Off-take of Food Grains from PDS and Its Determinants
5.9 Testing the Hypotheses-Multivariate Analysis
This section tests the casual relationship between the factors listed in Section V and the PDS lifting of food grains by BPL cardholders. Forming a quantitative supply side constraint, required for hypothesis testing is rendered difficult by the fact that the sample data did not suggest the presence of a pure supply side constraint pre-empting BPL lifting of PDS grains in any selected
district; irregular/non–demand for PDS cereals over a period coupled with infrastructural deficiencies and other weaknesses in the delivery system served as a disincentive to regular lifting by FPSs, eventually leading the PDS activity to a halt. Besides, the accepted criterion of not even a single household not drawing ration from a village as a measure of non-coverage of the village under PDS (Dutta and Ramaswami, 2001) cannot be applied to the PEO sample as the sample was drawn from the card assignment register of the selected FPSs (in the rural areas). Owing to this, the sample drawn from areas with extreme supply- cum-demand constraint has been explained in Section III as outliers and are omitted from the purview of hypothesis testing here.
Two equations are estimated here. Each equation tests a separate hypothesis; but both are explained together because of their being closely related. Equation (I) tests the expected relationship between the decision to buy or not to buy from the FPSs and its determinants while equation (II) tests the causality between the factors mentioned in Section VI and the monthly lifting of PDS grains averaged for those households, which reported PDS grain lifting. Such averaging helps in correctly fixing the factors behind determining the quantity to be lifted, once the decision has been made to purchase some food rice/wheat from PDS.
%PD=95.04+0.116(INS)–0.362(OWN)–0.142(ASTS)–0.165(PRE)+0.051(PR)–66.11(D)
(12.85) (1.91) (-1.67) (-2.08) (-2.42) (1.68) (-6.09)
Equation (I) R² = 0.62
d.o.f = 43
(Figures in brackets are t values).
(All variables are averaged for sample households at the district level.)
%PD = % of BPL respondents lifting PDS grains averaged at the district level.
INS = % of BPL respondents allowed to buy PDS grains in installments.
OWN = % of rice and wheat requirements of BPL respondents met out of own production and kind payment of wages.
ASTS = % of BPL respondents possessing specified assets.
PRE = % of BPL respondents preferring the local variety of rice and wheat strongly to their PDS variety.
PR = The ratio between the weighted average of market price of rice and wheat and the weighted average of the PDS price of rice and wheat. The weights employed are the normative entitlements to rice and wheat to a BPL cardholder; when the entitlement is only for one of the two grains, the weight for the other becomes zero and the average becomes a simple average for the entitled grain.
D = A dummy variable for the presence of two outlier observations in the dependent variable.
Qpd =-0.79+1.037(ENT) +0.020 (INS)–0.067(OWN)+0.003(ASTS)-0.002(PR)-0.031(FS) (-0.23) (6.91) (1.99) (-1.59) (0.189) (-0.3) (-00.54)
Equation (II) R² = 0.61
d.o.f = 43
(Figures in brackets are t values).
Qpd = Average PDS lifting of cereals by those BPL respondents reporting such lifting.
ENT = Average district-wise entitlement to PDS food grains to a BPL cardholder, not adjusted for supply side shortages.
FS = Ratio of number of households with size less than or equal to 2 in the total sample.
(The other variables are already explained).
In equation (I), all representative determinants, except the presence of foreign particles in PDS grains (omitted from the equation presented) are statistically significant at least at 10% level. The presence of foreign particles in PDS grains was strongly and positively correlated with the variable, PRE (% of BPL respondents preferring the local variety of rice and wheat strongly to their PDS variety), indicating that the adulteration/low quality of PDS grains got reflected in the preference for local varieties of grains. Equation (I) suggests that while INS and PR affect the decision to buy from PDS (%PD) positively, ASTS, OWN and PRE affect it negatively. Since the explained and the explanatory variables (except the dummy) are given in percentages, the regression coefficients serve as elasticities and are readily amenable to policy conclusions.
Equation (II) strongly suggests that once the decision to lift food grains is made, the quantity to be lifted is predominantly determined by the supply side factors, mainly, the quantity of food grains for which the household is entitled. This points towards the importance of having a clear policy towards fixation of
household entitlement and offering them to lift grains in convenient installments, which is crucially related to the viability of the delivery system, especially the FPSs. All the demand side factors turned out to be very weak in determining the quantity to be lifted, except, perhaps the contribution of own production and other non-market sources of cereals in the household’s cereal basket. To explain the equations together, for instance, the ownership of designated assets (ASTS) (suggestive of error of inclusion) affects the decision to buy from PDS or not (Equation (I)); but once the asset owner decides to buy from PDS, he buys as much as the poorer BPL cardholder does (Equation II). Again, the ratio of market price to PDS price affects only the decision to buy or not (Equation (I); it is weak in determining the quantity bought from PDS (Equation (II). Both the results are explicable; it is after thoughtful consideration of the open market alternatives (in the case of asset owner) and the narrow wedge between the BPL price of PDS grains and their market price in the case of grain surplus areas (PEO field notes & Srivasthava 2001) that the cardholder decides to allocate some quantity of his market purchase to PDS; but as the quantity to be bought from the market is as high as 90% of the total requirement (sample average), the BPL cardholder does not hesitate to buy almost 28% of that from PDS (sample average).