A STATISTICAL ANALYSIS OF THE PRODUCTION RESPONSE OF MALAYAN FISHERMEN TO PRICE CHANGES

Abstract

The purpose of this academic exercise is to employ statistical methods to study the behaviour of the supply and price of fish, as a preliminary, to the more important analysis of the production response of Malayan fishermen to price changes. The basic data of our analysis are the Singapore retail price of fresh fish and the fish landed in Malaya tine series, obtained from "Malayan Statistics".

The study of supply is divided into two sections the estimation of trend and the computation of seasonal indices. The annual series of fish landed, when tested for significance of trend by Kendall's S, is found to possess a significant trend. Mathematical and empirical methods are employed to estimate trend. The former is more precise and objective and the criterion used is the method of least squares. Linear, exponential and quadratic equations are used to fit the trend to the data. The trend of fish production is upward. The empirical trend, given by the 12-month centered moving averages, shows the same characteristic of an upward tread.

In studying the seasonal fluctuations of fish production, recourse must be made to the monthly data. Two methods, the ‘average-of-years’ method and the conventional ‘moving average’ method, are used to compute the seasonal indices, first from the original data and then from the data adjusted for calendar variation. This refinement eliminates from the data the variation of supply resulting from the differences in the number of days per month. We conclude from our analysis that there is a marked seasonality of supply, the peak season being from April to October.

The statistics regarding prices of fresh fish are of many kinds. All suffer from numerous limitations. The Singapore retail price is the series most useful and relevant to our study. The retail price of each type of fish is given but for analysis, we categorise the fish into ikan kurau, first, second and third quality fresh fish.

The annual data of each of the four categories possess a significant trend. The data, when fitted mathematically, reveal a strong falling trend. The empirical trends, when charted, reinforce our conclusion. The conventional 'moving average’ method is used to study the seasonal behaviour of fish prices. Considerable similarity exists in the general movement of the prices of the four categories of fish. Thus the seasons of low prices are from March to June and September to October, while the periods of higher prices are from November to February and duly to August.

In the inquiry into supply response of fishermen to price. Interest is focused on the derivation of the supply curve. The method consists of fitting simple linear or curve-linear regression equations to the historical series of price and supply of fish, regarding price as the independent variable and supply the dependent variable. Price here refers to the Singapore retail price of first quality fresh fish while supply refers to the series of fish landed in Malaya, adjusted for calendar variation.

Various models are constructed to study the response of fishermen to price and the null hypothesis is that the Malayan production of fish is not responsive to price fluctuations. Our models are mostly of the form St = a + bPt, where St is the total monthly production of fish and Pt the monthly average price in period t. The models are tested for significance by analysis of variance. In essence, this tests whether or not the regression coefficient, b, differs significantly from zero.

She variation in fish production is vary significantly explained by the linear regression of S on P when no time-lag la allowed, as well as for the models with one-month, two-month and three-month time lags. The variation in supply is also significantly explained by the quadratic and Cobb-Douglas models. A "peculiarity" exists, however, is that the regression coefficient is negative. This, in our supply equations, suggests a perverse relationship - that the fisherman produce more in periods of low price and less in high-price periods. This interpretation about the 'irrationality’ of Malayan fishermen is, however, not conclusive because of the many limitations inherent in our simple models, end needs to be supported by further empirical evidence.

The limitations suffered by our models may be divided into statistical and non-price factors. Statistically, the time-series available are too short for reliable conclusions to be dram. Also, the inability to compute a general 'price’ series reliable enough for regression analysis constitutes a further limitation. The non-price elements which affect variation is supply but which have not been eliminated from the data include the effects of technological improvements and mechanization, the lack of knowledge of the precise length of time-lags, the influence of input costs and prices of competing products on fishermens' response decision and the influence of biological, meteorological and related environmental factors. These external factors 'disturb’ our analysis and limit its applicability. They must therefore be noted in order to appreciate the significance, reliability and usefulness of our models, analysis and conclusions.