Average and marginal propensities to save in rural areas

Average and marginal propensities (APS & MPS) are two very important parameters for analyzing saving patterns. The limited availability of some authentic data from rural areas compels us to rely upon some fragmentary evidence in order to reveal an approximate picture. APS & MPS 1969-79 calculated by QURAISHI (1985:288) are reproduced below:

Table 3: Average and marginal propensities to save in the rural areas of Pakistan (in Rs.)

Average Propensity to Save (APS) and Marginal Propensity to Save (MPS) quoted from different studies⁶⁸ which gave APS figures of 0.364 in 1970 and 0.323 during 1971.

66 Transfer income includes income earned from seasonal migration, both within and outside Pakistan.

Income earned from the first kind of seasonal migration is treated as internal remittances; income from the latter as external remittances;

67 Miscellaneous expenditure in HIES 1990/91 (1991) is almost the same as ceremonial expenditure in the study.

68 See QURAISHI (1985) and other references quoted in the article.

Another study showed a very low APS of 0.178. During 1979, it was 0.590, which shows that an average rate of saving ranges from 0.178 to 0.59 rupees, from every rupee of disposable income.

For MPS, however, Table 3 shows the values of 0.725, 0.170 and 0.680 during 1969, 1978 and 1971 respectively. Values vary in different studies, e.g., a study conducted in 1978 reports a nominal value of 0.680 that is similar to previous results. This means that the saving out of every additional rupee of income earned is in the range of 0.170 to 0.725 rupee.

Table 3 shows an almost stagnant rate of APS and MPS from 1969-1971, because households were having hardly any access to income sources outside rural areas. From 1975 onwards, with the manpower transfer to the Middle East countries and Gulf states, rural areas of Pakistan got an access to new income sources (Table 8). Although this revolutionary increase in household income induced an increase in savings, it did not keep pace with the increase in income. For example, Table 3 shows a sudden rise in APS from 1971-1978, i.e., from 0.323 to 0.178. Even at such a high income level, almost 80 percent of the rural households were not saving from 1971-1979. The rural households’ marginal propensity to save (MPS) was negative for the low income group earning Rs. 271.74 p.m., in 1971-72, and Rs. 600.00 p.m., in 1979. The marginal propensity to save did not show a continuous rising trend with the increase in rural income. The ratio of saving for a household's monthly income level of Rs. 339.1 was 5.3 percent, whereas, at the income level of Rs. 448.8, Rs. 578.8 and Rs. 854.2, the ratio of saving was 2.95 percent, 6.0 percent and 28.6 percent, respectively. The ratio of saving to income was 20 percent at the next higher level of a household's monthly income level of Rs. 1198.6. It was 43.8 percent at the next higher income level of Rs. 2676.2. A considerable declining trend of the savings rate has been observed during the following years. This may be a result of the heavy expenditure on innovative technological items such as improved inputs, irrigation and mechanical devices like tractors, implements and other farm equipment.

The official statistics of Pakistan (e.g., KHALID 1988 & QURAISHI 1985) also supports the above-mentioned facts, „although large amount of remittances in the last decade could be diverted into monetary assets. "In the absence of facts and figures on monetary assets and deposits statistics with rural urban breakup, it is difficult to arrive at a meaningful conclusion. It is believed that about 40% of country's investment is being channelled in rural areas, whereas the share of rural areas in domestic savings is hardly about 15%"

(KHALID 1988:8).

The factors like a higher rate of public investments in agriculture, a constantly increasing flow of concessionary agricultural credit and more remunerative prices of agricultural crops played a very important role tobring a revolutionary change in the household income in the rural sphere, but its contribution towards increasing household savings remained very low. The savings rate was 12.1% in 1985-86, and 14.2% by 1986-87 (KHALID 1988:7). Table 4 shows the net rate of household saving as percentage of disposable income, and Table 5, national savings as % of GNP, which supports the above-mentioned statement.

Table 4: Net Household Savings Rates

Table 5: National Savings Rates

(Rs.)

National Savings (as % of GNP)

Year Overall Public Private National Savings

(as % of GNP)

The reasons why the increased rural income is not resulting in a substantial increase in the savings rate are due to a number of factors, among which extravagance in status-oriented demonstrative consumption and mistrust of financial institutions, which led to the purchase of real estate by way of transfer payments and avoided investment in financial assets, should specifically be mentioned. An acute lack of authentic statistical data further hinders the formulation of any proper strategy for the future. "According to a survey conducted by the Pakistan Institute of Development Economics in 1981 although this increased income has resulted in a tremendous increase in consumption expenditure in the rural areas, little of it is saved. The rate of financial savings, however, has remained particularly low, mostly because of the lack of access to financial institutions. 22 % of migrants' remittances are diverted to real estates, 11.5% to agricultural and commercial investments and only 1.5% to financial assets" (QURAISHI 1985:289).

Besides, the above-mentioned facts about the saving and income situation in rural Pakistan, the IFPRI survey of 1986-89 presents a more detailed analysis of five districts in rural Pakistan. Some of them are as follows:

− Low-income populations face more credit and liquidity constraints and, therefore, cannot fully adjust consumption in the face of negative shocks. Similarly, a tendency to consume income from positive shocks is observed at extremely low levels of income.

Table 6: Marginal propensities to spend out of transitory shocks disaggregated by the direction of the shock

Variable Positiv

Notes: The figures in parentheses are t-values.

a) A positive coefficient of a negative shock implies that when income declines, expenditure also declines.

b) Construction capital and other capital are combined.

