اثرات بالقوه از کاهش ارزش ارز نپال: یک رویکرد تعادل عمومی

This paper measures the potential impacts of the devaluation of domestic currency of the small, developing, landlocked and transition South Asian economy of Nepal, which is lagging behind in policy studies. The impacts on growth, distribution, price changes in factor and product markets, and on selected macroeconomic features are measured. Using a computable general equilibrium model applied to social accounting matrix data, we conclude that devaluation is expansionary but mostly benefits the rich, thus leading to a more uneven income distribution. In general, the expansion of economic activities occurs in agricultural and industrial sectors, whereas services activities contract. However, when the rate of devaluation is high, the agricultural sector also starts contracting. To this typical developing economy, devaluation causes an improvement in saving investment and export/import ratios, whereas the budget deficit widens.

Depreciation of the exchange rate of domestic currency is the most frequent outcome of exchange rate liberalisation in many developing countries. Moreover, devaluation of domestic currency is also one of the major components of the orthodox stabilisation strategy. Whatever the reason for the depreciation of the currency of many developing countries in the long run, the impacts are not only pervasive but also deep. In addition to some theoretical studies, some partial analyses in measuring the impacts of the devaluation of the currencies of developing economies are also apparent, while studies that followed the general system approach are quite limited. The latter have not yet covered the impacts in the poor, landlocked transition and developing economies of South Asia. This study tries to fill this gap. In this paper, we start with a review of some of the studies – both theoretical and empirical – that measured the impacts of the devaluation of domestic currencies of a wide range of countries. However, general equilibrium analyses in this regard are rather limited—more specifically to Asian poor economies. Therefore, we formulate a general equilibrium model and apply it to the case of Nepal and then measure major macroeconomic and distributional impacts of currency devaluation of this economy. Conventionally, nominal devaluation is understood to result in expenditure switching, increased production of tradables, higher exports and improvement in the current account in international trade. However, a number of authors, such as Edwards (1986, p. 1), attack this notion. According to them, though nominal devaluation may achieve its goal of generating a relative price readjustment, this stabilisation process may be very long and painful. This high cost is mainly because of the decline in total output. This is sometimes referred to as a contractionary devaluation problem. Why are devaluations sometimes contractionary? The first reason is the immediate contraction in aggregate demand. The devaluation raises the general price level of imports, and consequently, the import-based economy moves towards demand compression. This ultimately results in negative real balance effects. The second effect of devaluation on aggregate demand is through the change in income redistribution from groups with low marginal propensity to high marginal propensity to save, resulting in a decline in aggregate demand and output (see Alejandro, 1966 and Krugman and Taylor, 1978). Moreover, in a situation of low export and import price elasticities, the trade balance worsens, leading towards recession. There are also some supply-side analyses; Van Wijnbergen (1986) developed a model with intermediate goods and informal financial markets where, under certain conditions, the devaluation can result in an upward shift of aggregate supply. According to his model, once this supply side channel is introduced into the analysis, it is possible for devaluations to be contractionary even if the net effect on aggregate demand is expansionary. This is a situation where the expenditure-switching effect dominates the expenditure-reducing effect. To be more specific, according to the author, devaluation is contractionary under the conditions of: the domestic currency costs of intermediate imports, wage indexing (in the form of explicit contracts, implicit arrangements, or social pressure) with foreign goods present in wage earners’ consumption bundles (namely, food imports), and a reduced volume of real credit to firms. This last channel has its impact on the supply side of the economy because firms in need of funds to finance working capital are pushed into the informal financial market if bank credit is reduced; as a result, interest rates increase and the aggregate supply curve shifts back. This last contractionary effect is obviously exacerbated if the devaluation is accompanied by a cut in the nominal volume of bank credit, as is often the case. The existing partial analyses on the effects of devaluation on real economic activities are mixed, some suggest expansionary effects and others contractionary effects. Connolly (1983) analysed the effect of a nominal exchange rate on the rate of economic growth. The coefficient obtained was positive and marginally significant, providing some support to the hypothesis of expansionary devaluation. The study by Gylfason and Risager (1984), using the imputed parameter data, suggests that devaluations are generally expansionary in developed countries and likely to be contractionary in developing countries. The reason behind the contractionary devaluation in developing countries is the rise in the prices of imported intermediate products. This causes a decline in aggregate demand in the economy in both final consumption and intermediate consumption. Furthermore, this contraction is also reinforced by debt servicing. This is due to the price effect of devaluation. In the case of developed countries, this contraction also applies, but the force to raise aggregate demand is strong as well. The developed countries are able to reap the benefit of devaluation by export growth which they can enjoy with a fuller utilization of their production potential. On the other hand, developing economies face the resource crunch rather than enjoy the widened foreign markets when they undergo contraction. Likewise, the simulation model of Gylfason and Radetzki (1985) suggests that devaluation results in a decline in output and the extent of contraction increases in the presence of indexed wages. Christopoulos (2004), using panel data unit root tests and panel cointegration tests, examined the effect of currency devaluation on output expansion in a sample of 11 Asian countries over the period 1968–1999. The results suggest that, in the long run, output growth is affected by currency devaluation in the majority of countries in the panel as a whole. This finding stands at variance with other recent studies, which concluded that devaluation does not exert any important influence on aggregate output. In the distributional ground, Ripoll (2004) demonstrates that in the long-run real devaluation redistributes income towards unskilled labour, while real appreciation favours skilled labour. In addition to the partial analyses mentioned above, some general equilibrium analyses of the effects of devaluation are also available. The study by Branson (1986) supports the findings of Gylfason and Risager (1984) discussed already. On the other hand, Taylor and Rosensweig (1984), using a large computable general equilibrium (CGE) model for the Thai economy, simulated the effect of a 10% devaluation of the Bhat in real exchange and demonstrated an increase in real output by 3.3%. A more comprehensive study in this regard was conducted by Edwards (1986). This study, using the data of 12 developing countries for the period 1965–1980, showed that the immediate effect of devaluation is contractionary, after 1 year it is expansionary, and in the long run it is neutral. A similar finding was reported by Upadhyaya (1999). He estimated the effect of currency devaluation on the aggregate output level in six Asian countries. The estimated results suggest that devaluation, in general, is neutral in the long run. According to the CGE analysis conducted by Nu San et al. (2000) to study the real effect of devaluation on growth, production, deforestation, and income distribution in the Sumatra region of Indonesia, regional exports increase mainly from the non-agricultural sectors, and imports decline. The agricultural terms of trade – agricultural prices relative to non-agricultural prices – decline due to the smaller share of agricultural exports compared to other exports. Deforestation is likely to increase because the demand for forestry products increases, both as final products and as intermediate products for the wood processing industry, both of which are sold in international markets. An export tax on processed wood proved to be essential to discourage further deforestation in the region. Schweickert et al. (2005) simulated the macroeconomic and distributional effects of exchange rate policy in the highly dollarised economy of Bolivia. Accordingly, dollarisation appears to matter more through real than through financial sector effects. Furthermore, the potential of nominal devaluation to smooth the adjustment path after a negative shock primarily depends on the absence of wage indexation. Regarding the distributional effects, nominal devaluation in no circumstance reduces the poverty effect of the external shock. Lay et al. (2004), in their analysis of the effects of major external shocks and policy reforms of Bolivia to achieve pro-poor growth, concluded that the shocks have not only contributed to the economic crisis, but that they are also likely to impair Bolivia's medium-term development prospects, leading to marked increases in both urban and rural poverty. If the reform projects were implemented, their impact on growth would be large enough to slightly overcompensate the impact of the negative external shocks. The poverty increase caused by the shocks would be more than offset for urban households, but reinforced for rural households. Fagernäs (2004) examined the likely income distribution impacts of alternative policy instruments for stabilising the economy, specifically for eliminating the current account deficit, viz. devaluation or fiscal adjustment. This paper shows that devaluation is likely to be contractionary for GDP in all sectors except agriculture and export-oriented sectors. Profits and labour incomes in commercial farming rise, but fall in other sectors. In this paper, we use the general equilibrium approach to measure the impacts of potential depreciation of the domestic currency in exchange rate liberalisation, similar to the study Ferreira et al. (2004) conducted for the Brazilian economy. The social accounting matrix (SAM) at macrolevel is the major database to this study. Moreover, we have categorised households into poor and non-poor groups. The poor households have been further sub-divided into poor and poorer households (see Appendix A for details). We measure the changes in household income for rich, poor and poorer households following the policy simulation. The rationale behind this exercise is that the devaluation of the domestic currency has been one of the most important policy levers among different components of the conventional structural adjustment programme of the IMF and World Bank in developing countries such as Nepal. The major goal of this policy was to promote exports, reduce imports and thus narrow the deficit in the balance of payments. Nepal did devalue its currency by about 15% with reference to the currencies of major trading partners in 1987. Moreover, even after that, the Nepalese currency shows a long-term depreciation by about 100% during the last two decades in the course of liberalising its foreign exchange market; however, its explicit impact on the overall economy has seldom been assessed. Here we examine such effects using the computable general equilibrium (CGE) model. The rest of the paper is structured as follows. Section 2 explains the database of this study. Section 3 establishes the general equilibrium model and explains the slight modification of the conventional neo-classical version of the CGE model. Section 4 presents the simulation results of the devaluation of domestic currency in Nepal. We make three experiments in this section – devaluation by 5%, by 10% and by 20% – and analyse the results. Section 5 presents the changes accounted in selected macroeconomic features following these three simulations. Section 6 conducts some sensitivity analyses to find out whether the results of the simulations are significantly different with respect to the change in the values of some model parameters. Section 7 concludes the paper.

