Chapter

10 Government Spending, the Real Interest Rate, and Liquidity-Constrained Consumers’ Behavior in Developing Countries

Editor(s):
Mario Bléjer, and Ke-young Chu
Published Date:
June 1989
Share
  • ShareShare
Show Summary Details

Nicola Rossi*

I. Introduction

The importance of the mobilization of domestic savings in developing countries can hardly be understated, given the present state of international capital markets. Unfortunately, however, the understanding of the actual determinants of domestic savings in developing countries is still scanty. A number of empirical investigations (mainly reproducing the literature relating to developed economies) have been carried out in recent years, but the paucity of reliable data has made it difficult to test the underlying hypotheses and obtain results that warrant a reasonable degree of confidence.

The purpose of the present paper is to provide additional—and, it is to be hoped, more conclusive—empirical evidence on this topic. In order to allow a comparison with the most recent work on the subject, I shall estimate Euler equations for the representative consumer’s stochastic dynamic optimization problem. However, in contrast to much of the previous literature, the theoretical framework for the present paper is based on the consideration that some of the basic assumptions on which saving functions for developing countries have been estimated may not be entirely realistic. In particular, a significant fraction of the population in developing countries can be expected to be affected by liquidity constraints that substantially diminish consumers’ ability to substitute consumption intertemporally, as is assumed by the well-known life-cycle theory. This is attributable to a number of factors, including capital market imperfections. While the gravity of capital market imperfections continues to be a matter of debate, even in countries with apparently sophisticated financial institutions and well-developed capital markets (see Hayashi (1985) for a review and also Hubbard and Judd (1986)), this phenomenon has never been explicitly accounted for in developing countries, although there are several reasons why such imperfections are likely to be exacerbated in those countries (Blejer and Cheasty (1989)). Allowance is therefore made for departing from optimal behavioral rules by the representative consuming unit described by the theory.

In the context of a theoretically plausible model of consumer behavior that allows for borrowing constraints, two major issues are addressed.

First, the long-debated issue of the real interest rate elasticity of savings is taken up again. As is well known, the effects of the rate of return on the level of savings and the rate of capital formation are of central importance to both economists and policymakers, since they bear on a number of the central questions of macroeconomics. The relevance of the interest rate elasticity of savings is further enhanced in development economics, where some of the competing views on the role of financial conditions in the economic growth process rely crucially on the degree of responsiveness of aggregate savings to changes in the rate of return.

Notwithstanding considerable research to determine the interest rate responsiveness of savings behavior in developed economies, the traditional view that changes in the rate of return are likely to have only a minor effect on the savings rate holds (Modigliani (1986), p. 304), though controversy still exists (Summers (1984)). In the case of developing countries, the lack of empirical work on the responsiveness of savings in the 1960s and early 1970s is emphasized in the surveys by Mikesell and Zinser (1973) and Snyder (1974), who describe the evidence as sketchy at best.1 More recently, further attempts have been made (Fry (1978, 1980); Giovannini (1983, 1985); McDonald (1983); and Pereira Leite and Makonnen (1984)) to explore this relationship. However, they can be questioned on the basis of their limited geographical coverage; the unreliability of available data; and, in some cases, their underlying methodology. Consequently, their results cannot be used with confidence for policy analysis.

Fry (1978, 1980) estimates a (national) savings function for seven Asian countries2 for 1962-72. He apparently finds strong support for the hypothesis of a negative real interest rate elasticity of domestic consumption. He estimates this elasticity to be about -0.2. Similar conclusions are reached by McDonald (1983) and Pereira Leite and Makonnen (1984). McDonald focuses on factors determining saving behavior in 12 Latin American countries3 and provides evidence of a negative relationship between the real interest rate and private consumption in most of the countries examined of a magnitude roughly comparable with that found by Fry. Pereira Leite and Makonnen’s study, on the other hand, concentrates on six African countries4 and provides evidence of a limited, but positive relationship between private savings and the real interest rate.

The hypothesis of a positive and significant relationship between real interest rates and savings in developing countries is questioned by Giovannini (1983, 1985), who replicates Fry’s results and shows that “the apparent empirical success of the high interest elasticity hypothesis depends in a crucial way on the presence in the sample of a few observations that have a disproportionately large influence on the estimated” response of savings to the real interest rate (1985, p. 199). In this work, Giovannini extends the analysis to 18 developing countries5 and, bypassing many of the econometric problems with aggregate-savings equations, estimates the elasticity of intertemporal substitution in (private) consumption. Using annual data, Giovannini finds that in only 5 out of 18 countries is the intertemporal substitutability in consumption not likely to be very small,6 therefore implying, other things being equal, that the interest rate elasticity of savings is positive.

Giovannini’s work represents a considerable improvement in the knowledge of savings behavior in developing countries. However, it can hardly be considered conclusive. First, his main result relates to the difficulty of obtaining precise estimates of the relevant parameters. In 11 out of 18 cases, the coefficient of intertemporal substitution is positive, but with standard errors so large as to permit one to draw any sort of conclusion. Since Giovannini’s sample period in most cases covers the 1960s, this result is not unexpected, given the very low variability of real rates in that period. Second, as far as geographical coverage is concerned, Giovannini’s work does not provide evidence for those regions for which evidence is most lacking—that is, Africa and the Middle East.7 Third, as Giovannini points out, some assumptions under which the elasticity of substitution is estimated in his 1985 work may not be realistic in developing countries. In particular, some fraction of aggregate consumption is likely to be accounted for by consumption of liquidity-constrained individuals, for which the first-order condition on which estimation is based does not hold.8 While the existence of liquidity constraints implies a relatively small elasticity of savings9 and can therefore explain Giovannini’s results, it also implies misspecification of the estimated unrestricted first-order condition. Further investigation is called for.

Therefore, Section II provides a framework for the estimation of the degree of intertemporal substitution in consumption, in which liquidity constraints are explicitly allowed for. Section II addresses also the second major issue dealt with in this paper. It extends the representative consumer’s utility function to include government spending and assesses the role of public expenditure in private consumption decisions. The importance of the response of private spending to changes in government spending stems from the observation that, if government spending is a substitute for private spending, then policies seeking to restrain government expenditure are likely to induce higher private consumption. However, most adjustment programs in developing countries attempt to ensure that government deficits do not absorb an unduly high share of private savings. Indeed, the creation of public surpluses (presumably through tight expenditure policies) is often seen as a way of generating loanable funds savings for use by private sector investors that avoids the problems involved in implementing more traditional policies aimed at mobilizing private savings. This approach, however, disregards the fact that direct crowding out can partly or fully counteract government efforts.10 Indeed, for some Latin American countries, McDonald (1983) provides evidence of a sizable degree of substitution between private and public consumption. But his results rest on an inappropriate definition of disposable income and should therefore be investigated further.

As Section III makes clear, an effort is made to construct as accurate and extensive a data set as existing sources allow. In particular, the empirical analysis in Sections IV and V focuses on private savings behavior over the period 1973-83 in 49 developing countries, grouped in six sets of pooled time-series cross-section observations, each one referring to a single geographical region.

Finally, Section VI presents the main conclusions of the analysis and its policy implications.

II. Theory

Research on consumption in the early 1980s (reviewed in Deaton (1986) and King (1985)) has been marked by the important works of Hall (1978), Grossman and Shiller (1981), and Hansen and Singleton (1982) which open the possibility of a direct estimation of the parameters of the intertemporal utility function characterizing the behavior of a representative individual without requiring explicit solutions of the consumer’s dynamic-optimization problem. In addition, Hansen and Singleton (1982) have shown how one can test the overidentifying restrictions implied by the hypothesis of continuous optimization of a stable, additively separable objective function.

Following this line of research, I posit that aggregate consumption can be modeled as the outcome of optimizing decisions of a representative consumption unit (household).11 The household faces an economic environment in which future opportunities are uncertain, and it has a stationary utility function that is additively separable through time and is defined over a composite consumption good as follows:

with

In equations (1) and (2), Vt denotes expected utility at t, Et denotes the expectations operator conditional on information available at t, p denotes a constant-discount factor, Cτ denotes private consumption of goods at τ, and Gτ denotes government expenditure in period τ. The parameter γ in equation (1) controls intertemporal substitution: large and negative values of γ are characteristic of consumers who are willing to smooth consumption over time and who respond only to substantial changes in incentives. Finally, U denotes a function that is increasing and concave in a Cobb-Douglas aggregate of per capita private and public consumption.

The consumer (household) maximizes equation (1), subject to the following period-to-period budget constraint:

where Aτ denotes real assets at the end of period τ, Rτ denotes the real rate of return between periods τ - 1 and τ, and Yτ denotes real non-property income (net of taxes) in period τ. As long as the optimum path lies in the interior of the budget set, we can use simple perturbation arguments to establish certain characteristics of the optimal path. At any point along an optimal path the representative consumption unit cannot make itself better off by forgoing one unit of consumption at time t and using the proceeds to purchase any other good at any other point in time. Formally, at time t the marginal condition will be given by

which, apart from implicitly defining Ft + 1, is satisfied for any freely traded risky asset (even if other assets, such as human capital, cannot be traded freely) and holds for consumers who expect with certainty to be alive in the next period, regardless of the length of horizon of their maximization problem. Notice also that condition (4) does not depend on any assumption about expectations regarding future labor income, government spending, or rates of return.

Estimation of the first-order condition for utility maximization (e.g., equation (4)) represents an alternative approach to estimating standard consumption functions. The difficulties associated with the latter are well known and mostly concern the Lucas critique: the relation between consumption, income, and interest rates depends on the wider macroeconomic context and may not be stable over time, even though preferences remain unchanged. However, the research done so far has provided only limited support for the econometric restrictions implied by the Euler equation approach. Furthermore, the assumptions usually underlying the application of the Euler equation approach are far from being generally accepted. (See Ando and Kennickell (1986), Blinder and Deaton (1985), Deaton (1986), and King (1986).)

Under rational expectations and market clearing, the first-order condition (4) holds ex post except for an error term uncorrelated with information available to the consumption unit at time t. In other words,

where ∈t+1 denotes the forecast error having a zero mean and a constant variance (σ2).

