Previous Page Table of Contents Next Page


2.1 Methodology

The cost - benefit analysis at individual farm level was implemented with the following objectives:

Since this type of cost-benefit analysis was carried for the first time in the project area, it was decided to interview as many farmers as possible who could provide useful data for this survey. Therefore the sample of farmers was not picked at random. On the contrary, an effort was made to identify individuals who could answer reasonably well to the questionnaire. The possible impact of this choice is a sample biased in favor of farmers who keep records of their operations and therefore would be better managers and fish-farmers.

Therefore, the present exercise should be considered as frame survey which could be used as reference if a broader socio-economic survey was required. A more immediate use of the present survey is to assess the profitability of fish-farming for the individual farmers, in order to provide basic economic information for project management.

2.1.1 Questionnaire

The questionnaire used in the survey covered the following elements:

General information:



A specimen of the questionnaire can be found in Annex III.

Questions were oriented more towards physical quantities of inputs and outputs, prices and unit costs, rather than cash outlays. The reasons were that many expenses are non-cash and therefore had to be estimated based on physical quantities and some type of unit cost (ex: value of time spent by owner, value of inputs collected from the farm…etc.). In addition, the farmer could purchase items like feed and fertilizer, which could be used for the fish-pond as well as for other uses. Then in the course of the survey, it appeared that many farmers could well remember the amount of their sales and the different purchases they were making for their fish-farm. The same can be said for the opportunity cost of land, which is the net income they could get from the best alternative use for their fish-farm. As the survey was progressing, questions were therefore asked concerning all purchases. The answers could be more directly included in the income statement account after counterchecking with answers to questions on consumption of inputs.

2.2 Analysis of results

2.2.1 General information


All fish-farms surveyed were individual fish-farms save for one which had been built and operated by a women group (first interview). As womens' group fish-farms are community-based activities which are not operated on the basis of economic efficiency, it was decided to do other interviews with individual fish-farms only.

All other farms were operated by their owner, except for two where a manager was in charge of farm activities.


The sample included farms from the following districts:

Nyanza province:
Kisii (4), Nyamira (3), Migori (3);

Western province:
Siaya (3), Busia (3), Bungoma (4), Kakamega (3).

Number of years in fish-farming

On average, fish-farmers surveyed had seven years of experience in fish-farming, yet the median for the number of years of experience was five years, as the group included individuals who have been in the trade since 24 and even 34 years.

Distance from home

Most farmers surveyed live relatively close to their ponds. 14 live at 5 minutes walk or less, 5 at 5–10 minutes walk, 3 at 10–15 minutes walk and in one case, the farmer lived at one hour walking distance from his home. Therefore, in most cases, commuting between farm and home does not appear as a constraint for good fish-farm management, since in addition, the ponds are generally next to his other cultures.


Ponds visited were fully drainable, with one or two exceptions.

Water, soil, terrain conditions

All farms visited had ample water supply all through the year. The only problem related to water supply were occasional floods which caused destruction of dikes and escape of fish for a few farmers. 11 farmers found no difficulty in building their ponds. Others faced problems of rocky soils (3), hard soil (4), or swampy, muddy soils (4). Only one faced problems of seepage.

2.2.2 Construction

As explained, the survey asked farmers to answer questions on the construction costs for their most recent construction programmes. Out of a total of 23, 9 farmers built their ponds themselves, 8 hired laborers paid on a daily basis, 4 had given the job to a contractor on fixed term. In one case, construction was achieved through the efforts of community work. In two cases, no answer was provided as the owner himself was not available for answering questions.

Community effort dwells on participation and at least in this case, involved many more people than would be normally required. Therefore at the present daily wage rate, the ponds built through community effort would come to a total cost of 7298 K.Sh. per m2, a very high figure indeed, but which cannot be interpeted in a strictly economic sense (see Table 2.1).

Ponds have been built at different periods. Therefore, in order to be able to compare construction costs from one site to another, on-going daily wage rates have been applied. This could not be done however for those cases where contractors were hired, for which total cost figures were given.

Daily wage rates vary between 15 K.Sh. and 60 K.Sh. Strong variations can be seen even within the same district, depending on the labour market conditions. Rates decrease when moving away from big centers where demand for labour is higher.

