Back Matter

APPENDIX I Statistical Issues and Measurement of GDP

36. Official statistics have only recently attempted to include estimates of activity in the informal sector. Even in the formal sector, there is a bias towards the coverage of older enterprises and traditional sectors (such a heavy industry), which have seen some of the sharpest declines in output. Partial attempts to improve the coverage of statistics have, however, led the authorities to underestimate real growth and overestimate the GDP deflator.

37. Under the former Yugoslav system, with the vast majority of production taking place in socially owned enterprises, it was easy to keep track of aggregate production. With the shift to a more decentralized market economy it became necessary to overhaul statistical collection methods to capture the activity of new private enterprises and the growing informal sector. In addition, concepts such as Gross Social Product (which excluded output of the services sector and government) were replaced by market concepts such as GDP.

38. Initial estimates of GDP were derived from rather narrowly based volume indices (in particular the industrial production series discussed in section B above) and estimates of sectoral deflators. However, the Statistical Office (SO) is in the process of upgrading its surveys and techniques to generate consistent figures for GDP on a value added and expenditure basis in line with international standards.

39. In 1997 new estimates were introduced, based on surveys of value added at current prices which, for the first time, covered the informal sector as well as improving coverage of new enterprises in the formal sector. Nominal GDP for 1993 was revised upwards by 3.3 percent, while nominal GDP for 1996 was revised up by 8.1 percent. Current estimates for GDP growth in 1997 are based on the old methodology and growth in 1997 is likely to be revised upwards when estimates based on the new surveys are published.

40. Despite substantial changes in nominal GDP growth, there have been only minor changes to estimates of real growth. This is because volume indices have not yet been revised to capture new enterprises or activity in the informal sector. Deflators for each sector are calculated by dividing value indices by volume indices. As estimates of the informal and new formal sector are included in the numerator but not the denominator, a bias is introduced into the calculated deflators. In fact, the calculated deflators capture the entire growth of the informal and new formal sectors as measured by the new nominal surveys. This bias is a temporary phenomenon and will be corrected once the volume surveys have been updated. There are two ways to estimate the size of the bias. First, by assuming that the initial estimates of the GDP deflator were accurate and the higher nominal output captured by the survey reflected higher real output, real growth between 1994 and 1996 is estimated to have averaged 0.4 percent per year compared to the official -0.7 percent. An alternative approach is to construct a GDP deflator from a weighted average of CPI, producer prices and import and export prices. This implies an average real growth rate over this period of 0.9 percent (see Table 2).

Table 2.

FYRM: GDP Estimates, 1993–97

article image
Sources: The authorities, and IMF staff estimates.

This method becomes inaccurate at high levels of inflation.

APPENDIX II The Social Safety Net

Unemployment Benefits

156. Unemployment benefits account for 1.3 percent of GDP, or over one-third of all expenditure on social programs. As of end-1997, there were 258,000 registered unemployed, accounting for 32.2 percent of the estimated active working population. Of these, only 49,881, or 19 percent of the total, received unemployment benefits as only the unemployed with a work history are eligible for unemployment benefits and the eligibility period is time bound. As described before, recipients of unemployment benefits also have pension contributions paid on their behalf. As all registered unemployed receive free health insurance, it is widely suspected that a significant number of these individuals actually do work in the private sector.

157. The average unemployment benefit stood at 32 percent of the average wage in 1993. This replacement rate was raised in 1995 to 50 percent, a value higher than in most other transition economies. In July 1997 a new Unemployment Law was enacted, whose main provisions are as follows: (i) the replacement rate falls to 40 percent after receiving benefits for 12 consecutive months; (ii) the maximum duration of cash benefits for persons with 20-25 years of insurance is 18 months, while the previous law gave 24 months to persons with over 22 years of insurance; (iii) unemployed workers receiving cash benefits must register monthly with the Employment Fund—all the other unemployed must register every 6 months. Further, the special one-time payment to the unemployed who claimed a pension, of 40 percent of the average monthly net wage, was abolished.