In a setting characterized by intertemporal separability and constant relative risk aversion, given by equations (1) and (2), and with lowercase letters denoting natural logarithms and Δ the difference operator, equations (4) and (5) imply12

where ψc=ψc(ρ,σ2,γ,α);ψr=1[γ(α1)+1];andψg=γαψr.

Since ψr > 0, ψg is greater than, equal to, or less than zero, depending on whether γ is positive, zero, or negative. In equation (6), the error term ut+1 reflects the impact of “news” (or “surprises”) about current levels of income, interest rates, and government spending. It is therefore orthogonal to all past information.13

As it stands, though, equation (6) still disregards the possibility that some consumers may face constraints on the amount they borrow, or that loan rates available to them may be higher than the corresponding lending rates prevailing in the market. These situations may arise for a number of reasons, including imperfections in capital markets and tax policy. For example, the tax system can generate divergencies between after-tax rates on borrowing and lending. Alternatively, large transaction costs, the possibility of bankruptcy, and/or asymmetric information about credit-worthiness in the hands of lenders and borrowers can result in lenders denying loans to potential borrowers with particular characteristics.

Suppose, then, that the liquidity constraint takes the form of a restriction on the total net stock of traded assets, as follows:

where a negative value of At indicates net indebtedness in period t. The additional condition AT ≤ 0 provides the necessary endpoint constraint. Equation (7) is expressed in terms of the net position in order to allow the use of illiquid assets as collateral. According to the equation, potential lenders make the size of the loan conditional upon non-property income. Notice that the lending rule is time dependent, since the intercept φt is allowed to respond to changes in government legislation and macroeconomic conditions in general.

Expressing liquidity constraints in the form of equation (7), that is, exogenous stock constraints, is important because such borrowing restrictions can be exploited by stabilization policy (Hubbard and Judd (1986)). Of course, other alternatives are conceivable. For example, Hayashi (1985) discusses the case of imperfect information in the loan market and shows that it is not necessarily exploitable for stabilization purposes. (See also King (1986).)

Under the additional constraints given by equation (7), it can be shown (as in Muellbauer (1983, 1986 a) and Zeldes (1985)) that equation (6) has to be augmented by adding a term [ψτµt] where µt is an increasing function of the shadow price associated at t with credit rationing or, in other words, is the marginal increase in expected lifetime utility derived from a unit relaxation of the credit constraint in period t. Since agents are constrained from borrowing more, but not from saving more, µ, is zero when the constraint is not binding and positive when it is binding.

In principle, µ, could be derived by solving the whole intertemporal programming problem. However, a sufficiently general solution is hardly likely to be operational. As an alternative, Muellbauer (1986 a, p. 10) therefore suggests that, “if consumers are most likely to want to borrow and hence, other things being equal, to encounter credit restrictions when future income prospects look bright compared with current circumstances,” then, in the aggregate, µt is likely to depend positively on terms like Et(zt + 1 - ct) where Zt = At - 1(Rt - 1) + Yt—that is, real disposable income.14 In other words, consumers whose liquidity is constrained at t may not expect it to be constrained at t + 1 and may therefore be forced to let their consumption path follow their income path more closely. Equation (6) would therefore be rewritten as

which can be interpreted as an approximation to the Euler equation for consumption that incorporates credit constraints. Abstracting from real interest rate and government spending effects, as ψµ tends to one in equation (8), consumption developments increasingly mimic income developments. Notice that, for ψg= ψµ = 0, equation (8) reduces instead to Hall’s (1981) original formulation, which was estimated by Giovannini (1985).15

III. The Data Set

A thorough empirical analysis of private saving behavior in developing countries raises several difficult statistical problems, which stem mostly from the inadequacies in the data and their lack of comparability. A reasonable number of observations on aggregate time-series data is available, on a consistent basis, for only a few developing countries. In the great majority of cases, less than 20 annual observations are available.16 In such a situation, pooling cross-section and time-series data for a number of countries seems to be the most sensible procedure, provided that sufficient allowance is made for obvious institutional and cultural differences among countries.

Following this line of research, the empirical analysis of the present paper is based on six sets of pooled time-series, cross-section data, with each one referring to what, it is to be hoped, is a homogeneous geographical region.17 The first set includes 12 countries in sub-Saharan Africa. To give a different order of magnitude, this sample covers 40 percent of the 1975 gross domestic product (GDP) of the whole region, as defined in the World Bank’s World Tables. The second set includes five countries in North Africa and the Middle East that accounted for 61 percent of the whole region’s 1975 GDP. The third set covers nine countries in East and South Asia and the Pacific, or 46 percent of that region’s 1975 GDP. The fourth and fifth sets cover eight countries in Central America (including the Caribbean) and nine in South America, respectively, with coverage in terms of 1975 regional GDP of 76.2 percent and 83.1 percent, respectively.18 Finally, the sixth set of data includes six Southern European countries, accounting for 77 percent of the 1975 regional GDP. The sample as a whole contains 11 low-income, and 38 middle-income, countries. Low-income countries are, therefore, somewhat underrepresented.19Appendix I provides a detailed description of the data set.

It is important to recall that appropriate measurement is particularly difficult for real interest rates, where the problem of choosing a particular interest rate series from series that may be available is coupled with the question of appropriately deflating nominal interest rates (Khatkhate (1985)). In this respect, the approach described in the previous section turns out to be particularly useful, because the relationship represented by equation (6) should hold for all real rates of return on freely traded assets.

In order to provide an indication of the robustness of the results, two alternative measures of the nominal rate are used. On the one hand, domestic interest rates on time deposits of commercial banks, which constitute a relatively large segment of the financial system in developing countries, are considered.20 On the other, implicitly making reference to the small, open economy model, the nominal interest rate is derived as the relevant foreign interest rate adjusted for expected changes in the exchange rate. The latter alternative implies that the relevant real interest rate depends on the rate of change of the real price of home goods (Dornbusch (1983)). Of course, it may be argued that the small, open economy stereotype is inappropriate for most developing countries that are characterized by pervasive foreign exchange and trade controls. However, it has been suggested (by Tanzi and Blejer (1982)) that, even in countries with severe restrictions on capital movements and other exchange controls, it is unlikely that economic agents will be prevented from illicitly substituting foreign currency and foreign financial assets for domestic currency and domestic financial assets if incentives to do that are sufficiently strong.21

IV. Estimation

For estimation purposes, let us rewrite the theoretical model described in Section II as follows:

where the suffix i identifies the ith country in each of the geographical areas referred to in the previous section. In other words, the constant term in equation (9), being a function of the variance of the forecast error, is allowed to differ among countries because, for example, countries with a higher share of their gross domestic product originating in agriculture are likely to face higher uncertainty. In addition, the coefficient of the proxy for borrowing constraints, ψµ, is allowed to take different values in low-income countries (i.e., those countries identified by the superscript j and belonging to the subset identified by the suffix “ℓ”) and in middle-income countries (i.e., identified by the superscript k and belonging to the subset “m”). Low-income countries are taken to be the countries currently eligible for use of International Development Association (IDA) resources. Of course, according to the interpretation of ψµ used in this paper, one would expect ψµ,ℓ > ψµ,m.22 Finally, the original error term in equation (6)—that is, ut+1—is now linearly decomposed into three innovation terms referring to z, g, and r, respectively, as well as two random components that have zero means but are not necessarily homoskedastic, because the variance of different countries’ forecast errors may differ and this difference could be only partially incorporated into the innovation terms. The first component is country specific and is uncorrelated across countries (vt+1i), while the second one is an area-wide component, which equally affects all countries in a particular geographical area (v̄t+1).23 The obvious example of the latter type of component would be one indicating the effect of the recent drought in sub-Saharan Africa, provided the drought’s effects were not already incorporated in the income “news.”

Notice that the variable rti is alternatively defined as [qtiΔpti] where qt denotes the domestic nominal interest rate and pi denotes the (logarithm of the) consumer price level, or as [qt*+ΔetiΔpti] where, apart from pti, qt* denotes the representative nominal interest rate paid on foreign-currency assets, and et denotes the (logarithm of the) exchange rate, defined as units of domestic currency per unit of foreign currency.

Disregarding, for the time being, the expected (or unexpected) nature of the variables on the right-hand side of equation (9), the appropriate estimator for the kind of setting described by equation (9) is given by what is known, if we can regard each country as a group, as the between-within groups, fixed-effects estimator. As Mundlak (1978) shows, this estimator amounts to applying ordinary least squares to equation (9) expressed in terms of “transformed” variables—that is,

where the transformation takes the following form:

for a generic variable ϰt, and where T and N denote the number of time periods and the number of countries, respectively. In other words, the transformed variable is the original variable minus the country and time means plus the total mean. Notice that the transformation eliminates the constant term and the area-wide error term. In general, the transformation would eliminate all variables not simultaneously indexed on i and t. Therefore, if the nominal interest rate is given by the adjusted foreign interest rate, the term Etr¯t+1i reduces to Et[Δ(et+1ipt+1i)].

Reverting to the modeling of expected and unexpected (or “surprise”) variables, in equation (10) use is made of the well-known two-step procedure involving the estimation of an auxiliary set of equations describing the variables about which expectations are formed and then substituting the estimated residuals and predicted values as appropriate in the relevant structural equation.24 A vector autoregression (VAR) is estimated for the transformed variables z̄, , and r̄; the right-hand-side variables of the VAR include lagged consumption, lagged disposable income, lagged government spending, lagged nominal interest rate (or lagged devaluation), two lags of the price level, and a time trend.25 In general, the VAR equations for z̄ and fit the transformed data quite well, while, as one would expect (Hall (1981)), the real interest rate r̄ appears to be more difficult to predict. (See Appendix II.) Disposable income and government spending are strongly autoregressive; in addition, they help to predict each other, while increased inflation signals a future slowdown in the rate of growth.

Once anticipated and unanticipated series are available, equation (10) can be estimated by ordinary least squares. However, as shown by Pagan (1984), the two-step procedure does not yield correct estimates of all the standard errors. In particular, while the standard errors of the coefficients of the “surprise” variables are correct, standard errors for the remaining coefficients have to be obtained using a two-stage least-squares regression that omits the surprise terms and uses the VAR as the first stage.

