Continuation of sub-topic- Third cause of INR Shortage: Terms of Trade Deterioration
When trying to solve the INR problem, it is not only important to consider merchandize trade but also trade in services. Tourism, which is an invisible export, helps offset the current account deficit of our country. However, the per capita per night difference in gross earnings from regional tourists and international tourists must be calculated in a rigorous way to show which kind of tourism will aggravate current account deficit and which one helps on the margin. Gross earning per regional tourist is estimated to be 476 dollars while gross earning for an international tourist is estimated to be 5044 dollars. One additional international tourist can generate the same gross earning as 10 regional tourists. It is easy to see which one is smarter move. Bhutan should go more international tourist and minimize regional tourism. In 2012, gross earnings from $ tariff paying tourists was $ 62.8 m, royalty to the government was $ 16.6 m; tour operators net was $ 43.96 m. Total gross earning was estimated to be $ 227 m. In the same year, a regional tourist spent around INR 33,718 ($ 616.6) for a trip to Bhutan. Total gross earning of regional tourist was estimated to be $ 31.26 m. These numbers concerning regional tourism earning seems to be overestimated and needs further research. As the number of regional tourists increase, more budget hotels will have to be built, and construction materials and labour have to be imported. More fuel and food will have to be imported to support such tourism on a continuous basis. Promoting regional tourism will be fuel, car and budget hotel construction intensive compared to international tourism. Increasing imports of cars, fuel and construction materials, the main components of imports from India, will lead to current account deterioration. In this unexpected manner, regional tourism will contribute, though it will not be a decisive factor, to further deterioration in current account balance for some time.
Forecasting Balance of Payments
We know roughly the liabilities on reserves level arising from debt service payment. But we need anticipate with greater salience the pressure on the currency reserves from the current account and financial transaction accounts. They have implications on reserves level. Extrapolating from the past trend, imports of goods and services from India will place far more demand on INR reserves in future than the country has been accustomed. Merchandize imports from India have grown at an annual average rate of 17% between 2000 and 2011. INR import increased from 7.4 b in 2000 to 35.1 b.
The time path of INR earning and reserves should be foreseen for over a much longer period for two reasons. Firstly, it will be useful to assess the capacity for INR debt service payment. Secondly, it will be useful to assess the capacity for INR financing of current account deficit. Forecasting is difficult yet necessary to do taking into account probabilities of various factors including the likely level of electricity tariff.1 Changes in scenarios of future INR earnings and reserves level can be updated regularly by feeding information on changing circumstances bearing on energy prices, capital costs, construction schedule etc.
While long-term structural conditions for earning INR are being improved, expenditure in general has to be redirected to domestically produced goods and import intensive consumption component of the GDP have to be contained from sharp increase. Simultaneously, expenditure switching to domestic goods and expenditure reduction has to occur to bring both internal and external balance. Consumption, made up of government and private consumption, is growing at an impressive rate. It was about 22 b in 2006 and it is raised to 52 b in 2011. Investment was about 52 b in 2011, incidentally equal to the size of consumption. These were two parts in the GDP of 85 b for 2011 along with a current account deficit of 18 b. Until earning in INR improves substantially, financial stability has to be maintained through other means. However, the estimates need major improvement to have confidence in them. Statistical discrepancies in the estimates of consumption component of GDP from 2006 to 2010 are so large that interpretations have to be extremely cautious. For example, 2009 consumption level was estimated to be 44.6 b but statistical discrepancy was of huge order at 8.4 b, which is almost 19% of the total.
Conclusions for Decision Makers
Foreign currency reserves are affected by change in capital account, current account and financial transaction accounts. Better estimation of likely burden on foreign currency reserves in all three accounts over a longer period of time than it has been done in the past are necessary to avoid abrupt responses. So far the practice has been to estimate debt service payment on capital account only. Even on capital account, the volume of inflow during the construction period of hydropower and the estimation of the construction cost needs to be sharpened. The amount of INR that will flow out to repay the debt contracted for hydropower projects will have a critical impact on the INR reserves as well as revenue of the government. It is not only the amount but the timing of outflows and inflows that matters.
Current account deficit has been continually deteriorating. How much it will worsen should be forecasted and quantified in order to find ways of controlling it. Assessing the size of the current account deficit has become a critical exercise to be undertaken.