158. Any employer who hires an unemployed person receives, as a wage subsidy, any unpaid unemployment benefits, for up to a maximum of 18 months. On balance, this incentive may facilitate the re-employment of the unemployed, at little budgetary cost. However, wage subsidies are in general easier to justify if targeted at the long-term unemployed.

Social Assistance Program

159. About 1.7 percent of GDP is spent on the Social Assistance Program (SAP). This provides financial support to the disadvantaged through four distinct schemes. First, the SAP provides Social Assistance Benefits to persons who are able-bodied but ‘materially endangered’, that is, do not receive sufficient income from other sources (such as wages, pensions, or unemployment benefits), and own few assets. This scheme, which had 64,000 recipients at end-1997, accounts for most users of the SAP and 85 percent of its 1997 expenditures. Typical recipients include: (i) any unemployed who are ineligible for unemployment benefits, submit documentation of their poverty, and have their indigence confirmed by site visits; (ii) employed persons whose wages have not been paid for more than 3 months; (iii) pensioners receiving the minimum pension. Each recipient receives the difference between the national poverty line36 and his ‘potential income’. The latter is defined as the recipient’s net income, excluding any government compensation for handicaps or natural disasters; examples of such excluded items are civil disability pensions, allowances for disabled children, and disaster relief compensation.

160. Second, the SAP provides Cash Compensation for Care of the Disabled to anyone who regularly helps the disabled with assistance in everyday life. This scheme accounts for 8 percent of SAP expenditures. Third, the SAP provides Permanent Cash Compensation to anyone who is incapable of work and is materially endangered. Examples include people with mental or physical disabilities, and orphaned children. This scheme covered over 10,000 people at end-1997, and accounts for 4.5 percent of SAP expenditures. Finally, the SAP provides One-Time Assistance for emergency purposes. This includes disaster relief compensation, as well as assistance to people who are eligible for but have not yet received Social Assistance Benefits. On average, this scheme is used by 6,000 people every year and accounts for 2.5 percent of SAP expenditures.

Withdrawals from Frozen Foreign Currency Deposits for Social Needs

161. For the last four years individuals have been able to make withdrawals from frozen foreign currency deposits on the grounds of ‘social need’. Such withdrawals amounted to 0.5 percent of GDP in 1997. Three distinct types of withdrawals are allowed. First, individuals who are ‘materially endangered’, and present a certificate to this effect from the Center for Social Affairs, can withdraw up to DM 100 per month. This accounts for the vast majority of withdrawals. Second, one-off withdrawals of DM 1,000 are allowed in connection with family weddings or deaths. Finally, unlimited withdrawals are allowed to finance the education abroad of one’s children.

APPENDIX III Balance of Payments Statistical Issues

200. The compilation and coverage of balance of payments data has improved over the last four years. The customs offices were fully computerized in 1996 following the ASYCUDA system, while the NBM redesigned the international transactions reporting system (ITRS) in 1994 and has since then continued to strengthen the methodology to estimate the current account. Nevertheless, errors and omissions and unidentified flows have remained high, accounting for as much as 7 percent of GDP in 1997.

Revisions in trade data

201. Responsibility for balance of payments data is shared by the SO, Customs Administration (CA), and the NBM. The SO publishes trade data on a monthly basis and provides additional data to the Ministry of Development and the NBM for the compilation of the balance of payments statistics.

202. A number of errors were discovered in the external trade data in 1996 and 1997, prompting substantial revisions and a review of procedures. A committee comprising officials of the SO, the CA, and the NBM was set up in June 1997 to oversee the collection and the compilation of balance of payments statistics, implement a single customs documentation for reporting exports and imports, and revise, where appropriate, the methodology used to compile the trade statistics. A lack of experience in data collection,52 and lack of appropriate human and technical resources are being addressed with financial assistance through the EU PHARE program and technical assistance from the Fund. New quality controls have been introduced for the processing of data at the SO and the NBM is also providing technical assistance in the formulation of a new data processing software at the SO to computerize the compilation of trade statistics under the new single customs documentation. The SO, NBM, and CA will shortly sign a binding protocol that will ensure the quality of the collection, transfer and processing of trade data in the future.