A devalued exchange rate reduces imports, increases exports and enhances production and domestic supply in developing economies, for example Nepal, which is consistent with neo-liberal doctrine. Moreover, devaluation as a major policy lever of structural adjustment seems quite useful from the short-run viewpoint of growing domestic activities and the foreign exchange balance. But the growth is not at unison. Agricultural and industrial activities expand, whereas services activities contract. This relates to the nature of Nepalese economic fundamentals. Nepalese exports mainly consist of industrial products; therefore, industrial production grows faster after devaluation. Moreover, because of the agro-based nature of Nepalese industries, the agricultural sector also expands. These expansions in agricultural and industrial activities require the partial transfer of factors from services activities. Consequently, contraction occurs among the latter. However, if the rate of devaluation is high, contraction also occurs in agricultural activities. In that situation, only the industrial activities – the most tradable activities – undergo expansion at the cost of contraction of all other activities. Overall, GDP grows. Household incomes also grow, more so among richer household groups and less among poorer household groups. Therefore, devaluation is not so attractive from the distribution viewpoint in Nepalese economy. Furthermore, the government budgetary situation worsens owing to the rise in fiscal deficit. Nevertheless, the saving-investment ratio and export–import ratio improve. This paper draws some important policy implications for a typical developing and transition economy similar to Nepal. Although a small and regular depreciation of the domestic currency contributes to economic growth, the growth is pro-rich because the rate of returns to high-skilled labour and capital grow faster as compared to that of the low-skilled labour. Therefore, conventional stabilisation policy may bring pro-poor growth under depreciation of the domestic currency only if accompanied with a skill-upgrade of low-skilled labour belonging to poor household groups. Furthermore, the government must be cautious of the growth of the fiscal deficit as an impact of currency depreciation.