) found that 90 percent voted in favour of food. Only in Doma was it found that 83 percent of sample
respondents preferred a cash wage over food. This was largely because of fears of delays in food payments
resulting from the logistical difficulties of reaching isolated locations.

Given that income transfer is one of the key objectives of such projects, the more important question is: What is
the real value of the alternative wage forms to the participants themselves? Holt (1983) estimates that in the
early 1980s the value of a cash wage was roughly 2.5 birr, whereas the average cash value of an FFW rations at
local market retail prices was at least 2.7 birr. (...) Taking the period from late 1986 to early 1988, Erni (1988)
found that only during the first half of 1987 would a fixed cash wage have had a higher market value to
recipients than the food wage actually received. He concludes that the food wage was more urgently needed and
more popular than a cash wage would have been.

Source: P. Webb, P. and J von Braun, Famine and Food Security in Ethiopia, IFPRI 1994.


- 205 -