203. While the revisions to the trade data for 1997 were implemented, data prior to 1997 have not yet been revised implying a potential discontinuity in the time series.

Errors and omissions

204. The balance of payments also suffers from large errors and omissions which reflect an incomplete coverage of current and capital account transactions. Three areas of the external sector have been identified where coverage is particularly weak: remittances; the drawdown of foreign exchange held; and the valuation of trade.

205. There are an estimated 500,000 citizens from the FYRM living abroad, compared with a total population within the FYRM of 2 million. As in other cases, it is customary for ex-patriots to remit some of their foreign earnings to relatives at home, either in the form of foreign currency or goods. In 1996, out of 4,800 registered traders that were solely importers, 4,000 imported less than US$75,000 each, for a total of US$73.8 million (about one-third the size of errors and omissions). A large part of these imports are thought to have been financed directly by family members living abroad by drawing down foreign exchange deposits in the country of export. These transactions would imply an under recorded level of remittances in the balance of payments, contributing to an overestimated current account deficit.

206. Anecdotal evidence suggests that citizens living within the FYRM hold large foreign currency deposits abroad. The international transaction reporting system used by the NBM does not record changes in foreign currency deposits held abroad, implying that any movements in these assets that are used to finance the current account are not reflected in the capital account. They thus accrue to errors and omissions. Without further information, it is not possible to tell though whether this would imply a higher or lower capital account balance.

207. In addition, there is anecdotal evidence that FYRM citizens hold an unusually high level of foreign currency outside of the banking system. Unofficial estimates put the amount of foreign currency circulating in the economy as high as US$800 million (25 percent of GDP). This level of currency substitution has clear implications for the balance of payments: as in the case of other transition economies,53 any reversal in currency substitution would be reflected in large unidentified financing of the current account, leading to high errors and omissions. An example of this phenomenon may have actually taken place after the July 1997 devaluation, when private transfers doubled in the second half of that year, and errors and omissions jumped to 7 percent of GDP from 4 percent a year earlier.

208. Finally, there is anecdotal evidence that trade flows are undervalued. A number of exporters, for example, are reported to understate export volumes in customs declaration, particularly when barter trade is involved, so as to meet the requirement that a fixed percentage of the barter be transacted in cash. There is also evidence of unrecorded cross-border trade with FRY and Albania, which is thought to result in additional net exports. Similar unrecorded cross-border trade is reported to occur with Bulgaria and Turkey, but is thought to lead to additional net imports. On balance, it is hard to assess whether unreported trade would increase or reduce the trade deficit in FYRM.


  • Baliño, T., J. Dhawan, and V. Sundararajan, 1994, “Payments System Reforms and Monetary Policy in Emerging Market Economies in Central and Eastern Europe,IMF Staff Papers, Vol. 41, No. 3.

    • Search Google Scholar
    • Export Citation
  • Kaufmann, D., and A. Kaliberda, ‘1996, “Integrating the Unofficial Economy into the Dynamics of Post-Socialist Economics; A Framework for Analysis and Evidence.” In, “Economic Transition in Russia and the New States of Eurasia,Kaminsky, ed. Armonk NY: N.E. Sharpe.

    • Search Google Scholar
    • Export Citation
  • Halpern, L., and C. Wyplosz, 1997, “Equilibrium Exchange Rates in Transition Economies,IMF Staff Papers, Vol. 44, No. 4.

  • International Monetary Fund, 1998, “Trade Liberalization in IMF-Supported Programs

  • National Bank of the Republic of Macedonia, 1997, Annual Report

  • Suklev, B., 1996, “Privatization in the Republic of Macedonia”, Eastern European Economics, November-December.


This section is based on official statistics which may underestimate informal activity. This issue is discussed in more detail in Appendix 1.


This method will underestimate activity if producers have become more efficient uses of electricity. However, this is less of a problem than in other transition economies because the price of energy has not changed as much during transition. This method overestimates value added to the extent that large, energy intensive loss makers continue to operate.