* Significant at the 10 percent level.

** Significant at the 1 percent level.

Source: ALDERMAN/GARCIA (1993:45) based on IFPRI Rural Survey of Pakistan.

Table 6 shows the amount of different kinds of savings as a result of positive shocks (increase in income) and then the utilization of this surplus and amount of debt (when it is insufficient) in the case of negative shock. In the case of physical saving, for example, a marginal propensity to save out of a positive shock is 0.149, but households increase their debt by 0.296 when incomes temporarily decline. There exists, in addition, a surprising difference in the tendency to increase ceremonial expenditures when incomes increase, yet not to appreciably reduce them when incomes decline⁶⁹ (ALDERMAN/GARCIA 1993:44).

It is also noteworthy that households in all but the wealthiest quartile increase their consumption with positive transitory incomes but do not decrease their consumption with an income decline.

− The poorest households allocate all of their negative income shocks usually to financial dissavings. Restated in more intuitive terms, these households increase their debt when incomes decrease. They have relatively few assets to sell and, therefore, cannot dissave in terms of physical assets as the wealthier households do.

− Contrary to the popular notion that remittances are primarily used to raise consumption, the marginal propensity to save out of external remittances is 0.84. Most of these savings were either in the form of bank deposits or in paid-off loans.

69 ALDERMAN/GARCIA (1993:44) observe that some ceremonial expenditures - for example - may correlate with negative individual income shocks, but such shocks are not included in the transitory income as measured here.

− Pensions are saved in much the same way as remittances from abroad. Many of these pensions are lumpsum payments on retirement and are apparently viewed as transitory.

− The marginal expenditure on construction out of international remittances is less than it is out of domestic remittances or transitory income. The tendency to use international remittances for immediate construction is not different from that of other transitory sources. It should be noted, however, that the study deals with remittances per se, not retained earnings of returned migrants. Moreover, if remittances are initially banked and withdrawn in a subsequent year, the use of remittances for construction would not be directly apparent⁷⁰.

− Domestic remittances are not treated in the same manner as international remittances and pensions. According to these results, marginal consumption out of local remittances is higher than out of other transitory sources, a result that holds even for low income groups. Nevertheless, these local remittances are utilized mostly for physical and ceremonial expenditure.⁷¹ These remittances appear to encourage financial dissavings in the form of increasing net debt.⁷²

− A comparatively higher coefficient of local remittances for medical and ceremonial expenditures could indicate transfers effected by the local family and friends when household expenses for medical care and ceremonies increase. Local remittances, then, may fulfill a risk-sharing function.⁷³

− Credit may be considered as a subset of the category of financial savings in this study.

More than 90 percent of the households in the survey reported obtaining informal loans in a given year.⁷⁴ Village shopkeepers, for example, extend credit (the functional equivalent of credit cards in developed countries) routinely to obtain an advantage over larger market centers. According to the results of IFPRI Rural Survey of Pakistan 1986/87-1988/89 (ALDERMAN/GARCIA 1993), eighty percent of the annual loans obtained by the survey households were from the informal sector (90 percent of the total number of loans). Of these, 40 percent came from shopkeepers, and 45 percent from relatives and friends.

70 These results on spending out of international remittances are somewhat at odds with previous studies on average spending out of remittances (AMJAD 1986, for example). The study differs, in part, in the method of measurement; the marginal savings are measured here as a part of an entire household accounting, which is more accurate than recalling the spending out of a particular income source that is fungible over all resources. Moreover, financial savings (including net debt) are included in this study, which proves to be a major share of all uses of remittances. The results also differ from the analysis of aggregate data by BURNEY (1987). In part, the difference is based on the types of data used and conceptual issues;

households sales and purchases of assets or contraction of debt may either increase or reduce the individual savings position in accord with the desire to smooth consumption, yet have a small impact on aggregate savings.

71 Remittances are taken as exogenous in this study. There is, however, some element of causality in that relatives may increase transfers when a ceremony occurs.

72 Since the share of informal-sector loan is predominant in the total credit flow, the source of credit may be the same as the source of remittances. Some double counting or even misclassification is possible.

73 See ROSENZWEIG (1988).

74 Although this finding is hardly acceptable to the conventional wisdom regarding credit constraints.

Table 7: Loans obtained from the informal sector, by expenditure quintile Loans Obtained Net Increase of Debt*

Expenditur

Source: ALDERMAN/GARCIA (1993:49) based on IFPRI Rural Survey of Pakistan, 1986/87-1988/89.

Note: The figures in parentheses are standard errors.

* includes formal and informal loans

− Table 7 shows that a substantial amount of the total credit is obtained from the informal sector. During the survey period, however, the net increase in loan (including formal and informal loans) was more than half the value of loans. The percentage of loans taken during a year and that remains outstanding at the end of the year (two months after the rabi harvest) is roughly half as large for shopkeepers as for friends and relatives. This implies that loans from friends have comparatively longer-term conditions or are implicit transfers.⁷⁵

− More than 14 percent of the households are reported to have domestic bank accounts.⁷⁶ The percentage was highest in Attock and lowest in Badin.⁷⁷ The average of all annual gross deposits was Rs. 1,600 in the first year, rising to Rs. 2,000 in the second year, and nearly to Rs. 2,600 (constant) in the third year.