When Ukraine gained independence, it faced a need to reform its statistics system, inherited from the Soviet era. So the country opted for international statistics standards, and implemented an internationally recognized system of national accounts (SNA) with GDP as its key macroeconomic indicator, estimated on a monthly basis. Only two other countries in the world - Australia and Finland - make monthly estimations of GDP. Another innovation is high-quality sample surveys of household living conditions and the population’s economic activity. Ukraine was elected a member of the UN Statistics Commission.

However, there is still too little trust in statistics in a society where the public authorities’ trust rating is as low as 6-8%. No wonder that all of the statistical data, users are most cautious about major indicators affecting the people’s interests. One of those is the consumer price index (CPI).

First of all, we need it for an accurate summarized measurement of the population’s purchasing capacity and a forecast of changes in consumer prices for goods and services. However, it has a much broader application, including gauging the population’s real income and calculating about 60% of GDP accounting for household expenditures. Trade unions and employers use it as a benchmark while concluding collective agreements, and budget-funded institutions use it while making decisions on salary increases.

The index is essential for the economy-related executive power bodies - the Ministry of the Economy, the Ministry of Finance, the Ministry of Labour and Social Policy; for the NBU and its Board, commercial banks, UN agencies and other international organizations, especially the IMF and the ILO. Therefore requirements about the accuracy and reliability of CPI calculations are fairly strict. So are requirements about the transparency of methodologies and practices of its computation and use.

Unfortunately, the current level of disclosure of methodologies and practices of CPI calculations by the State Statistics Committee does not meet even the minimum requirements. Until 2002, descriptions of underlying methodologies in the 60-70 page long “Consumer Price Indices” yearbooks did not exceed seven paragraphs. The deficiency of explanations and interpretations was so striking that in October 2001, the President of Ukraine had to issue an instruction entitled “On Improving the Monitoring of Prices and Tariffs”, demanding that the methodology for price index calculation be enhanced.

Following the President’s instructions, Ukraine’s statistics authorities expanded the methodological interpretations in their yearbooks from seven to nine paragraphs! By way of comparison, explanatory notes accompanying the CPI data in the USA amount to 46 pages, in Poland, to 31 pages. In Ukraine such laconic brevity is the exception rather than a rule: the State Statistics Committee publishes several yearbooks that offer detailed explanations of applied methodology (e.g. “Economic Activity of the Ukrainian Population” - 40 pages, “Expenditures and Resources of Households in Ukraine” - 35 pages). Yet this is not the case with CPI data: in the latest publication in the 2004 yearbook, the methodology description shrank again to one page.

Recently the State Statistics Committee started publishing a more detailed description of methodology for CPI computation on its official website, but the on-line version is also far from perfect. Nor does it address the universal concern over the possible falsification of the CPI data. However, analyzing the information available on the methodology and practices of CPI calculations, including those by regions, on the data aggregation and publication, one could conclude that falsifications are most unlikely. For example, if the calculations show that the index is 111.1% as compared to the previous month but somebody wants to change it to a more acceptable 101.1%, it is not feasible for a number of reasons. In theory, it could be done once or twice, but very soon, such “adjustments” would surface and fall apart as a house of cards because, for one thing, all initial CPI calculations are made regionally, upon which they are collected and aggregated in Kyiv. For another thing, all aggregated and summary calculations are made automatically. Besides, dozens of people take part in index computation in the regions and in the centre, so information on falsifications would inevitably leak. Finally, the IMF closely watches all developments relating to CPI, sending experts to Kyiv to check the dependability of data it feeds to the international community.

However, the lack of transparency and any public control, and serious flaws in the methodology of CPI calculations undermines experts and society’s trust in statistical data on the country’s economic situation and, more generally speaking, in the incumbent administration.

Three years ago, in February 2003, the Supreme Rada of Ukraine adopted a law entitled “On Incorporating Amendments into the Law of Ukraine ‘On Indexation of Monetary Incomes of the Population’”. Articles 11-13 of this law are directly applicable to the consumer price index in that they stipulate that the monitoring of all changes in prices and tariffs should be carried out in a transparent manner and involve the general public, trade unions, women’s, youth and veteran associations, consumers’ NGOs, etc.

The law envisioned establishing CPI expert boards under the Cabinet of Ministers and regional state administrations, with an equal representation of executive power bodies and civil society. Those boards were supposed to study the monthly CPI data and invite independent auditors to revise it, if need be. The Cabinet of Ministers was mandated to develop regulations on the expert boards’ membership and rules of procedure by May 2003.

The government was not in a hurry to comply with the law: the regulations were developed only in January 2005. According to them, the central board was to comprise representatives of the Secretariat of the Cabinet of Ministers, the Ministry of Labour and Social Policy, the Ministry of Finance, Ministry of Economy and State Statistics Committee, leaving behind NBU and its Board, and academic and research institutions. A year has elapsed since the Cabinet passed a relevant resolution but the expert board has not been formed, even though the Trade Union Federation of Ukraine and the Ukrainian Consumers’ Association have nominated their representatives to the committee and notified the Cabinet thereof.