Official statistics showing GDP on an expenditure basis only exist up to 1994, Figures since then are based on staff estimates and are subject to wide margins of error.


Within-year data on fixed capital formation is inconsistent with data collected on an annual basis. The coverage of enterprises is much narrower and the series is highly volatile. However, broad trends can be distinguished and suggest that investment was weak in the first half of 1997 but bounced back in the second half.


The law covered about 80 percent of employees including those in companies which had not fully paid off privatization installments.


For more details on the law, see Chapter VI.


In theory, management buyouts were not necessarily to existing management but to those who would take on the task of management. In practice, this was nearly always existing management.


The BPO, which was required to resubmit enterprises after accounts were blocked for an additional 60 days, found they were sometimes submitting the same enterprise many times.


Listed companies are required to publish their accounts, hold their share registries at the central share registry and are not permitted to restrict trading of their shares. To avoid these regulations, many enterprises choose not to list. During 1997, all companies are required to move their share registries to the central share registry, remove restrictions on share trading and the larger companies are required to provide accounts to shareholders. This will eliminate the most important barriers to listing.


Non-performing loans defined as loans classified into risk category ‘C’ or worse.


Commercial banks must hold 100 percent reserves against resident foreign currency deposits in the form of liquid foreign currency assets with overseas banks.


Defined as the 20 largest enterprises that had any loans from banks classified as ‘B’ or worse. Exposure includes new credits, accrued interest, as well as off balance sheet items (guarantees and letters of credit). The figure does not adjust for exchange rate valuation effects. Given estimates of the share of foreign currency exposure to these 20 enterprises as of September 1997, the inferred growth of total exposure between March and December, adjusting for exchange rate effects, is about 14 percent.


This calculation compares interest income reported as collected by banks—adjusted for provisioning—in their profit and loss accounts, with their annual average loans outstanding.


Comparing guaranty capital with risk weighted assets as per the Basle norms.


While there is some variation across banks, all banks met the Basle minimum standards. One savings house did not fulfill the minimum capital adequacy standards and is under close surveillance by the Supervision Department.


Banking system return on assets increased from 1.2 percent in 1996 to 2.2 percent in 1997, and return on equity from 4.3 percent to 9.3 percent.


The NBM has announced its intentions to vary this prudential ceiling as a means of signaling its credit stance to banks in advance.


In Macedonian, Zavod za Planten Promet (ZPP)—the Institute for Domestic Payments.


Large net float positions impede the ability of a central bank to effectively implement monetary policy. This has been a major focus of the inter-action between payments system reform and the conduct of monetary policy in other transition economies where large floats impeded early attempts at establishing monetary control. For example net float as a share of reserve money in the first half of 1992 averaged some 5 percent in Poland and 16 percent in Bulgaria, compared with 0.09-0.027 percent in the United States (Baliño, Dhawan, and Sundararajan (1994)) and virtually zero in FYRM.


In addition to a cut in the average effective tariff rate, preferential trade agreements were signed with Slovenia (February 1996), Yugoslavia (October 1996), Bosnia (January 1997), and Croatia (March 1997). Such liberalization more than offset the positive revenue impact of the tariffication of quotas.


However, some repayment schedules stretch out until the year 2000.


However, the presumptive nature of the FWL as regards social security contributions is reducing the incentive for enterprises to under-report their total wage-bill, which should have a positive impact on income tax collection.


After adjusting for changes in the accounting treatment of local taxes.


Although the supplementary budget was not approved by parliament until September 1997, lower spending targets were enforced from July.


Including assumed pre-privatization from the IFC for Macedonia Telecom.


More precisely, the exemption applies to new employees who: (i) were registered with the Employment Fund as being unemployed on December 1, 1997; or (ii) have either been unemployed for over 1 year, or fall into one of two categories: workers whose firm went bankrupt or workers who were made ‘technologically redundant’.


Prior to 1998, the central government did not contribute to the financing of the Health Fund. However, with the introduction of the PfE, the central government will provide direct transfers to the Health Fund in the event of a shortfall in contributions.