V. Empirical Results

Tables 1-6 report the estimates of the coefficients of equation (10) for the six geographical regions described in Section III. Before examining the tables in detail, it is worth emphasizing their main implications. First, the omission of liquidity constraints appears to consistently and seriously bias downward the estimates of the intertemporal elasticity of substitution. Second, where liquidity constraints are substantial (as in regions where the use of IDA resources is common), intertemporal substitution is weak and very large changes in incentives are necessary to induce postponement of consumption. Third, as expected, low-income countries suffer most from liquidity constraints and therefore react strongly to expected income changes, although there is no clear-cut pattern in the way different countries react to unexpected income shocks. In short, the picture that emerges from the evidence is a highly coherent one in which differences in behavioral responses appear to be linked more to the stage of development of different areas or countries than to unexplained shifts in preferences.

Table 1.Sub-Saharan Africa: Parameter Estimates and Test Statistics1
ri = qi − Δpiri = q* − Δ(ei/pi)
(i)(ii)(iii)(iv)
ψr0.06 (0.38)0.33 (0.43)−0.04 (0.16)2
ψg−0.32 (0.21)−0.25 (0.21)
ψµ,m0.22 (0.08)0.23 (0.08)
ψµ,t0.72 (0.19)0.70 (0.19)
ξz0.41 (0.07)0.42 (0.07)
ξg0.02 (0.08)0.02 (0.08)
ξr0.04 (0.10)−0.01 (0.07)
γ−16.27 (112.89)−3.25 (5.73)
γ(1 − α)−2.24 (4.48)
α0.31 (0.22)0.20 (0.13)
α/(1 − α)0.45 (0.46)0.25 (0.21)
αˆ0.070.050.070.05
R20.510.720.510.72
D-W1.741.811.751.82
Chow0.5930.7640.6030.715
n. ob.104104104104
Note: D-W denotes the Durbin-Watson statistic; Chow denotes the Chow test; and n. ob. denotes the number of observations.
Table 2.Middle East and North Africa: Parameter Estimates and Test Statistics
ri = qi − Δpiri = q* − Δ(ei/pi)
(i)(ii)(iii)(iv)
ψr0.99 (0.77)0.98 (0.68)0.23 (0.44)1.17 (0.51)
ψg11
ψµ,m0.22 (0.09)0.41 (0.14)
ψµ,t
ξ0.38 (0.14)0.39 (0.11)
ξg0.09 (0.14)0.08 (0.14)
ξr0.02 (0.42)−0.12 (0.19)
γ−0.01 (0.79)−0.02 (0.71)−3.38 (8.41)0.15 (0.37)
γ(1 − α)
α
α/(1 − α)
αˆ0.070.060.070.05
R20.060.390.010.47
D-W1.771.861.781.68
Chow0.4320.5430.3020.333
n. ob.44444444
Note: D-W denotes the Durbin-Watson statistic; Chow denotes the Chow test; and n. ob. denotes the number of observations.
Table 3.East and South Asia and Pacific: Parameter Estimates and Test Statistics
ri = qi − Δpiri = q* − Δ(ei/pi)
(i)(ii)(iii)(iv)
ψr0.07 (0.17)0.18 (0.18)–0.04 (0.10)0.09 (0.11)
ψg–0.03 (0.10)1
ψµ,m0.17 (0.14)0.23 (0.16)
ψµ,t0.79 (0.39)0.65 (0.33)
ξz0.58 (0.04)0.63 (0.05)
ξg–0.05 (0.04)–0.04 (0.05)
ξt0.12 (0.08)–0.01 (0.02)
γ−12.79 (31.78)−4.74 (5.82)−10.23 (13.74)
γ(1 − α)−5.40 (5.73)
α0.03 (0.12)
α/(1 − α)0.04 (0.13)
αˆ0.030.020.030.02
D-W0.010.650.010.63
D-W2.112.332.142.16
Chow0.6020.7330.6220.674
n. ob.84848484
Note: D-W denotes the Durbin-Watson statistic; Chow denotes the Chow test; and n. ob. denotes the number of observations.
Table 4.Southern Europe: Parameter Estimates and Test Statistics1
pi = qi − Δpiri = q* - Δ(eipi)
(i)(ii)(iii)(iv)
ψr0.08 (0.08)20.05 (0.06)0.17 (0.05)
ψg22
ψµ,m0.39 (0.09)0.49 (0.14)
ψµ,t
ξz0.61 (0.11)0.58 (0.10)
ξg−0.58 (0.10)−0.61 (0.11)
ξt−0.04 (0.05)−0.09 (0.05)
γ−11.66 (13.24)−18.44 (21.74)−4.98 (2.15)
γ(1 − α)
α
α/(1 − α)
αˆ0.040.030.040.03
R20.0300.710.0300.77
D-W1.601.581.611.84
Chow0.4630.5440.3530.535
n. ob.56565656
Note: D-W denotes the Durbin-Watson statistic; Chow denotes the Chow test; and n. ob. denotes the number of observations.
Table 5.Central America and Caribbean: Parameter Estimates and Test Statistics
rt = qi − Δpiri = q* − Δ(ei/pi)
(i)(ii)(iii)(iv)
ψr0.37 (0.16)0.35 (0.17)−0.05 (0.19)1
ψg11
ψµ,m0.22 (0.21)0.34 (0.19)
ψµ,t
ξz0.45 (0.10)0.49 (0.10)
ξg0.02 (0.06)0.03 (0.06)
ξt0.19 (0.07)−0.13 (0.03)
γ−1.86 (1.32)−1.86 (1.39)
γ(1 − α)
α
α/(1 − α)
αˆ0.050.040.050.04
R20.080.440.010.39
D-W1.281.461.131.49
Chow1.2921.1931.5421.124
n. ob.74747474
Note: D-W denotes the Durbin-Watson statistic; Chow denotes the Chow test; and n. ob. denotes the number of observations.
Table 6.South America: Parameter Estimates and Test Statistics
ri = qi - Δpiri = g* − Δ(ei /pi)
(i)(ii)(iii)(iv)
ψt0.09 (0.07)0.01 (0.05)−0.31 (0.15)1
ψg2
ψµm0.65 (0.12)0.71 (0.10)
ψµt
ξz0.48 (0.09)0.48 (0.09)
ξg0.02 (0.05)−0.04 (0.05)
ξr0.01 (0.04)−0.05 (0.04)
γ−9.83 (8.11)−154.52 (1,153.8)
γ(1 − α)
α
α/(1 - α)
αˆ0.050.040.050.03
R20.060.570.060.60
D-W1.661.511.471.64
Chow1.2032.2441.1832.575
n. ob.88888888
Note: D-W denotes the Durbin-Watson statistic; Chow denotes the Chow test; and n. ob. denotes the number of observations.

In Tables 1-6, the columns labeled (ii) and (iv) report the coefficient estimates for the two measures of the real interest rate and the six subsamples, respectively, their heteroskedasticity-consistent (White (1980)) standard errors (derived as above), as well as some diagnostic statistics, such as the Durbin-Watson statistic for fixed-effects models given in Bhargava and others (1982), and a Chow stability test across the 1981-83 period.26 This period coincides with the downward trend of oil prices (in U.S. dollars) and, therefore, also with substantial (and, in recent times, unprecedented) shifts of real income from oil exporting to oil importing countries. In addition, the same period witnesses the emergence of the debt crisis. Stability tests are expected to detect possible structural breaks related to these events. To allow for comparison with previous work, columns (i) and (iii) in Tables 1-6 report the results derived by following Giovannini (1985) and estimating Hall’s (1981) original formulation (corresponding to equation (10) with ψg = ψµℓ = ψµm = ξz = ξg = ξr = 0).

Tables 1-6 also report the estimates of the implied behavioral parameters, as well as some interesting functions of the same parameters, along with their standard errors derived by linearizing the underlying nonlinear functions.27 In particular, the tables show estimates of the parameters γ and α. In the restricted model, the former parameter controls the inter-temporal elasticity of substitution in consumption, which is given, instead, by γ(1 − α) in the full model (equation (10)). The latter parameter defines the weight of government spending in the Cobb-Douglas consumption index (equation (2)) and, if nondistortionary taxes are available and perfect transformation in production is assumed, it also defines the optimal provision of public goods as a percentage of private ones (i.e., α/(1 − α)).

In general, the full model constitutes a substantial improvement over its restricted version. The available diagnostic does not suggest misspecification, and, in particular, the hypothesis of parameter constancy across the 1981-83 period cannot be rejected except in South America.

Contrary to Giovannini’s (1985) findings, there is quite clear-cut evidence of a positive relationship between the rate of growth of per capita consumption and the expected real interest rate. Furthermore, in three regions out of six (Middle East and North Africa, Southern Europe, and Central America), the coefficient ψr also turns out to be positive and significantly different from zero, although this result depends on the definition of the real interest rate. It may be argued that assets denominated in foreign currency are unlikely to be a significant item in private portfolios in sub-Saharan Africa, while indications are that the reverse is true in the Middle East and North Africa and in Southern Europe.28 In general, however, the restricted model estimated by Giovannini (1985) tends to bias downward the estimate of ψr.

Notwithstanding these results, ψr still remains quite small and the intertemporal elasticity of substitution therefore tends to take on negative values that are larger, in absolute terms, than those observed in developed economies.29 In addition, with the exceptions of the Middle East and North Africa, and Central America and the Caribbean, the estimates of γ do not tend to differ widely across regions, and they indicate a reduced response by the consumers to changes in incentives. It is important to stress that if the sample excluded 1982 and 1983, South America, too, would show a positive ψr coefficient that was significantly different from zero.30 As is apparent from the results of the Chow test, however, the relationship weakens considerably in the early 1980s. Therefore, although the extent of the misspecification of the restricted model is apparent and substantial, the main thrust of Giovannini’s work remains largely unaffected.

