Other Checking Simulations

In our next check, we shifted export demand and import supply curves vertically upwards by 10%. This implies a 10% increase in the foreign currency price. If the exchange rate is exogenous and held constant, all endogenous nominal variables should be changed by 10%, while there is zero change in endogenous real variables.

Elasticities and parameters 5.3.4

Behavioural elasticities and parameters for ORANIMON are presented in Table 5.5, along with their algebraic notations. The adopted and estimated values for each of them are shown in Appendix 3 and Appendix 4. Thus this section’s purpose is to describe the methodologies and sources, to discuss ongoing and potential analysis regarding elasticities and parameters.

Table 5.5 Elasticities and parameters

Name Algebraic

notation Elasticity of substitution between primary factors 𝜎1𝑃𝑅𝐼𝑀(𝑖) Elasticity of substitution between labour occupations 𝜎1𝐿𝐴𝐵(𝐿,𝑖) Elasticity of transformation between industry outputs 𝜎1𝑅𝑈𝐶(𝑖) Armington elasticity of substitution between domestic and imported

intermediate inputs 𝜎1(𝑐)

Armington elasticity of substitution between domestic and imported inputs

to capital formation 𝜎2(𝑐)

Armington elasticity of substitution between domestic and imported

commodities – household consumption 𝜎3(𝑐)

Export demand elasticities, by commodity and by trading partners _𝛾(𝑐)

Household expenditure elasticities 𝜀(𝑐)

Frisch parameter 𝐹𝑅𝐼𝑆𝐶𝐶

We classify and analyse these elasticities and parameters in four groups. The first group is composed of the parameters related to input demand and commodity supplies in MONAGE. As discussed in Chapter 3, there are two tiers in the production technology, showing the relationship between each industry’s inputs and its activity level, and the relationship between each industry’s activity level and its outputs.

149 The first two elasticities in the first group are related to the substitutability between factors of production. The elasticities of substitution between primary factors,

𝜎1𝑝𝑟𝑖𝑚(𝑖), are concerned with the relationship between each industry’s primary inputs and its activity level. We adopted the values for the elasticity of substitution between primary factors from GTAP 7. Mapped values of elasticities of substitutability between primary factors for 55 industries are included in Appendix 3.

We note that the attempts have been made to estimate the elasticity of substitution between labour and capital across broad industry classes in the case of Mongolia. Firstly, we have attempted to extend an approach initially used by Phipps (1983), and further developed by Rimmer (1990), adopting the zero pure profit constraint, in keeping with the assumptions in ORANI. The attempts to utilize the panel estimation techniques with the fixed and random effects, as well as the error correction models, have been made. Due to lack of detailed data, these are still a work in progress. Ideally, we could estimate short- and long-run general equilibrium elasticities if there is adequately detailed data. However, we will see in the next chapter that the validity and sensitivity analysis reveal the robustness of elasticities– in particular, of the elasticities of substitution between primary factors.

For the elasticity of substitution between labours of different occupations – 𝜎1𝐿𝐴𝐵(𝐿,𝑖) we adopted the MONASH value of 0.35. The last parameter 𝜎1𝑅𝑈𝐶(𝑖) in the first group is a vector of the elasticities of substitution which govern the choice of alternative outputs in industries with CET output functions. We borrowed the ORANI value of 0.50 across all industries for this study.

Second group elasticities are vectors of elasticities which govern the substitutability of commodities from different sources for producers, investors and households, respectively. ORANIMON treats domestic and imported products as imperfect substitutes, with the degree of substitutability governed by these Armington elasticities. These elasticities are important for determining the behaviour of trade flows and are explained in Chapter 3. However, they are very difficult to estimate, and the available estimates vary widely due to the availability and quality of data for their estimation, as well as the differences in the econometric models used to estimate them (Hertel et al. 2007; McDaniel & Balistreri 2003; Okagawa & Ban 2008). Due to the lack of any estimate of these elasticities of substitution between domestic and foreign sources of supply for Mongolia, we adopted the elasticities from the GTAP 7 database. We assume

150 that the commodity-specific Armington elasticities are the same for producers, investors and households.

The third group of parameters is a vector of foreign demand elasticities for Mongolian exports – 𝛾(𝑐). Export demand elasticities are crucial in determining the effects of changes in the volumes of exports on the changes of the terms of trade and hence in analysing the impacts of mining boom.

We calculated the export demand elasticities for commodities in ORANIMON through a synthetic method often used in COPS-style modelling. We note that the attempts to estimate export elasticities at the commodity level by trading partners have been undertaken as well. For this study, however, we use the general commodity-specific export elasticities.

In Chapter 3, ORANIMON distinguishes two major groups of export commodities. The first group of commodities, which exports 20% or more of their total sales, are considered individual exports. They have individual export demand curves, and thus require individual export demand elasticities. There are 28 such commodities in the ORANIMON benchmark database. The remaining 27 commodities are considered collective exports; their export volumes move with the average price index for the collective group.