As for the number of man/days required for building ponds, it also varies considerably. Figures given can be as low as 10.3/100 m2 which seems abnormally low and as high as high as 184 m/d/100 m2. In total, 6 farmers have stated figures which would imply less than 50 man/days for building a 100 m2 pond, out of a total of 18 for whom such figures could be computed. Those farmers had all built their pond themselves or with members of their family, with the exception of one who had used hired labor.

Construction work carried by hired labor generally required more manpower, with the exception of the case mentioned above. The lowest manpower use was 51.4 man/days/100 m2. In all other cases, uses of manpower exceeded 100 man/days/100 m2, going from 104.5 m/d up to 171.1 m/d, with 5 cases between 104.5 and 113.3 m/d.

Excluding the women group fish-farm, the average cost for all ponds was 2943 K.Sh./100 m2. For construction given to contractors, the average cost was 2725 K.Sh./100 m2; for family built ponds, the average cost was 2533 K.Sh./100 m2, while it was 3513 K.Sh./100 m2 for construction using hired labour. It must be assumed that farmers do not record use of their own time as well as use of their money, which would explain at least in part some very low figures given for manpower requirement in pond construction.

Out of a total of 23 farms, 16 were not protected with fences, and among those equipped with fences, most were only partly enclosed. Fences have been installed to keep cattle from the pond site or to prevent theft. Different types were used:

Theft was not identified as a serious problem among farmers interviewed. This fact does not infer that fences are not justified; yet, it was found preferable to leave this element out of the cost-benefit analysis.



Fish-farm identification number1234567891011121314151617181920212223
Number of ponds3131234591033344134225319
Total area (sq. m.)5316040846624631359211451800447050831048070575624555012505003607587433805
Type of workCommunity workOwnerHired LabourHired LabourOwnerOwnerContractHired LabourContractContractOwnerOwnerOwnerOwnerHired LabourHired LabourHired LabourHired Labour--OwnerHired LabourContractOwner
Area of last pond built (in 100 m2)
Peak size of work team306101112--44--114131565--36321
Number of wheelbarrows302202162--01101523--12----
LABOR INPUT (man/days per 100 m2)                       
Compost crib16450.312--1----10.321--0.444--62----
TOTAL MAN/DAYS (per 100 m2)291.918451.4171.135.6142.5--111.9----2710.3126.519.1109.9104.510812.8--15.7113.375210
Unit cost of labor (per man/day)252543305024--40--3515303540154042.560--50456030
COST OF LABOR (per 100 m2)7297.54600221251331780342012504476265025004053094427.57641648.441804590768--7855098.545006300
COST OF CONSTRUCTION3875027609025239204379107057400512504770011175020579582125253861246210241252459600--28263864733435239715

2.2.3 Production

In order to estimate the costs and benefits of fish-farming operations, the questionnaire inquired about the yields and income obtained from at least one recent harvest.

Fish species

All farmers were producing O. nilotica except one who had started in 1967 with T. zillii and did not want to change. Another farmer was practicing polyculture with O. nilotica and catfish captured from the wild. This technique had just been recently applied and no harvest had yet been recorded based on using polyculture.


Only two fish-farmers were stocking with monosex, all others were using mixed stocking. The standard for stocking density is 2/m2 but this standard is very loosely applied. One factor was the practice of selling fingerlings by weight, which was changed only in February of this year. In general, fish-farmers met by the survey team did not seem very convinced of the importance of keeping the stocking standards set by the project.

Production cycle

Fish-farmers in the project area follow long production cycles (the data on this subject and other items covered in this section are taken from Table 2.6). Two farmers had harvested after 6 months, 11 between 7 and 12 months, 3 after 13 to 18 months, 2 between 19 and 24 months and three had cycles of 5 years or more (in two cases the length of production cycle could not be estimated).


In addition, most farmers practice partial harvests, leaving fingerlings in the pond to harvest at a later time. Harvesting is done when convenient, not when optimal yields can be expected. The results are very uneven yields. As many as 8 fish-farmers had yields inferior to 2t/ha/yr, 5 had yields between 2 and 3 t/ha/yr, 7 had yields between 3 and 4 t/ha/yr and one had reached 4.8 t/ha/yr (yields could not be estimated in two cases). The highest yields were reached in a farm where ponds were stocked with males only.