Wages earned prior to January 1970 are not used in the calculation of the pension base given the lack of complete earnings records.


Out of about 225,000 pensioners, 59,000 received survival pensions, and 48,000 received disability pensions.


The number of people aged 65 years or more, to those aged from 15 to 64.


For instance, in a survey of 247 rural primary health-care centers (PHC), virtually every center lacked some basic equipment for routine examination, minor surgical intervention, sterilization and waste disposal. Around 30 percent lacked blood pressure gauges and stethoscopes, 70 percent lacked infant scales and refrigerators, and 90 percent lacked an otoscope. PHC staff was also narrowly trained, and lacked a broad preventive and curative approach to family medicine.


In 1995, the occupancy rate was 61.9 percent in general hospitals, 74.7 percent in clinics, 82.2 percent in specialized hospitals, and 22.7 percent for inpatient facilities. Contrasting FYRM with the UK, the average length of a hospital stay in days was as follows: for infectious and parasitic diseases, 22 versus 7.3; for neoplasms, 18 versus 9.5; for complications of pregnancy and childbirth, 16 versus 3.5.


For instance, a recent World Bank mission found that market prices for a sample of essential primary care drugs were 5 to 20 times the world market level.


Until late 1997, there were two separate poverty lines, one for rural and one for urban residents. The rural poverty line was set at a lower level, allegedly reflecting the lower rural cost of living.


An initial undershooting of the equilibrium real exchange rate was a phenomenon observed in many other transition economies. See Halpern and Wyplosz (1997).


There has been no parallel market in more recent years.


The sanctions included a trade embargo which was first imposed in May 1992 and then intensified in November 1992 and April 1993. The embargo was lifted in November 1995 as part of the Dayton Peace Accord.


These growth numbers should be taken with caution since the authorities are in the process of reviewing the quality of the data before 1997.


The destination of exports was considerably distorted in 1992-95 by the effects of sanctions, with a sharp increase in exports to Bulgaria partly displacing a decline in exports to Serbia. More normal shares of exports to these countries were restored in 1996-97.


Year-to-year changes are subject to potentially large measurement errors. In addition to possible compilation errors, the level of imports in 1995 is probably overstated due to the re-export activity under the UN sanctions implying stronger underlying growth in 1996.


The distortion of the shares of imports from Bulgaria and Serbia in 1992-95 is also similar to the distortions to the destination of exports.


Against this trend, imports from Albania doubled between 1996 and 1997, partly due to better monitoring of trade flows at the border.


The terms of the tradable bonds are: 15 years maturity, with a four-year grace period, and an interest rate of LIBOR plus 0.8125, of which only 3.5 percent will be paid for the first four years, while the remainder is capitalized.


The CSTR index is used by IMF staff to quantify the overall trade restrictiveness in a specific country (see Box). See Annex I, International Monetary Fund (1998).


Additional information is being collected on these standards to assess whether they are consistent with internationally accepted standards.


Ms. Sosa (MAE) and Ms. Weeks (LEG) contributed to this section.


All transactions must be completed by 11 a.m. to allow for the timely processing of the payment order; all “large” payment orders (denar 5 million, equivalent to about US$ 90,000) have to be processed by noon.


In addition to prudential limits on their open foreign exchange position, banks are subject to minimum and maximum foreign exchange liquidity limits, which the NBM sets quarterly for each bank in relation to its capital and volume of foreign exchange transactions.


The NBM sets the indicative rates for each day on the basis of the weighted average of DM/denar rates obtained in transactions among enterprises, between enterprises and banks, among banks, and between banks and the NBM during the previous dealing day. The average rate constitutes the middle exchange rate, from which a buying and a selling rate are calculated by applying a margin of 0.3 percent. Once the DM/denar rate is thus calculated, the cross rates for 22 other currencies are calculated on the basis of quotations in Frankfurt.


In the former SFRY, the bureau of statistics in Belgrade collected and processed all balance of payments data.


See for example the case of Croatia in 1994, or Georgia in 1995.