Sadly, the Secretariat of the Cabinet of Ministers has not done what it should have to enforce the law and the governmental resolution that require neither budgetary funds nor complex organizational effort. Surprisingly, the Supreme Rada and Cabinet of Ministers have not controlled the implementation process, although recently the Speaker challenged the CPI data, suggesting that real inflation was twice as high as the official indicators suggested…

The UN agency responsible for the improvement of CPI calculation methodologies worldwide is the International Labour Organization (ILO). In late 2003, at the XVIIth ILO Conference on Labour Statistics, a new advanced resolution on consumer price indices was adopted. It is a substantial document of 28 pages. In practice, though, it is the IMF, rather than the ILO, that monitors national consumer price indices and publishes a monthly report - International Financial Statistics - summarizing financial and economic data on 170 countries of the world, including CPI (line 64), and providing a brief description of national methodologies for data generation.

As mentioned before, the IMF periodically sends its missions to various countries to conduct on-site monitoring and assist them with index computation. Ukraine has hosted at least five such missions. Of course, their assistance with improving multiple aspects of the national CPI calculations was invaluable and we should be grateful to them for that. However, it is natural that the IMF experts could not possibly understand all of the national “specifics” of the CPI calculation methodology and perennial problems remain unsolved. It could be a good idea, at least in the case of transition economies such as Ukraine, to invite national experts tо participate in the IMF missions.

International organizations attach great importance to CPI. Thus in 2004, six of them - ILO, IMF, OECD, EuroStat, UN and the World Bank - prepared and published a 535-page long volume called “Consumer Price Index. Theory and Practice”. During his visit to Ukraine in August 2005, IMF Managing Director Rodrigo de Rato met with President Yushchenko and, reportedly, discussed with him the IMF’s concerns over high inflation rates in Ukraine. It is noteworthy that almost all countries in the world, except Russia, Ukraine, a couple of other CIS member-states and China, measure their inflation rates based on the average annual indicators, rather than against those of December of the previous year. In Ukraine, when measured in these two different ways, the inflation indicators differ: CPI grew by 10.3% as compared to the December 2004 indicator but by 13.5% in the annual average terms, recognized by IMF. The difference is fairly substantial.

What are the main deficiencies in the methodology of our CPI calculations, and how do they affect the final results? There are four major issues here.

Only experts know that our CPI does not cover the rural population, i.e. one third of the country’s population. In fact, it should be defined as “CPI for the urban population”. The exclusion of the rural population from the assessment of such a significant social and economic indicator has very serious implications for home policy.

Nowhere in its official documents, including the “Consumer Price Index” yearbooks, does the State Statistics Committee openly admit that it takes no account of rural population in its computations. In justifying its current practice, the Committee maintains that collection of data on consumer prices in rural areas is a costly and labour-consuming exercise, which is true, of course. Yet the CPI does not aim to capture prices for certain goods and services expressed in absolute terms but, rather, major trends or dynamics in such prices. So the ILO recommends taking the rural population into account for the purposes of the national CPI calculation, emphasizing that absolute prices for some goods in rural areas could differ from those in towns and cities while the price trends in rural areas should be similar to the dynamics of urban price changes.

That is why all EU member-states incorporate data on rural populations into their national CPI. The only reason for not including those data in national CPI could be the absence of statistics on the structure of rural residents’ monetary consumer expenditures. Yet this is not the case in Ukraine, where data are collected simultaneously for the rural and urban population with the same degree of detail and according to the same procedure.

One should also remember that rural population purchases most goods and services in cities, which account for 88-90% of retail trade turnover and paid services. In order to ensure a more accurate measurement of price changes in rural areas, statistics authorities should organize an additional telephone survey of the rural population covering prices for five to seven key products, such as bread, flour, liquefied (bottle) gas and some others.

Given the identical, or almost identical, composition of monetary consumer expenditures of urban and rural population, exclusion of the latter from the current CPI calculations could have had no effect whatsoever on the accuracy of measurements had Ukraine been similar to other European countries. Yet it is different, in that most Ukrainian rural residents have homestead land where they grow various agricultural produce, thus having no need to buy those. As a result, the composition of their monetary consumer expenditures differs from that of their urban compatriots.

In 2004, rural residents spent more than urban ones for such foodstuffs as bread and bakery foods - by 1.8 times (29.1% and 16.6% of their monetary expenditures, correspondingly), vegetable oils - by 1.6 times (7.7% vs 4.8%), sugar and confectionery - by 1.7 times (13.5% vs 8.1%); and much less for other groups of foodstuffs, including: meat and meat foods - by 1.7 times, milk, cheese and eggs - by 2.2 times, potatoes - by 5 times, vegetables - by 1.6 times, fruit - by 2.1 times. Further, they spent 1.9 times less on eating out.

When prices for goods entailing higher expenditures for the rural residents than for the urban ones outgrow CPI, the overall index will be underestimated, and vice versa. This can be illustrated with the 2003 change in prices for bread and bakery foods, when general CPI (calculated against December 2002) increased by 8.2% and prices for foodstuffs rose by 10.9%, but prices for farinaceous foods grew incomparably: for bread - by 36.0%, macaroni - by 33.7%, flour - by 75.1%. Had the rural population, whose expenditures for these products were higher than those of urban residents, been included in the CPI calculations, the total index would have been higher than the officially published one.