Government spending never appears to play a substantial role in the regressions. No definite pattern of substitution emerges from the estimates. On the contrary, private consumption is mostly insensitive to the expected path of government spending, with the exception of sub-Saharan Africa where, for what it is worth, the implied estimate of the optimal provision of public goods (as a percentage of private consumption) exceeds the average government spending/private consumption ratio over the 1973-83 period (i.e., 0.27).

Most of the improvement with respect to the restricted formulation shown in columns (i) and (iii) is therefore clearly attributable to the liquidity-constraint proxies and to the impact of “news” on disposable income. The “surprise” variables explain a substantial amount (from 10 percent to nearly 35 percent)31 of the variance of the error in the regressions reported in columns (ii) and (iv) in Tables 1-6, as the rational-expectations approach to the consumption function suggests. Statistically, the innovation in income is the most important such variable.

However, it cannot be safely said that only unexpected changes in income cause consumption to change, as modern versions of the permanent-income hypothesis suggest. The coefficients of the liquidity constraint proxies (ψµ,ℓ and ψµ,m) are always positive, of substantial magnitude, and significantly different from zero.32 In addition, ψµℓ turns out to be always greater and significantly different from ψµ,m, and both coefficients are roughly of the same order of magnitude across regions.33 Interestingly, as one would expect, the relationship between the rate of growth of consumption and the expected real interest rate shows up more clearly and strongly where the proxy for liquidity constraints plays a minor role.

In order to clarify further these issues, Table 7 focuses on the effects of changes in interest rates on consumption. To characterize fully the consumer’s response to random shocks, a closed-form solution to the stochastic control problem described in Section II would be needed. Since such solutions remain intractable, this paper follows Mankiw and others (1985) and concentrates on the effects of interest rate changes in a deterministic environment. The elasticities reported in Table 7 illustrate the changes in consumption at t in response to temporary changes in the real interest rate, from t to t + 1. These are short-run elasticities, in the sense that the effect of such changes after t + 1 is ignored.34

Table 7.Six Regions: Real Interest Rate Elasticities of Consumption1
ri = qi - Δpiri = q* − Δ(ei/pi)
(i)(ii)(iii)(iv)
Sub-Saharan Africa−0.06−0.25
Middle East and North Africa−1.05−1.04−0.24−1.25
East and South Asia and the Pacific−0.08−0.18−0.09
Southern Europe−0.05−0.18
Central America and the Caribbean−0.37−0.37
South America−0.10−0.01

Figures in Table 7 describe the percentage change in consumption following a 1 percent change in the variable (1 + r). Hence, if the real rate of interest jumps from, say, 3 to 4 percent in sub-Saharan Africa, the corresponding reduction in consumption as implied by model (10) is about 1/4 of 1 percent.

Table 7 conveys the same message as Tables 1-6, although in a different form. In particular, the relationship between the degree of responsiveness of consumption to changes in the real interest rate and the magnitude of liquidity constraints (as described by the coefficients ψµ,ℓ and ψµ,m) is, if anything, emphasized.

Assuming that ψµ = 0 yields the Euler equation satisfied under market clearing, or, in other words, assuming that {ψµ[Et(zt+1) − ct]}can provide an estimate of the Lagrange multiplier associated with transferring resources between tomorrow and today,35 the pervasiveness of liquidity constraint can be seen easily by computing the (lower) rate of growth of consumption that would have taken place in the absence of such constraints.36 It turns out that sub-Saharan Africa, North Africa and the Middle East, and South America, which witnessed average growth rates of per capita consumption of about 0.6 percent, 4.8 percent, and 1.7 percent, respectively, in 1973-83, would actually have experienced much lower rates of growth (about −0.4 percent, 3.2 percent, and 0.9 percent, respectively). In contrast, East and South Asia and the Pacific, and Southern Europe, whose per capita consumption levels grew by 3.2 percent and 2.4 percent, respectively, over the same period, would have had annual growth rates approximately 0.3 percent lower. The only region that actually experienced, on average, negative values of the [Et(zt+1) − ct] variable over that period—Central America and the Caribbean—is the only region to show an estimate of ψµ that is not significantly different from zero.

VI. Policy Implications

Analyses based on intercountry data are subject to several well-known caveats, and the results obtained should be viewed with caution. This is even truer when, as in the present case, data problems are known to be substantial. Nonetheless, the results of the present study appear to be robust in most respects. They provide a coherent picture of private saving behavior in developing countries, offer reasons for the existing behavioral differences among geographical regions, and suggest a number of important policy implications.

With respect to the issue of the real interest rate elasticity of savings, the available evidence indicates that in all regions considered, the expected growth of consumption does change with changes in the real interest rate. In addition, in regions such as the Middle East and North Africa, Southern Europe, and Central America and the Caribbean, the response of consumption growth to the expected real interest rate is also significantly different from zero. However, if the magnitude of the estimated parameters is to be taken seriously, the effective mobilization of domestic savings through changes in saving incentives is likely to require changes in the real interest rates, which, given the existing constraints, may prove infeasible, especially in low-income developing countries. In such a case, a viable alternative is the one considered by Blejer and Cheasty (1989)—that is, the generation of budgetary surpluses. As long as these are derived by expenditure restraints, they are not likely to crowd in additional private expenditure and thereby be counteracted by private agents’ behavior.

More far reaching, however, are the implications of the existence of pervasive liquidity constraints for fiscal policy design and implementation. The small size of current resources, compared with lifetime resources, and consumers’ limited ability to borrow against future income clearly affect the way they look at issues such as the efficacy of temporary tax cuts and the effects of government budget deficits on aggregate demand. In Tobin’s (1980, p. 57) words, liquidity-constrained consumers are not “indifferent to the opportunity to defer tax payments. Even if they themselves must pay the taxes later, they will increase their consumption now. In effect, the government lends to them at its borrowing rate of interest, an option not otherwise available in the credit market.” Fiscal policy ineffectiveness arguments are therefore affected if a substantial number of consumers are liquidity constrained,37 although, in assessing the effects of debt-financed tax cuts, the distribution of tax changes across consumers has to be considered. Of course, the reduced responsiveness of saving to changes in the real interest rate further emphasizes the role of traditional stabilization policies.

The same arguments that, under borrowing constraints, can lead to countercyclical policies on efficiency grounds also impinge on a number of issues in tax policy evaluation and tax reform. If liquidity constraints are present, traditional arguments in favor of wage and consumption taxation or proportional taxation and against capital taxation and progressive income taxation lose some of their appeal. For example, Hubbard and Judd (1986, p. 27) show that “a switch from progressive to proportional income taxation would speed up tax collection, raising tax rates on low-income consumers and reducing their consumption substantially when liquidity constraints are important.” In other words, tax exemptions, as well as other forms of social insurance, would not only obey considerations of equity but would also be grounded in efficiency considerations.

Similarly, the usual conclusion suggesting that substantial efficiency costs are likely to characterize capital income taxation, as opposed to labor income taxation, is likely to be reversed to some extent when liquidity constraints are introduced. Again, this is because capital income taxation effectively delays the collection of tax payments over an individual’s life cycle.

It should be stressed that tax policies designed to lessen the burden of borrowing constraints may induce substantial welfare gains if the public does not substitute easily between present and future consumption. If, as appears to be the case in developing countries, people prefer even consumption paths and show low elasticities of intertemporal substitution, the welfare cost of borrowing constraints is likely to be exacerbated. In this respect, the results of this paper underline the role of financial reforms in developing countries.

In recent years, a substantial amount of work has been carried out in developed economies on the effects of liquidity constraints on consumers’ behavior. Given the importance of considering capital market imperfections as pre-existing distortions in normative and positive economic analyses, it is surprising that liquidity constraints have received so little attention in the analysis of saving behavior in developing countries, where they seem to be a simple matter of common-sense observation.

Appendices
I. The Data

The data set for the present study has been constructed by assembling information from all available international sources: United Nations, National Accounts Statistics; International Monetary Fund, International Financial Statistics Yearbook and Government Finance Statistics Yearbook; and International Bank for Reconstruction and Development, World Tables; as well as national sources as needed.

As is well known, because of the unreliability and internal inconsistency of data and the varying methodology in different countries, data in the present sample may be subject to a wide margin of error. In addition, in assembling different sources of information, attention should be paid to conceptual differences and their implications. These remarks apply, in particular, to the construction of the variable Zt (i.e., the per capita private disposable income, in constant prices), to the estimation of the real interest rate, and to the definition of Gt.

Per capita private final consumption expenditure, in constant prices (index: 1980 = 1), is denoted by Ct. Sources: United Nations, National Accounts Statistics: Main Aggregates and Detailed Tables, 1983 (New York, 1986), Tables 1.1 and 1.2, for consumption data; various issues of International Bank for Reconstruction and Development, Economic Analysis and Projections Department, Social Indicators of Development, for population data.

Per capita government final consumption expenditure, in constant prices (index: 1980 = 1), is denoted by Gt. Sources: same as for Ct above. According to the definitions in the United Nations’ National Accounts Statistics, this item comprises compensation of employees and other purchases of goods and services. This study therefore disregards capital expenditure and, what is more important, neglects the long-debated question of the correct definition of current, as opposed to capital, expenditure.

Per capita private disposable income, in constant prices (index: 1980 = 1), is denoted by Zt. Defined as gross national product (GNP), minus consumption of fixed capital (CFC, when available), plus net transfers from abroad (NTA, when available), minus tax revenue (TR), plus subsidies and current transfers (SCT, when available), deflated by private final consumption implicit price index. Table 8 reports the availability of these data for the 49 countries in the sample. Sources: United Nations, National Accounts Statistics: Main Aggregates and Detailed Tables, 1983 (New York, 1986), Table 1.12 for gross national products, consumption of fixed capital, and net current transfers from abroad; United Nations, National Accounts Statistics: Main Aggregates and Detailed Tables, 1983 (New York, 1986), Table 1.4 and/or International Monetary Fund, Government Finance Statistics Yearbook (Washington, 1985), Summary Table and Table C for tax revenue and subsidies and current transfers; and International Bank for Reconstruction and Development, Economic Analysis and Projections Department, Social Indicators of Development, for population data. Notice that, while national disposable income (i.e., gross national product minus consumption of fixed capital plus net current transfers from the rest of the world) is often reported in the United Nations’ National Accounts Statistics, the same is seldom true for the general government current receipts and disbursements, and, in particular, for the current tax revenue and for subsidies and current transfers. Therefore, in such cases, use was made of International Monetary Fund, Government Finance Statistics Yearbook, thereby combining transactions recorded on a payments basis and flows measured and classified by their characteristics at the time of transaction (as in the Government Finance Statistics Yearbook), with transactions recorded on an accrual basis and flows measured and classified by future use or purpose (as in the United Nations’ National Accounts Statistics).