For the individual exports, we calculate export demand elasticities using the estimates of importers’ elasticities of substitution between different sources of imports and the theory suggested by Dixon and Rimmer (2002).

Foreign importers are assumed to be profit-maximisers who consider importing various commodities from different sources. In addition, they treat these commodities as imperfect substitutes. Finally, foreign importers choose import commodities from different sources to minimise their costs, subject to a CES function. Solving this optimisation problem results in Mongolia’s export demand elasticity for commodity c

with regard to its FOB export price as:

𝛾(𝑐) =�𝜂(𝑐)𝑆𝑚𝑎𝑎(𝑐)− 𝜙(𝑐)�1− 𝑆𝑚𝑎𝑎(𝑐)�� 𝑆𝐹𝐹𝐹(𝑐) (5.4)

where 𝛾(𝑐) is export demand elasticity for commodity c from Mongolia;𝜂(𝑐) is the price elasticity of world demand for commodity c; 𝑆_𝑚𝑎𝑎(𝑐) is a share of Mongolia in ROW’s imports of good c; 𝜙(𝑐) is a foreign importers’ elasticity of substitution between

151 alternative sources of supply; and 𝑆_𝐹𝐹𝐹(𝑐) is the proportion of the FOB price of commodity c from Mongolia in the purchaser price of c in foreign countries.

(5.4) provides a way of calculating export demand elasticities that are consistent with the Armington parameters in a global model such as the GTAP.

If Mongolia is very small in international trade for commodity c (𝑆_𝑚𝑎𝑎(𝑐)≈0), and if there is no difference between FOB price and purchaser price of c (𝑆_𝐹𝐹𝐹(𝑐)), then the export demand elasticity of commodity c (𝛾(𝑐)) would be equal to the negative of the foreigners’ Armington elasticity of substitution between alternative imports (𝜙(𝑐)). This is, in fact, the case for most of Mongolian exports commodities. However, Mongolia is likely to have non-trivial shares of the foreign markets for main commodities (notably, copper ore and cashmere articles) of its exports. Thus we calculated the values for

𝑆𝑚𝑎𝑎(𝑐) and 𝑆𝐹𝐹𝐹(𝑐) to compute Mongolia’s export demand elasticities.

The elasticities were calculated first for 55 commodities using the values of 𝜙(𝑐) from the GTAP 8 database on world imports and Mongolia’s imports and exports of commodities. Mongolia’s export share in world imports of commodity c was calculated as:

𝑆𝑀𝐼𝑀(𝑐) = 𝑀𝐿𝑛𝑓𝐿𝑙𝑖𝑎

′_{𝑖𝐶𝑥𝑝𝐿𝑟𝐿𝑖𝐿𝐿𝑐}

[𝑅𝐿𝑟𝑙𝑙𝐼𝑚𝑝𝐿𝑟𝐿𝑖𝐿𝐿𝑐 − 𝑀𝐿𝑛𝑓𝐿𝑙𝑖𝑎′_{𝑖𝐼𝑚𝑝𝐿𝑟𝐿𝑖𝐿𝐿𝑐]} (5.5)

Mongolia’s export share in world imports of the ‘wol’ (wool, silk-worm cocoons) commodity in the GTAP was the largest at 2%, followed by ‘omn’ (minerals nec) and ‘col’ (coal) with 0.5 and 0.1%, respectively. In the calculations, we adopted the world price elasticity of commodity c pf -0.5 (𝜂(𝑐)= −0.5), the value for the share of FOB price in the importers’ purchaser price of 0.7 (𝑆_𝐹𝐹𝐹(𝑐) = 0.7) for all merchandise commodities and of 1 for all services, following Dixon and Rimmer (2002). The range of 𝛾(𝑐) are included in Appendix 4. The last group in Table 5.5 contains elasticities and a parameter relating to household consumption. We adopted expenditure elasticities from the GTAP 7 database and then scaled them to satisfy the Engel aggregation property of demand systems. The aggregation requires that the sum of the products of income elasticity of each good and its budget proportion must equal unity.

The Frisch parameter shows the relationship between households’ total expenditure and their supernumerary expenditure in the Klein-Rubin utility function. Frisch parameters are used in evaluating own- and cross-price elasticities of demand for households, and

152 in calculating the changes in the subsistence component of household consumptions. The Frisch parameter is defined as the negative of the ratio between total final household expenditure and household supernumerary expenditure. The Engel law states that, as income increases, the proportion of income spent on foods decreases. Similarly, we can expect that the proportion of income spent on subsistence items falls as income increases. That is, on the other side of the token, the supernumerary proportion of household consumption rises as income increases. Hence, the Frisch parameters for developing countries are generally higher than those for developed economies. Likewise, the Frisch parameters for low-income groups are expected to be higher than those for higher-income groups in Mongolia. Even though MONAGE is capable of having a number of household types, we include, for this study, a representative household.

Additional Data for MONAGE