In containing current account deficit, suppressing import oriented private consumption through taxation is unlikely to be effective. Private consumption is anyway not growing at a fast pace unlike government consumption and investment. Changing the scale of government consumption and government investment will be instrumental in containing current account deficits. Increasing import duties on certain imports will reduce current account deficits if the goods have elastic demand, i.e., price elasticity of demand is greater than one. It is doubtful whether this is the case with many of the goods such as cars currently subjected to high import tax although it generates revenue for the government. Motor vehicle tax contributed revenue of 177.5 m in 2011/122 up from 117 m in 2008, while the green tax if imposed will add another 3 m assuming 3000 plus light vehicles are imported. Ultimately it is not import tax on few selected imported items that will make a difference to the current account deficit but how the government expenditure in totality can be switched to domestic products, after it is raised from import tax and other taxes or is financed by foreign aid. So the attention to fiscal policy in terms of expenditure efficiency, expenditure reduction and expenditure switching to the domestically produced goods has to be the focus far more than how to raise taxes to match rising recurrent expenditure.
Controlling the civil service size and para-statal organizations by increasingly substituting permanent recruitment by contractual services, decentralization to gewog functionaries and generating employment elsewhere are keys strategies of lowering recurrent expenditure. The returns to the recurrent expenditure might have leveled off already because of low average productivity brought about by a large number. The civil service system is becoming too complex and unaffordable. Taxation and revenue base have to be increased substantially if the civil service size keeps on increasing every year by 5% or so. It is not only the salaries and wages that will continue to impose huge costs on society. The pension system will become negative in 2033 under present system and a substantial injection of public resources will be needed to keep it operational.
Employment has to be however generated even though the civil service size can be contracted without any loss in its output. The substitution of foreign labour by Bhutanese would also make a substantial difference to the INR problem. INR 5 b goes out as remittances. But the recruitment of Bhutanese people will drive up the wage-costs and result in higher construction costs and this will itself lead to slowing down of construction sector, which is desirable in the foreseeable future, until the current account deficit becomes manageable. Where there is substitution of foreign labour by Bhutanese labour those goods produced will also become more expensive. If labour intensive exports are produced, exports will fall because of uncompetitiveness. Labour share of value added in exportable may not be very high in Bhutan; so it will not affect export competitiveness, except agricultural produce like apple, orange, etc. But in agricultural sectors, foreign labour is not involved.
The designation of urban centres by the Ministry of Human Settlements and Roads has spurred urbanization. There are now 60 urban sites waiting for credit. Without changing the nature of urban planning and methodology of construction including the Building Code and the minimum plot size, proliferation of concrete buildings will stimulate credit creation, import of building materials and contribute to the perpetuation of INR shortage. The National Urbanisation Strategy 2005 needs to be re-examined not only from structural perspective. It is causing cultural transformation, outmigration from the rural areas, and abandonment of agriculture and food insufficiency. In its place, we need a clear set of policies to support rural renewal. The set of policies should include a redirection of credit from financial institutions. At present, the financial institutions supply only a negligible amount of credit to rural sector.3 For the last 50 years of development, the banking sector has focused on urbanization. The rapid urbanisation came about due to two fundamental economic factors: easier credit and cheaper labour related to rural areas and changes in them can equally change human settlements pattern in favour of what can be financially and ecologically more sustainable.
While substantial progress has taken place in improving national accounts statistics and consumer price index, there are uncomfortably large statistical discrepancies in national accounts statistics. It is also possible that there is huge overestimation in some sectors of the economy such as agriculture and livestock while there is usual underestimation in hotelling and restaurant. In the agricultural and livestock sectors, extrapolating data from the census of 2010 by annual upward adjustment is not at all suitable practice. Yield based on given acreage of Thimphu district is for example extrapolated. According to Agriculture Statistic 20104, Thimphu should have 879 acres of paddy land producing 1,499,000 Kgs of paddy at an average yield of 1,706 kg per acres. But the land use pattern has changed rather radically since 2005 by allocating substantial part of paddy fields for urban development. In addition to land use pattern changes, labour costs have changed rapidly in places such as Paro and Thimphu for the production costs down the value change analysis not to change and these factors would change the gross output, gross value and intermediate costs in agriculture and livestock sectors.