It must be explained at this point that except in the Siaya district, data on production had to be obtained from the farmer himself as the extension staff in most districts do not keep proper records which would allow to measure yields on a farm by farm basis.

Price of fish

Price of table-size fish has been found to be the most important factor of variation in net incomes among fish-farmers. It had to be computed by dividing income from sales by volume of sales, as given by the farmer. Sales prices for the most recent harvests were distributed as follows:


Price of fish
Number of farmers
0 – 104
11 – 205
21 – 304
31 – 402
41 – 502
51 – 601
61 – 702
71 – 800
81 – 901

Sales price is thus clearly the most important element to be assessed when analysing the cost-benefit of fish-farming in the project area. Even on one farm, prices will vary widely from one harvest to the next. The factors behind those important price differences deserve to be investigated in a more thorough way. One thing must be reminded is that the net income a farmer can derive from his work of many months will largely depend upon his capacity in efficiently marketing his harvest, fish being a perishable commodity that needs to be sold within a few hours. Extension work usually put much emphasis on yields, assuming that fish can be sold at a reasonable price wherever there is a demand. However, demand will meet supply provided that information on supply has been transmitted through existing communication channels. This element is not well controlled and the fish-farmer will sometimes be forced to sell at low prices because few people show up at harvest time. This can be due to a weak demand which would put into question the feasibility of fish-farming in that area, but quite often it is caused by insufficent marketing.

Commercial aspects of fish-marketing need to be given more consideration as for all farmers save one, income derived from fish-farming through sales is far more important than income from self-consumption.


Income from fish production was obtained by applying the yield estimate to the total area under water and multiplying by the price figure. It thus includes sales and self-consumption, valued at the sale price. In two cases, yields could not be estimated and sales were computed from the sales figure given for the last harvest.


Farm NoTotal area
Total Income
Income per are

Although no farm exceeds half a hectare, there are sizeable differences between sales volume, not only because of differences in farm area but also because of differences in sales per are. Income per are can be as high as 2000 K.Sh. and as low as 68 K.Sh.

Differences in sales per are occur mainly because of differences in sale prices but also because of differences in yields (see Table 2.6).


All fish-farmers interviewed fed their fish with either agricultural by-products they would buy, or some spoiled food they would collect from their farm. Most farmers spend money for feeding their fish, either through their purchases or through the use of marketable products from their own farm. Only in three cases farmers manage to feed the fish without spending money, by using spoiled maize.

Maize bran was the most commonly used feed input (7 farms), mostly in the Nyanza province, together with maize meal, also used in 7 farms. Dairy meal (3 farms) and Growers mash (2 farms) were also used, as well as rice bran (3 farms). Those inputs are complemented by bloodmeal, fishmeal (omena), kale and kitchen remains. Many farmers also use as feed weeds, cut from their lot.

Ranges of prices payed for these inputs (except weeds) are as follows:


ProductPrice range
Maize bran2.5 – 7
Maize flour5.6
Maize meal2 – 5.33
Maize meal (spoiled)3.75
Maize flour (sweepings)2.8
Dairy meal2.97 – 6.22
Grower's ma sh4.67
Rice bran1.60 – 5.33
Omena8 – 12
Butchery waste0.4 – 1
Brewery waste0.5
Kitchen remains0.5

Feeding ratios tend to follow a fixed pattern of giving one tinful of feed, whether it is maize bran, rice bran or other, two times per day, using the 2-kg container of cooking fat. Only a few farmers recognize that feeding ratios need to be adapted to fish-size and number.

This is complemented by other feeds higher in protein-content, like bloodmeal, kitchen remains, omena, etc. No data was available concerning conversion ratios for those different feeds. Yet the prices given for these different feeds, when compared to the prices at which most farmers manage to sell their fish, suggest that the farmers would have to achieve very good conversion ratios in order to even cover the cost of feeding with the inputs they are using. In fact, the cost of feed exceeded the estimated total value of production in 9 cases out of the total of 23.


Two farmers in the group surveyed used chemical fertilizer, namely DAP. In both cases it was used as supplement to direct feeding. The price paid for DAP was 300 K.Sh./50 kg bag in one case and 490 K.Sh./50 kg bag in the other. In both cases, the cost of DAP represented 5% or less of the total cost of purchased inputs.