Table 8.Six Regions: Availability and Sources of Input Data
Private Disposable IncomeReal Rate

of Interest
CountryGNPCFCNTATRSCTNRIR
Sub-Saharan Africa
Botswananasnasnasnasnasbr, ifspcd, nas
Burundinasn.a.n.a.gfsn.a.dr, ifspcd, nas
Cameroonnasnasnasnasnasbr, ifspcd, nas
Ethiopiawtn.a.n.a.gfsgfstr, ifspcd, nas
Ghananasnasnasgfsgfsbr, ifspcd, nas
Kenyanasn.a.nasgfsgfsdr, ifspcd, nas
Liberianasn.a.n.a.gfsgfsdr, ifspcd, nas
Malawinasn.a.n.a.nasgfsbr, ifspci, ifs
South Africanasnasnasnasnastr, ifspcd, nas
Swazilandnasn.a.nasgfsgfsdr, ifspcd, nas
Zambianasnasnasgfsgfsdr, ifspci, ifs
Zimbabwenasn.a.nasnasnasdr, ifspcd, nas
North Africa and Middle East
Irannasnasnasnasnasbr, ifspcd, nas
Jordannasnasnasgfsgfsbr, ifspci, ifs
Morocconasn.a.nasgfsgfsbr, ifspci, ifs
Syrianasn.a.n.a.gfsn.a.br, ifspcd, nas
Tunisianasnasnasnasnasbr, ifspcd, nas
East and South Asia and Pacific
Fijinasn.a.nasgfsgfsbr, ifspcd, nas
Indianasnasnasnasnasdr, dkpcd, nas
Indonesianasnasn.a.gfsgfsdr, ifspcd, nas
Korea, Republic ofnasnasnasnasnasdr, ifspcd, nas
Malaysianasn.a.n.a.gfsgfsdr, ifspcd, nas
Pakistannasnasn.a.gfsgfsdr, dkpcd, nas
Philippinesnasnasnasnasnasdr, dkpcd, nas
Sri Lankanasnasnasnasnasdr, ifspcd, nas
Thailandnasnasnasnasnasdr, ifspcd, nas
Southern Europe
Cyprusnasnasnasgfsgfsdr, ifspcd, nas
Greecenasnasnasnasnasdr, ifspcd, nas
Israelnasnasnasnasnasdr, nspcd, nas
Maltanasnasnasnasnasdr, ifspcd, nas
Portugalnasnasnasnasnasdr, ifspcd, nas
Turkeynasnasnasgfsgfsdr, ifspcd, nas
Central America and Caribbean
Costa Ricanasnasnasnasnasdr, itspcd, nas
Dominican Republicnasnasnasgfsgfsdr, dkpcd, nas
El Salvadornasnasnasgfsgfsdf, dkpcd, nas
Guatemalanasn.a.nasgfsgfsdr, dkpcd, nas
Hondurasnasnasnasnasnasdr, ifspcd, nas
Jamaicanasnasnasnasnasdr, ifspci, ifs
Mexiconasnasnasgfsgfsdr, ifspcd, nas
Panamanasnasnasnasnasdr, ifspcd, nas
South America
Bolivianasn.a.nasgfsgfsdr, dkpcd, nas
Brazilnasnasn.a.nasnastr, ifspcd, nas
Chilenasnasnasgfsgfsdr, ifspcd, nas
Colombianasn.a.nasnasnasdr, ifspcd, nas
Ecuadornasnasnasnasnasdr, ifspcd, nas
Paraguaynasnasn.a.nasnasdr, dkpcd, nas
Perunasnasnasnasnasdr, dkpcd, nas
Uruguaynasnasnasgfsgfsdr, dkpcd, nas
Venezuelanasnasnasgfsgfsdr, dkpci, ifs
Notes:CFC: Consumption of fixed capital.GNP: Gross national product.IR: Interest rate deflator.NR: Nominal rate of interest.NTA: Net transfers from abroad.SCT: Subsidies and current transfers.TR: Tax revenue.br: Bank rate and discount rate.dk: Khatkhate (1985).dr: Deposit rate.gfs: International Monetary Fund, Government Finance Statistics Yearbook (Washington,1985).ifs: International Monetary Fund, International Financial Statistics (Washington, 1985), various issues.n.a.: Not available.nas: United Nations, National Accounts Statistics: Main Aggregates and Detailed Tables,1983 (New York, 1986).ns: National sources.pcd: Implicit price index of final private consumption.pci: Consumer price index.tr: Treasury bill rate.wt: International Bank for Reconstruction and Development, World Tables: The Third Edition, Volume I: Economic Data (Baltimore: Johns Hopkins University Press, 1981).

The real interest rate is denoted by rt. It is defined as (1 + ri) = (1 + qi)/(1 + Δpi) or, alternatively, as (1 + ri) = (1 + q*)[(1 + Δe)/(1 + Δpi)] Table 8 reports the definitions of the domestic nominal interest rate and of the inflation rate adopted for each country. Sources (apart from national sources): United Nations, National Accounts Statistics: Main Aggregates and Detailed Tables, 1983 (New York, 1986), Tables 1.1 and 1.2 (for the private final consumption deflator); and International Monetary Fund, International Financial Statistics Yearbook (Washington, 1985) for interest rates, exchange rates, and the consumer price index.

For each country in the six subsamples, Table 9 reports the time period considered, the average and the standard deviations of the ratio to GNP of gross private savings as derived by subtracting private final consumption from the measure of disposable income mentioned above, as well as the average and the standard deviations of the ratio to GNP of gross private savings as derived by adding the current account surplus to gross capital formation and subtracting government gross savings. The comparison of the two average ratios is a useful check on the quality of the approximation embodied in our definition of private disposable income. As is apparent from the table, in most cases the two averages are quite close to each other. However, substantial discrepancies arise in a few cases, such as South Africa, Iran, Jamaica, Greece, and Israel. Tracing the reasons for these discrepancies is, of course, far from easy. However, the discrepancies are likely to be due partly to the fact that our approximation to the concept of disposable income disregards interest payments on the public debt and, therefore, in some cases, substantially underestimates income.38 Unfortunately, there are very few countries for which statistics are available that allow one to isolate the volume of interest payments on domestic public debt paid to the private sector. In addition, in cases such as South Africa, the difference partially derives also from a sizable statistical discrepancy that allows the reconciliation of the national accounting aggregates.

Table 9.Six Regions: Coverage and Main Characteristics of Data
Private Savings/GNP Ratios
EstimatedResidual
Periodmean (S.E.)mean (S.E.)
Sub-Saharan Africa
Botswana1973-810.111 (0.086)0.068 (0.071)
Burundi*1973-810.038 (0.037)… …
Cameroon1973-810.069 (0.043)0.112 (0.041)
Ethiopia*1973-810.094 (0.036)0.100 (0.029)
Ghana*1973-810.081 (0.020)0.129 (0.032)
Kenya*1973-820.206 (0.034)0.163 (0.033)
Liberia*1973-820.316 (0.045)0.291 (0.071)
Malawi*1973-830.171 (0.071)0.122 (0.075)
South Africa1973-830.114 (0.032)0.262 (0.036)
Swaziland1973-820.227 (0.140)0.127 (0.127)
Zambia*1973-820.116 (0.076)0.203 (0.072)
Zimbabwe1973-820.235 (0.023)0.213 (0.020)
North Africa and Middle East
Iran1973-790.462 (0.072)0.233 (0.078)
Jordan1973-830.496 (0.093)0.466 (0.096)
Morocco1973-830.207 (0.030)0.147 (0.031)
Syria1973-810.220 (0.035)0.229 (0.081)
Tunisia1973-830.082 (0.037)0.137 (0.022)
East and South Asia and Pacific
Fiji1973-820.185 (0.048)0.161 (0.053)
India*1973-830.134 (0.016)0.201 (0.016)
Indonesia1973-830.091 (0.024)0.106 (0.026)
Korea, Republic of1973-830.133 (0.027)0.190 (0.024)
Malaysia1973-810.288 (0.023)0.238 (0.024)
Pakistan*1973-820.085 (0.013)0.110 (0.014)
Philippines1973-820.122 (0.012)0.207 (0.015)
Sri Lanka*1973-820.112 (0.029)0.127 (0.031)
Thailand1973-830.140 (0.023)0.194 (0.016)
Southern Europe
Cyprus1973-830.180 (0.024)0.250 (0.023)
Greece1973-830.095 (0.036)0.249 (0.030)
Israel1973-830.065 (0.049)0.361 (0.051)
Malta1973-830.250 (0.051)0.171 (0.041)
Portugal1973-810.189 (0.051)0.234 (0.063)
Turkey1973-810.119 (0.023)0.136 (0.036)
Central America and Caribbean
Costa Rica1973-830.055 (0.028)0.119 (0.031)
Dominican Republic1973-810.076 (0.036)0.141 (0.020)
El Salvador1973-820.140 (0.024)0.164 (0.026)
Guatemala1973-830.139 (0.027)0.123 (0.028)
Honduras1973-830.059 (0.048)0.096 (0.033)
Jamaica1973-820.015 (0.046)0.146 (0.036)
Mexico1973-830.188 (0.024)0.218 (0.041)
Panama1973-800.201 (0.017)0.219 (0.041)
South America
Bolivia*1973-830.172 (0.070)0.115 (0.071)
Brazil1973-820.144 (0.035)0.184 (0.021)
Chile1973-830.035 (0.049)0.079 (0.035)
Colombia1973-830.178 (0.016)0.157 (0.011)
Ecuador1973-830.110 (0.038)0.154 (0.030)
Paraguay1973-830.108 (0.032)0.185 (0.016)
Peru1973-830.068 (0.023)0.042 (0.064)
Uruguay1973-830.104 (0.035)0.118 (0.027)
Venezuela1973-820.215 (0.051)0.204 (0.041)
Notes: An asterisk (*) indicates that a country is designated as eligible to borrow from the International Development Association. S.E. denotes standard error.