In general, the present simplistic method of extrapolating GDP growth on the basis of past trends is fraught with misleading statistics which will in turn mislead policy makers. GDP forecast for the 11FYP and beyond are admittedly done on this basis, although hydropower investment is taken into account. Even in hydropower investment effect on construction sub-sector, it seems that the estimation is not rigorous. Nonetheless, GDP thus derived are then used as denominator for estimating tax ratios, budget deficit, and a series of other ratios. While this simplistic method might not be inappropriate in some sub-sectors, it is extremely risky for sub-sectors where structural changes are occurring rapidly. Production, cost and revenue functions in such sub-sectors are susceptible to shifts. Microeconomic sample studies should be carried out to build sharper estimates that would contribute to better national accounts statistics and thus to macroeconomic forecasting and policy making.
Statistical discrepancies in the estimation of consumption, saving rates, etc. make it difficult to calculate any elasticity that can be inputs in forecasting, for example, imports. There is an urgent need to improve statistics, not necessarily by widening or increasing the frequency of collection of information but by improving the accuracy of what is being collected. Methodological and definitional issues further plague several areas including employment statistics.
As a part of fiscal expansion, leapfrogging pay and salary of some part of public sector employment have been responsible for wage push inflation. The claim that the higher wages and salaries are reward for performance in terms of profit or greater productivity in financial terms is not fully defensible. Such justification is based on narrower reasoning. Increases in their productivity or company’s profit is the result of capital investment made by the state or by resources and factors which are collective, and there is no deep justification for privatising the returns in the form of higher salaries and entitlement of certain public sector employees. Furthermore, these companies do not operate in a competitive environment. Abnormal profits can be made in spite of normal performance by workers and management because they are monopolies.
Over the last two years the control of M2, the measure of broad money, has been fairly effective. It has not grown much over last 2 years. Likewise the level of reserve money has fallen slightly. However this has not been translated equally into control of domestic credit and consequent reduction of INR shortage.
It is not so clear at this stage as to which measure of money fits the best definition of money in the context of Bhutan. Supply and demand functions for money have to be modelled in order to understand their behavioural functions in the context of Bhutan. We could set up simple regression models in order to estimate parameters of the relevant independent variables with the objective of leading us to better monetary planning and policy. M2, reserve money and M1 are, as mentioned above, have been kept stable in the last two years. In addition, a variety of instruments to control domestic credit, have been applied. Yet domestic credit is on an upward trajectory. Domestic credit was 53.4 b in March 2013 up from 45.3 in March 2012. Money supply is a product of money multiplier and high-powered money that includes reserves. Since money supply as measured by M2 or M1 has not increased significantly in the last two years, it is the money multiplier, along with the variables that affect it, that has to be targeted more clearly. Money multiplier is dependent on the behaviour of the public and the banks.5 In any case, much more quantitative work with a view towards building predictive frameworks are needed rather than simply gathering data which are not interpreted insightfully for policy purpose.
The volume of government expenditure is likely to increase on average 15% or so annually due to the increase in the size of 11 FYP. At the same time, hydropower investments is likely bring exponential rise in expenditure in the country. The exact amount invested each year in hydropower sector will depend on the number of projects that will be executed every year. One can easily calculate different scenarios depending on the cost escalation and staggering of hydropower project construction schedule. Different scenarios give us rise different policy options for debt stock, debt repayment and revenue streams. Such exercises should further inform decisions on public expenditure choices.
Given the low absorption capacity an unusual increase in expenditure (E) will lead to current account deficit of magnitude not experienced so far. Technically, absorption capacity E ≡ C+I+G, where C is consumption, I is investment and G is government consumption. Total spending is usually summed up in GDP, though gross national disposable income (GNDI) is the most appropriate measure of spending for Bhutan. GNDI is 11-15% higher than GDP because of relatively large transfers to Bhutan. But let me use GDP in this example. In identity terms, absorption rate is given by GDP – E ≡ X-M, where X is exports and M is imports of goods and services. As spending rises, spending on domestic goods rises but it rises at a rate far less than income because part of income is saved but an overwhelming part of it is spent on imports. Bhutan’s GDP was 85 b in 2011 when its current account deficit was 18 b. Expenditure increase will lead to GDP increase and given the high marginal propensity to import and high money multiplier, it will lead to current account deficit that will not be fully covered by grants and capital inflow in hydropower investments.
To be continued
By Dasho Karma Ura