Fish-farmers surveyed use “jembes” (hoes) and shovels, slashers and “pangas” (machetes), which cost less than 100 K.Sh. per piece and last each a little less than a year. About half the fish-farmers stated that they used a wheelbarrow for their operations, mainly for fertilizing. A wheelbarrow costs between 800–1000 K.Sh. and lasts around three years. However, those tools are also used to varying extent in other farm operations and the cost of using them cannot be fully allocated to fish-farming. Therefore, an allocation of 200 K.Sh. has been set for the cost of supplies, for each 1000 m2 of ponds (200 K.Sh. for 0–1000 m2, 400 K.Sh. for 1001–2000 m2, etc)


Labour is required for feeding, fertilizing, maintenance, and finally, stocking and harvesting. The amount of labour required, as estimated from the answers given by the farmers, varies from 6 man-days/are/year to 52 man-days/are/year.


Number of pondsNumber of man-days per are
  118 – 49
  212, 52
  36 – 41
  48 – 50
  56, 25

All pond farms visited were operated by one individual, whether the owner himself or a pond manager (3 farms), except at harvesting time when additional help is required. Also, outside help might be hired for simple maintenance work, mainly cutting the grass. The answers given, suggest that a larger scale of operations will bring savings in labour inputs, although variations are quite important within a given size of operations.

The most labor-consuming activity is feeding, requiring on average 55% of total labour, with a peak of 84% and a minimum of 16%. Fertilizing requires on average 13.6% of labour input, although in this case, the range is very high with a minimum of 1.5% and a maximum of 63.2% (standard deviation: 17.6).

In the cost-benefit analysis, labour has been valued at the on-going daily rate, as it was the most readily available measure for the cost of labour. This approach leads to an overestimation the cost of labour. For instance, on one farm where both rates were given, 40 K.Sh. was the daily rate while the monthly rate was 600 K.Sh., only 15 times as much. On another farm, the monthly rate, 1000 K.Sh. was 16.67 times the daily rate 60 K.Sh. In another case, the monthly rate (600 K.Sh.) was only 10 times the daily rate (60 K.Sh.). Considering that fish-farming requires labour input 7 days per week and it is being done mostly by the owner or his family on an on-going basis, it was decided to value the cost of labour at ½ the local daily rate.

As explained in 2.2.2, daily wage-rates vary between 15 K.Sh. and 60 K.Sh. It must be kept in mind that in most cases, this wage is an income that the farmer pays to himself. It must be taken into account since the farmer could be using the time and effort he is putting in his ponds in some other useful activities.

For the group surveyed, the cost of labor amounts in average to 22% of total operational costs, with a minimum of 6% and a maximum of 52%.

2.3 Financial results

Initial investment

In order to take into account other elements that the farmer has to purchase initially in order to start fish-farming (pipes, initial stocking of ponds, etc..), a 5% margin has been added to the cost of construction in order to estimate the initial investment.

Gross margin

Gross margin is defined here as sales income minus purchases of fingerlings, feed and fertilizer. For farmers who operate their fish-farm without outside help, this is an important performance indicator.

The survey shows that, if the results from the last harvest are applied to the whole farm, 14 out of the 23 farms surveyed (61%) have a positive cash balance after deduction for outside purchases. Obviously, the 2 fish-farms who do not buy feed or fertilizer have a positive gross margin. Also, farms who have high yields and/or high sales price will have bigger gross margins.

Cost of operations

The cost of operations includes total purchases and the cost of supplies and labour.

Profit on operations

Profit on operations is defined as income from sales and self-consumption minus the cost of operations. 8 farms out of a total of 23 (35%) show a profit on operations.


Depreciation is taken on a straight line basis, over a period of 20 years. It is based on the whole initial investment made the farmer.

Net profit

Net profit is obtained after deducting depreciation from profit on operations.

What was found was that six farms show a net profit. Four out of those six farms had a sales price per kg of more than 50 K.Sh. Another one (No.23) although selling at only 17 K.Sh./kg, shows good yields (3.2 t/ha). No.9 did was selling at 19 K.Sh./kg only but his “input-output coefficient” (ratio of purchases to value of sales), at 0.31 is very favorable. High sales prices cannot guarantee positive financial results. Farm No.4 and farm No.18 also had high sales prices but their net profit was negative because of low yields and high feed costs respectively.