As Table 9 shows, the sample is characterized by a substantial variability across time and across countries, with the latter variability seen both between and within regional subsets.

II. VAR Estimation

This appendix reports in detail the VAR equations estimated for the six geographical regions and the two alternative definitions of the real rate of return. In all tables (Tables 10-21), the symbols are the same ones used in the text of this paper.

Table 10.Sub-Saharan Africa: VAR Estimation
Dependent Variable
ΔztΔgtrt
zt - 1−0.52 (0.15)0.11 (0.09)0.07 (0.07)
gt-1−0.12 (0.09)−0.34 (0.07)−0.06 (0.06)
ct - 10.18 (0.10)−0.08 (0.08)−0.03 (0.05)
qt - 1−0.30 (0.55)0.07 (0.66)0.87 (0.38)
Δpt - 1−0.17 (0.14)−0.18 (0.15)0.11 (0.15)
pt - 10.06 (0.05)0.02 (0.05)−0.09 (0.07)
d−0.06 (0.06)−0.04 (0.03)0.01 (0.02)
t−0.02 (0.02)0.004 (0.02)−0.02 (0.01)
D-W1.761.911.86
R0.220.230.08
σ̂0.090.080.07
n. ob.104104104
σ (x)0.0960.0840.066
σ(x̂)0.0450.0400.019
σ(x - x̂)0.0850.0730.063
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. The variable d takes values of 1 in 1974 and of −1 in 1975 for Swaziland. Figures in parentheses are White’s (1980) standard errors.
Table 11.Sub-Saharan Africa: VAR Estimation
Dependent Variables
ΔztΔgtΔ(ep)t
zt-1−0.52 (0.15)0.11 (0.09)0.06 (0.12)
gt-1−0.11 (0.09)−0.34 (0.06)−0.07 (0.09)
ct-10.20 (0.10)−0.09 (0.07)0.02 (0.09)
Δet-10.09 (0.10)0.04 (0.08)0.39 (0.19)
Δpt-1t-1−0.14 (0.14)−0.18 (0.15)0.61 (0.28)
pt-10.03 (0.05)0.03 (0.05)−0.16 (0.11)
d−0.06 (0.06)−0.05 (0.03)−0.06 (0.06)
t−0.02 (0.02)−0.03 (0.02)−0.04 (0.02)
D-W1.731.911.89
R0.230.230.18
αˆ0.090.080.10
n. ob.104104104
σ(x)0.0960.0840.106
σ(x̂)0.0460.0400.045
σ(x − x̂)0.0840.0730.096
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. The variable d takes values of 1 in 1974 and of −1 in 1975 for Swaziland. Figures in parentheses are White’s (1980) standard errors.
Table 12.Middle East and North Africa: VAR Estimation
Dependent Variable
ΔztΔgtrt
zt - 10.02 (0.12)0.20 (0.12)0.04 (0.03)
gt - 1−0.27 (0.17)−0.61 (0.13)−0.09 (0.03)
ct - 1−0.58 (0.30)−0.59 (0.33)−0.18 (0.07)
qt - 13.40 (3.73)1.08 (3.22)3.45 (0.64)
Δpt - 1−0.20 (0.41)−0.42 (0.49)−0.31 (0.10)
pt - 10.06 (0.21)−0.54 (0.26)0.23 (0.05)
t0.02 (0.04)−0.08 (0.04)−0.01 (0.01)
D-W2.081.652.35
R0.260.420.44
αˆ0.090.090.02
n. ob.444444
σ(x)0.0920.1060.028
σ(x̂)0.0470.0680.018
σ(x - x̂)0.0800.0810.021
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 13.Middle East and North Africa: VAR Estimation
Dependent Variable
ΔztΔgtΔ(etpt)
zt - 1−0.01 (0.10)0.15 (0.12)−0.14 (0.06)
gt - 1−0.26 (0.17)−0.59 (0.13)−0.08 (0.07)
ct - 1−0.52 (0.27)−0.53 (0.34)0.04 (0.13)
Δet - 10.07 (0.39)−0.37 (0.43)−0.31 (0.17)
Δpt - 1−0.14 (0.45)−0.22 (0.62)0.004 (0.21)
pt - 1−0.15(0.19)−0.58 (0.21)0.01 (0.11)
t0.03 (0.04)0.09 (0.04)0.03 (0.02)
D-W1.971.622.17
R0.240.420.37
αˆ0.090.090.04
n. ob.444444
σ(x)0.0920.1060.047
σ(x̂)0.0450.0690.028
σ(xx̂)0.0800.0800.037
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 14.East and South Asia and Pacific: VAR Estimation
Dependent Variable
Δzt - 1Δgtrt
zt - 1−0.56 (0.14)0.28 (0.18)−0.15 (0.11)
gt - 10.04 (0.06)−0.45 (0.08)−0.05 (0.05)
ct - 10.35 (0.15)0.12 (0.18)0.04 (0.11)
qt - 10.02 (0.27)−0.69 (0.35)0.67 (0.19)
Δpt - 10.13 (0.09)−0.02 (0.14)−0.08 (0.10)
pt - 1−0.004 (0.04)−0.17 (0.05)0.14 (0.04)
t0.01 (0.02)0.01 (0.02)−0.001 (0.01)
D-W1.781.621.95
R0.190.330.31
αˆ0.040.050.03
n. ob.848484
σ(x)0.0450.0590.037
σ(x̂)0.0200.0340.021
σ(x - x̂)0.0400.0480.031
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 15.East and South Asia and Pacific: VAR Estimation
Dependent Variable
ΔztΔgtΔ(et - pt)
zt - 1−0.55 (0.14)0.27 (0.17)−0.25 (0.29)
gt - 10.02 (0.06)−0.38 (0.06)0.02 (0.15)
ct - 10.36 (0.14)0.12 (0.17)0.04 (0.23)
Δet - 10.10 (0.06)0.05 (0.04)−0.19 (0.09)
Δpt - 10.13 (0.08)−0.09 (0.16)−0.11 (0.17)
pt - 1−0.01 (0.04)−0.14 (0.05)0.25 (0.07)
t0.01 (0.02)0.02 (0.01)0.001 (0.003)
D-W1.821.741.92
R0.220.300.13
αˆ0.040.050.08
n. ob.848484
σ(x)0.0450.0590.085
σ(x̂)0.0210.0320.030
σ(xx̂)0.0400.0490.079
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 16.Southern Europe: VAR Estimation
Dependent Variable
ΔztΔgtrt
zt - 1−0.24 (0.13)0.31 (0.12)0.09 (0.16)
gt - 1−0.24 (0.12)−0.38 (0.11)0.32 (0.20)
ct - 1−0.16 (0.16)−0.44 (0.16)−0.10 (0.23)
qt - 10.50 (0.29)−0.001 (0.25)0.27 (0.68)
Δpt - 1−0.15 (0.14)0.07 (0.11)−0.57 (0.25)
pt - 1−0.05 (0.04)−0.05 (0.03)0.10 (0.07)
d1−0.10 (0.03)0.07 (0.03)0.09 (0.03)
d20.04 (0.02)0.02 (0.02)−0.03 (0.03)
t0.02 (0.02)0.02 (0.02)0.02 (0.03)
D-W1.722.061.49
R0.450.290.32
αˆ0.050.050.08
n. ob.565656
σ(x)0.0630.0530.093
σ(x̂)0.0430.0290.052
α(xx̂)0.0470.0450.077
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. The variables d1 and d2 take values of 1 in 1974 for both Cyprus and Portugal. Figures in parentheses are White’s (1980) standard errors.
Table 17.Southern Europe: VAR Estimation
Dependent Variable
ΔztΔgtΔ(etpt)
zt - 1−0.20 (0.12)0.27 (0.10)−0.65 (0.25)
gt - 1−0.24 (0.12)−0.30 (0.09)−0.29 (0.19)
ct - 1−0.23 (0.16)−0.32 (0.16)0.25 (0.28)
Δet-1−0.03 (0.07)0.20 (0.05)−0.54 (0.16)
Δpt - 1−0.19 (0.15)−0.09 (0.11)0.30 (0.24)
pt - 1−0.01 (0.03)−0.05 (0.02)0.04 (0.05)
d1−0.09 (0.03)0.06 (0.03)0.03 (0.04)
d20.02 (0.02)0.03 (0.02)0.01 (0.03)
t0.02 (0.02)0.01 (0.02)0.06 (0.03)
D-W1.802.192.05
R0.430.440.46
αˆ0.050.040.09
n. ob.565656
σ(x)0.0630.0530.114
σ(x̂)0.0420.0350.077
σ(xx̂)0.0480.0390.083
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. The variables d1 and d2 take values of 1 in 1974 for both Cyprus and Portugal. Figures in parentheses are White’s (1980) standard errors.
Table 18.Central America and Caribbean: VAR Estimation
Dependent Variable
ΔztΔgtrt
zt - 1−0.85 (0.14)0.08 (0.22)−0.49 (0.23)
gt - 10.17 (0.07)−0.38 (0.19)0.14 (0.08)
ct - 10.84 (0.14)0.48 (0.18)0.55 (0.20)
qt - 1−0.18 (0.21)−0.30 (0.20)−0.12 (0.28)
Δpt - 1−0.22 (0.10)−0.14 (0.12)−0.48 (0.14)
pt - 10.16 (0.05)0.05 (0.07)0.17 (0.06)
t−0.03 (0.01)0.001 (0.02)−0.01 (0.01)
D-W1.331.151.70
R0.400.220.25
αˆ0.050.080.06
n. ob.747474
σ(x)0.0640.0820.072
σ(x̂)0.0410.0380.036
σ(xx̂)0.0500.0720.062
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 19.Central America and Caribbean: VAR Estimation
Dependent Variable
ΔztΔgtΔ(et - pt)
zt - 10.81 (0.13)0.10 (0.21)0.48 (0.32)
gt - 10.17 (0.07)−0.38 (0.19)0.07 (0.10)
ct - 10.78 (0.15)0.40 (0.29)−0.40 (0.29)
Δet - 1−0.12 (0.05)−0.06 (0.05)0.9 (0.14)
Δpt - 1−0.03 (0.12)−0.03 (0.17)−0.11 (0.22)
pt - 10.13 (0.05)0.002 (0.07)0.07 (0.11)
t−0.03 (0.01)0.001 (0.02)0.01 (0.02)
D-W0.1241.141.99
R0.430.210.08
αˆ0.050.080.10
n. ob.747474
σ(x)0.0640.0820.104
σ(x̂)0.0420.0380.030
σ(x - x̂)0.0490.0720.099
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 20.South America: VAR Estimation
Dependent Variable
ΔztΔgtrt
zt - 1−0.90 (0.20)−0.13 (0.24)2.13 (0.77)
gt - 10.23 (0.07)−0.37 (0.13)0.05 (0.21)
ct - 10.65 (0.20)0.28 (0.24)−1.53 (0.71)
qt - 10.01 (0.03)−0.02 (0.03)0.53 (0.12)
Δpt - 1−0.10 (0.05)0.02 (0.05)−0.20 (0.21)
pt - 10.03 (0.01)0.01 (0.02)0.05 (0.04)
t−0.04 (0.03)−0.02 (0.04)−0.05 (0.07)
D-W1.841.421.66
R0.420.200.45
αˆ0.060.080.16
n. ob.888888
σ(x)0.0700.0850.204
σ(x̂)0.0450.0380.137
σ(xx̂)0.0530.0760.151
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
Table 21.South America: VAR Estimation
Dependent Variable
ΔztΔgtΔ(et - pt)
zt - 1−0.87 (0.20)0.05 (0.21)0.42 (0.33)
gt - 10.21 (0.08)−0.41 (0.12)0.06 (0.14)
ct - 10.63 (0.20)0.25 (0.22)−0.36 (0.33)
Δet - 1−0.14 (0.08)−0.23 (0.09)0.22 (0.14)
Δpt - 10.05 (0.08)0.22 (0.10)−0.32 (0.14)
pt - 10.02 (0.02)−0.01 (0.02)−0.02 (0.02)
t−0.04 (0.03)−0.01 (0.04)0.03 (0.04)
D-W1.821.541.90
R0.450.250.18
αˆ0.050.080.10
n. ob.888888
σ(x)0.0700.0850.103
σ(x̂)0.0470.0420.043
σ(xx̂)0.0520.0740.093
Notes: D-W denotes the Durbin-Watson statistic, n. ob. the number of observations. Figures in parentheses are White’s (1980) standard errors.
References

    Ando, A., and A.Kennickell, “How Much (or Little) Life Cycle Is There in Micro Data? The Cases of U.S. and Japan,” in Macroeconomics and Finance: Essays in Honor of Franco Modiglianied. by R.Dornbusch, S.Fischer, and J.Bossons (Cambridge, Massachusetts: MIT Press, 1986).

    Baltagi, BadiH., and James M.Griffin, “Short and Long Run Effects in Pooled Models,”International Economic Review (Osaka), Vol. 25 (October1984), pp. 631-45.

    Bean, CharlesR., “The Estimation of ‘Surprise’ Models and the ‘Surprise’ Consumption Function,”Review of Economic Studies (Edinburgh), Vol. 53 (August1986), pp. 497-516.

    Bhargava, A., and others, “Serial Correlation and the Fixed Effects Model,”Review of Economic Studies (Edinburgh), Vol. 49 (October1982), pp. 533-49.

    Blejer, MarioI., and AdrienneCheasty, “Fiscal Policy and the Mobilization of Savings for Growth,”Chap. 2 in this volume.

    Blinder, AlanS., and Angus S.Deaton, “The Time Series Consumption Function Revisited,”Brookings Papers on Economic Activity: 2 (1985), The Brookings Institution (Washington), pp. 465-511.

    Davidson, JamesE.H., and others, “Econometric Modelling of the Aggregate Time Series Relationship between Consumers’ Expenditure and Income in the United Kingdom,”Economic Journal (Cambridge, England), Vol. 88 (December1978), pp. 661-92.

    Deaton, AngusS., Life-Cycle Models of Consumption: Is the Evidence Consistent with the Theory?, Working Paper No. 1910 (Cambridge, Massachusetts: National Bureau of Economic Research, 1986).

    Dornbusch, Rudiger, “Real Interest Rates, Home Goods, and Optimal External Borrowing,”Journal of Political Economy (Chicago), Vol. 91 (February1983), pp. 141-53.

    Fry, MaxwellJ., “Money and Capital or Financial Deepening in Economic Development?”Journal of Money, Credit and Banking (Columbus, Ohio), Vol. 4 (November1978), pp. 464-75.

    Fry, MaxwellJ., “Saving, Investment, Growth and the Cost of Financial Repression,”World Development (Oxford, England), Vol. 8 (April1980), pp. 317-27.

    Giovannini, Alberto, “The Interest Elasticity of Savings in Developing Countries: The Existing Evidence,”World Development (Oxford, England), Vol. 11 (July1983), pp. 601-607.

    Giovannini, Alberto, “Saving and the Real Interest Rate in LDCs,”Journal of Development Economics (Amsterdam), Vol. 18 (August1985), pp. 197-217.

    Grossman, SanfordJ., and Robert J.Shiller, “The Determinants of the Variability of Stock Market Prices,”American Economic Review: Papers and Proceedings of the Ninety-Third Annual Meeting of the American Economic Association (Nashville, Tennessee), Vol. 71 (May1981), pp. 222-27.

    Gupta, K. L., “Personal Saving in Developing Nations: Further Evidence,”Economic Record (Burwood, Victoria, Australia), Vol. 46 (June1970), pp. 243-49.

    Hall, RobertE., “Stochastic Implications of the Life Cycle-Permanent Income Hypothesis,”Journal of Political Economy (Chicago), Vol. 86 (December1978), pp. 971-87.

    Hall, RobertE., Intertemporal Substitution in Consumption, Working Paper No. 720 (Cambridge, Massachussets: National Bureau of Economic Research, 1981).

    Hansen, LarsP., and Kenneth J.Singleton, “Generalized Instrumental Variables Estimation of Nonlinear Rational Expectations Models,”Econometrica (Bristol, England), Vol. 50 (September1982), pp. 1269-86.

    Hayashi, Fumio, Tests for Liquidity Constraints: A Critical Survey, Working Paper No. 1720 (Cambridge, Massachusetts: National Bureau of Economic Research, 1985).

    Hubbard, R.Glenn, and Kenneth L.Judd, “Liquidity Constraints, Fiscal Policy, and Consumption,”Brookings Papers on Economic Activity: 1 (1986), The Brookings Institution (Washington), pp. 1-59.

    Jackman, Richard, and JohnSutton, “Imperfect Capital Markets and the Monetarist Black Box: Liquidity Constraints, Inflation and the Asymmetric Effects of Interest Rate Policy,”Economic Journal (Cambridge, England), Vol. 92 (March1982), pp. 108-28.

    Khatkhate, DeenaR., “Assessing the Level and Impact of Interest Rates in Less Developed Countries” (unpublished, International Monetary Fund, 1985).

    King, MervynA., “The Economics of Saving: A Survey of Recent Contributions,” in Frontiers of Economics, ed. by Kenneth J.Arrow and SeppoHonkapohja (Oxford, England and New York: Basil Blackwell, 1985), pp. 227-94.

    King, MervynA., “Capital Market Imperfections’ and the Consumption Function,”Scandinavian Journal of Economics (Oxford, England), Vol. 88 (No. 1, 1986), pp. 59-80.

    Krinsky, Itzhak, and A. LeslieRobb, “On Approximating the Statistical Properties of Elasticities,”Review of Economics and Statistics (Cambridge, Massachusetts), Vol. 68 (November1986), pp. 715-19.

    McDonald, DonoghC., “The Determinants of Saving Behavior in Latin America” (unpublished, International Monetary Fund, 1983).

    Mankiw, N.Gregory, and others, “Intertemporal Substitution in Macroeconomics,”Quarterly Journal of Economics (Cambridge, Massachusetts), Vol. 100 (February1985), pp. 225-51.

    Mikesell, RaymondF., and James E.Zinser, “The Nature of the Savings Function in Developing Countries: A Survey of the Theoretical and Empirical Literature,”Journal of Economic Literature (Nashville, Tennessee), Vol. 11 (March1973), pp. 1-26.

    Modigliani, Franco, “Life Cycle, Individual Thrift, and the Wealth of Nations,”American Economic Review (Nashville, Tennessee), Vol. 76 (June1986), pp. 297-313.

    Muellbauer, John, “Surprises in the Consumption Function,”Economic Journal (Cambridge, England), Vol. 93, Supplement (March1983), pp. 34-50.

    Muellbauer, John, (1986 a), “Notes on Uncertainty, Liquidity Constraints and Aggregation in the Consumption Function” (unpublished: Oxford, England: Nuffield College, 1986).

    Muellbauer, John,(1986 b), Habits, Rationality and Myopia in the Life-Cycle Consumption Function, Discussion Paper No. 112 (London: Center for Economic Policy Research, June1986).

    Mundlak, Yair, “On the Pooling of Time Series and Cross Section Data,”Econometrica (Evanston, Illinois), Vol. 46 (January1978), pp. 69-85.

    Pagan, Adrian, “Econometric Issues in the Analysis of Regressions with Generated Regressors,”International Economic Review (Osaka), Vol. 25 (February1984), pp. 221-47.

    PereiraLeite, Sergio, and DawitMakonnen, “Saving and Interest Rates in the BCEAO Countries: An Empirical Analysis” (unpublished, International Monetary Fund, 1984).