Rate of return

Two measures of financial return have been computed. The first is the percentage of net profit over fixed assets. It represents a crude measurement of the return on investment for the farmer, taking into account what he has invested in the construction of his ponds. The second is more comprehensive and includes the cost of working capital, estimated at half the value of the operating costs.

Of the six farmers who have a positive net profit, two manage to have rates of return of over 10% (for both measurements). The others can only be said to be marginally profitable, as their rates of return are inferior to the rate of inflation. In other words, they are not profitable enough to justify further investments, based on strictly financial considerations.

Cost of land

The analysis has not put a cost figure on the use of land. It would have to be taken into account if farmers had had to pay a rent, which was not the case. There is nevertheless a cost in building ponds where other crops could be grown, which is the maximum net income that could be realized if the pond site was allocated to its best alternative use. The farmer would not be justified in allocating his lot to fish-farming if his net profit did not exceed the net income he could derive from allocating that same lot to its best alternative use.

All but one of the farmers surveyed stated that they could put their pond site to an alternative use, generally a crop requiring much water. Three farmers gave estimates of the net incomes they could derive from alternative uses for their fish-farm:

No.20:1,500 K.Sh./yr(360 m2 water area)
No.21:21,000 K.Sh./yr(758 m2 water area)
No.23:16,000 K.Sh./yr(3,805 m2 water area)

Not having sufficient data the cost of land has not been included in the cost-benefit calculations. It nevertheless would have to be taken into consideration in further analysis.


Fish-farm identification number1234567891011121314151617181920212223
Starting year1986196719851988198919861985198819891989198819751983--199019901990198619871982198619801957
Number of ponds3131234591033344134225319
Total area (m2)5316040846624631359211451800447050831048070575624555012505003607587433805
Initial investment (K.Sh.)406872898947625116459811240777053813500851173382160100622315565613085107532650710080407929674057935107251701
Yield of latest harvest (t/ha/yr)
Length of cycle (months)16>60142819576111261191111--9138231212>608.5
Price of fish (K.Sh./kg)8--262.5503485919302781570203025506016372517
INCOME (TABLE-SIZE FISH)1062027717484920851110701237786624138205784334561135117204533357519375330018438694111520699
INCOME (FINGERLINGS)000000000026000002001800000000
Dairy meal0000000002271100000360001456000000
Growers marsh0000000000000000000084000
Maize bran82122430156018002433000000125000000010220000
Rice bran000000000--0000000010950048001600
Maize meal000000324413557240003483000003500--011200768
Blood meal5460000000000000000000000
Total purchases875824301560180024330324418761240022711378312519203100400464013954145601095010220244002368
GROSS MARGIN-7696-2430-1283-5224878517826-1752454661427-146671815368251132093-85794815-7650-83776254111518331
Stocking and harvesting1178329206--1526252152723--20251148
TOTAL NUMBER OF MAN/DAYS32265585302764686306781144356192120224978818619272211
Unit cost of Labor (per man/day)2525433050243240383515303540154042.5605550456030
TOTAL COST OF LABOR400325118312757503241024136011970503121525207120144024004760291024204650432021603165
COST OF OPERATIONS93582955294332753383524446820521399723711448615404640104202040724018914178701357015070696023606333
PROFIT ON OPERATIONS-8296-2955-2666-152715373276602-192843869427-2169-697-1184931-320-2507-135391505-10270-132271734-124514366
NET PROFIT-10330-3100-3140-27831307-2356213-219751365-5440-2277-747-2300648-974-3045-148641001-10474-13375-295-30001781
PROFIT OVER FIXED ASSETS (%)-25.4-107-33.1-11.128.4-2.180-40.82.7-4.6-105.4-74.3-10.311.5-7.4-28.3-56.19.9-256.8-450.8-0.7-8.50.7
PROFIT OVER TOTAL ASSETS (%)-22.8-70.8-28.7-10.420.8-262.1-34.32.6-4.2-51.7-42.1-9.36-6.9-21.2-41.35.3-96.4-127.4-0.7-8.30.7

Previous Page Top of Page Next Page