    Snyder, DonaldW., “Econometric Studies of Household Saving Behaviour in Developing Countries: A Survey,”Journal of Development Studies (London), Vol. 2 (January1974), pp. 139-53.

    Summers, LawrenceH., “The After-Tax Rate of Return Affects Private Savings,”American Economic Review: Papers and Proceedings of the Ninety-Sixth Annual Meeting of the American Economic Association (Nashville, Tennessee), Vol. 74 (May1984), pp.249–53.

    Tanzi, Vito, and Mario I.Blejer, “Inflation, Interest Rate Policy, and Currency Substitution in Developing Economies: A Discussion of Some Major Issues,”World Development (Oxford, England), Vol. 10 (September1982), pp. 781-89.

    Tobin, James, Asset Accumulation and Economic Activity: Reflections on Contemporary Macroeconomic Theory (Oxford, England: Basil Blackwell, 1980).

    White, Halbert, “A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity,”Econometrica (Bristol, England), Vol. 48 (May1980), pp. 817-38.

    Williamson, J. G., “Personal Saving in Developing Nations: An Inter-temporal Cross-Section from Asia,”Economic Record (Burwood, Victoria, Australia), Vol. 44 (June1968), pp. 194-210.

    Zeldes, S., “Consumption and Liquidity Constraints: An Empirical Investigation,”Working Paper No. 24–85 (unpublished, Philadelphia: University of Pennsylvania, Rodney L. White Center for Financial Research, 1985).

This paper was first published in the March 1988 issue of International Monetary Fund Staff Papers. I wish to thank Mario I. Blejer, Thanos Catsambas, Riccardo Faini, Alain Ize, T. Jappelli, Jeroen Kremers, and Vito Tanzi for discussions and comments that contributed substantially to the development of this paper. I am also very grateful to Bruce Fuller for kindly providing some of the data.

Leaving aside studies on the relationship between changes in interest rates in the organized money markets and the volume of saving done through financial intermediaries, both surveys report the works of Williamson (1968) and Gupta (1970). The first author examines the role of real interest rates in determining personal saving in selected Asian countries (Burma (now Myanmar), India, Japan, the Republic of Korea, the Philippines, and Taiwan Province of China) over the period 1950-64 and concludes that real rates of interest are, if anything, negatively correlated with national savings. Williamson’s results, particularly as they apply to India, are disputed by Gupta, who questions his savings data as well as his choice of the real interest rate. After examining data for a longer period than Williamson used, Gupta finds, instead, that interest rates play a significant role in determining household saving behavior in India.

Burma (now Myanmar), India, the Republic of Korea, Malaysia, the Philippines, Singapore, and Taiwan Province of China. The main data sources are various issues of International Bank for Reconstruction and Development, World Tables, and International Monetary Fund, International Financial Statistics.

Argentina, Chile, Colombia, Costa Rica, Guatemala, Haiti, Honduras, Mexico, Panama, Paraguay, Peru, and Uruguay. The main data source is International Monetary Fund, International Financial Statistics, various issues.

Benin, Burkina Faso, Côte d’Ivoire, Niger, Senegal, and Togo. The main data source is International Monetary Fund, International Financial Statistics, various issues.

Argentina, Brazil, Colombia, Jamaica, and Mexico in Latin America; Burma (now Myanmar), India, Indonesia, the Republic of Korea, Malaysia, the Philippines, Singapore, Taiwan Province of China, and Thailand in Asia and the Pacific; Greece, Portugal, and Turkey in Southern Europe; and Kenya in Africa. The main data source is International Monetary Fund, International Financial Statistics, various issues.

Burma (now Myanmar), Greece, India, Jamaica, and Turkey.

In terms of 1975 regional gross domestic product (GDP), Giovannini (1985) covers just 2 percent of sub-Saharan Africa, 43 percent of East and South Asia and the Pacific, 72 percent of Latin America, and 64 percent of Southern Europe. In contrast, Fry (1978, 1980) and McDonald (1983) cover only approximately 45 percent of both East and South Asia and the Pacific, and Latin America. Finally, Pereira Leite and Makonnen (1984) limit themselves to nearly 6 percent of sub-Saharan Africa.

Giovannini (1985, p. 215) mentions that “some preliminary experiments where the presence of liquidity constraints was allowed in the model, were not very satisfactory, but yielded the same estimates of the intertemporal substitution elasticity.”

Indeed, over some range, liquidity-constrained individuals can be totally unresponsive to changes in real interest rates. See, however, Jackman and Sutton (1982).

Notice that this discussion disregards the question of whether consumption is sensitive to choice of tax versus debt financing of current government expenditure and concentrates instead on the extent to which government spending directly substitutes for private consumer expenditure. Both are cases of direct crowding out, but their “dimensions” are different: the Ricardian-equivalence proposition is concerned with what is regarded as income and wealth by the private sector, as opposed to what is regarded as consumption by the private sector in the latter case.

Incidentally, notice that casting the analysis in terms of a household makes the “immortality” assumption, which is required for an aggregate version of the Euler condition to hold, slightly more palatable. See, however, Deaton’s (1986, p. 13) comments.

It is assumed that rt + 1, gt + 1, and ct + 1 follow a joint lognormal distribution. See Hansen and Singleton (1982).

Notice that formulation (6) does not allow transitory elements of consumption owing to imperfect execution of plans, which would introduce a first-order moving-average component into the error term. This assumption is not as strong as it seems at the aggregate level, since transitory elements should be uncorrelated between individuals and should therefore average out.

Non-property income would certainly be a more appropriate variable, since credit-rationed consumers are likely to show only a minimal level of assets. Disposable income is, however, preferred in the light of the information available. See, however, Appendix I.

Interestingly, equation (8) bears a close resemblance to the empirically successful consumption function attributable to Davidson and others (1978).

For example, McDonald’s (1983) and Giovannini’s (1985) regressions rarely present more than 15 degrees of freedom.

Stratifying country observations on a geographical basis is just one of the many possibilities, although it seems to be the most obvious one if preference parameters are, to some extent, influenced by institutional and cultural differences. Alternative criteria include, among others, size, economic performance, and per capita income. The last one is indirectly taken into account in what follows.

In addition to the countries listed in Table 8 of Appendix I, the South American sample originally also included Argentina. As it turned out, however, the Argentine subsample, ranging from 1973 to 1980, was dominated by two large outliers in 1974 and 1976. Given the small size of the subsample, it was therefore decided to omit the country altogether.

Reference is made here to the developing countries eligible to use the International Development Association’s (IDA’s) resources. On the basis of that classification, low-income countries are approximately two fifths of the 142 developing countries.

Nevertheless, in a few cases, it proved necessary to use discount rates. See Appendix I for details.

It should be recognized that in most developing countries, the capital market is small and usually confined to one central city, and wealth is held in the form of consumer durables, such as jewelry and livestock. In such cases, rates of return on financial instruments are likely to be largely irrelevant.

It may be argued that since the intensity of borrowing constraints varies across countries, so can the parameters of the underlying representation of preferences vary. On the basis of the available evidence (Zeldes (1985)), this does not appear to be the case.

Countries are not randomly selected, and therefore the area-wide shock cannot be analyzed in an “error-component” kind of model. Besides, the available evidence suggests that the fixed-effects estimator is robust with respect to various forms of dynamic misspecification. See Baltagi and Griffin (1984).

Besides being simpler computationally, the two-step procedure (like other limited-information methods) reduces the contamination of the estimated coefficients in the structural equation by specifying errors in the auxiliary equations.

In order to identify the system given by equation (10) and the vector autoregression, the strong assumption of zero covariance between νt + 1 and the “surprises” in disposable income, government spending, and the real interest rate is necessary. Furthermore, it should be noticed that the system given by equation (10) and the vector autoregression is observation ally equivalent to the system given by equation (10) without “surprises,” but with simultaneity affecting z, g, and r. Therefore, the interpretation used in the present paper relies on the author’s choice.

Estimation and hypothesis testing were carried out by means of the PC version of the Time Series Processor (TSP) (Version 4.01).

These standard errors should be interpreted with some care, considering the poor approximation usually provided by the linearization. See Krinsky and Robb (1986).

The apparent negative and strong relationship between consumption growth and the expected real interest rate (defined in terms of the world interest rate) in South America should not be taken too seriously in the light of the quite poor performance of the underlying VAR equations in that case.

Using U.S. data, Hansen and Singleton (1982) find values of the intertemporal elasticity of substitution of between plus and minus unity. Summers (1984) presents various estimates of γ ranging from -18.0 to 0.4, while Bean (1986) estimates it to be about -1.5.

Equal to 0.09 (0.04) and implying that γ is equal to -10.4 (6.1).

Between 10 percent and 15 percent in sub-Saharan Africa, and North Africa and the Middle East; nearly 20 percent in Latin America; and about 30 percent in both Southern Europe and East and South Asia and the Pacific.

Comparable estimates of ψµ for developed economies are available for only the U.S. economy: Muellbauer (1986 b) estimates it to be about 0.1.

The exception is South America, where the coefficient of the liquidity-constraint proxy for the low-income country (Bolivia) turns out to be negative and to possess a large standard error. In addition, ψµ, m tends to take on higher values than it does in other regions.

Changes in the subsequent period are, however, mediated through changes in future wealth. Hence, for consumers with long horizons, this is likely to be a valid approximation.

Of course, one expects {ψµ[Et(zt+1) - ct]} ≥ 0, at least on the average across time periods and countries. In a world with borrowing constraints, as opposed to one without, consumption can be expected to grow faster, but never slower.

All the computations that follow are based on the estimates reported in column (ii) of Tables 1-6. Calculations made on the basis of the estimates of column (iv) do not change the picture.

Actually, the simulation study by Hubbard and Judd (1986, pp. 33-43) clearly points out that, in the discussion on the Ricardian-equivalence proposition, borrowing constraints are likely to be substantially more important than the absence of intergenerational wealth redistribution (i.e., finite horizons).

This implies, however, that the variable Zt is nearer to the concept of net non-property income required by the theory of Section II.

    Other Resources Citing This Publication