Moreover, a greater part of bread and bakery foods in rural areas are made by individual entrepreneurs. They could hardly benefit from administrative measures aiming to stabilize flour prices and were hardly subject to price limits imposed in towns and cities for several sorts of bread. Therefore when the price of flour rose by 75%, bread prices in rural areas should have increased much more. Under the circumstances, the 2003 CPI should have been at least 0.3-0.5 percentage points higher than the official one.

Perennial underestimation of the general CPI is fuelled by totally incomprehensible calculations of individual price indices for potatoes, many vegetables and fruits.

In December 1999, the total weight of these products in CPI was 5.413%, or about one twentieth of all consumer expenditures. This indicator for potatoes was 2.462%. In subsequent years, this share must have reduced but it is next to impossible to obtain more recent data on such shares.

Annual data on prices for these foodstuffs are given in Table 1, according to which prices annually declined by 30-35%, on average, while potato prices almost halved every subsequent year. The only exception is 2005, when prices for ten quoted products rose by 6% on average, and potato prices fell by 9%. According to our CPI, over the last six years prices for fruit, vegetables and potatoes have not reduced by 8.5 times (to 11.8 %) as compared to the prices of December 1999, while the potato price has decreased 30-fold!

*Price of 1 kg of potatoes in December 1999

The above indices imply that whereas in December 1999, a kilo of potatoes cost about UAH 1.5, last December it shouldn’t have cost more than UAH 0.05, although on the very eve of 2000, a kilo of potatoes in Kyiv cost UAH 0.9. And the opposite is true: given the average potato price of UAH 2.2 per kilo in Kyiv markets in December 2005, six years ago it must have cost about UAH 65 per kilo, i.e. six times as much as pork sirloin. Put differently, the potato price index in the overall CPI for the last six years is underestimated by 50 times!

In reality, though, potato prices in Kyiv grew by 2.3 times in 2005, having reached the level of prices for rice and buckwheat. According to official statistics, in 2005 the total yield of potatoes was 19.5 million tons, only 6.2% less than in 2004.

Obviously incorrect computations of price indices in this group of products, characterized by clear seasonal fluctuations, weaken society’s trust in the official CPI. Table One also provides data on real December price for potatoes on 300 markets in 70 towns and cities of Ukraine. The data were available in the Trade Statistics Department of the State Statistics Committee, located in the same building as the Price Statistics Department responsible for CPI computations.

The observed discrepancies between the calculated CPI and actual prices for potatoes, fruit and vegetables arise from faulty methodology, whereby prices for agricultural produce of new harvest should be correlated with those of the previous harvest, due to which the overall CPI, particularly those for June and July, prove to be grossly underestimated. One could expect the statistics authorities to pay attention to the obvious defects in methodology and try to eliminate them. Instead, the authorities chose to drop potato price indices from their annual reports and “Consumer Price Indices” yearbooks...

It is easy to calculate that, even if potato and vegetable prices had remained at the 1999 level and the share of these prices in overall CPI has not changed over the last six years, the general CPI should have been higher anyway: in 2000 - by 2.3 percentage points (i.e. 128.1% instead of 125.8%), in 2001 - by 1.3, in 2002 - by 1.6, in 2003 - by 2.1 and in 2004 - by 2.2 percentage points. In 2005, it should have been higher by 0.9 percentage points (111.2%instead of 110.3%). Yet again, these calculations would not take into account the price changes over the last six years, while in fact those grew notably: for potatoes - by 1.7 times, for cabbages, onions, beetroots and carrots - by 2.1-2.6 times.

Correct CPI calculations largely depend on realistic weighting of each of 270 goods and services included in its composition. In Ukraine, data from sample surveys of household expenditures are used for this purpose, with no further adjustment or fine-tuning.

The national methodology and procedure for household surveys are amongst the world’s best, and the sampling meets all the appropriate criteria. However, sample survey findings have a serious drawback - incompleteness due to voluntary participation. Some households, predominantly wealthier ones, are traditionally left uncovered by the surveys, which eventually distorts the CPI composition where the weights of durable goods (cars, refrigerators, TV sets), petrol, as well as expenditures on recreation and entertainment, education and the like tend to be understated.

That is why weights of some goods and services in the overall CPI look paradoxically low. For example, in 2000, the proportion of expenditures on cars in the general CPI was 0.074%, whereas that of toilet paper and napkins was 0.077% (sic!); for petrol - 0.463%, whereas for onions - 0.509%; for colour TV sets - 0.157%, whereas for beetroots - 0.186%. This should be the reason why the State Statistics Committee would not provide data on a weighted CPI composition, broken down by all 270 goods and services, which is at variance with Articles 19 and 29 of the Law of Ukraine “On Information”.

The weighted CPI composition should be revised to reflect, more realistically, expenditures of the wealthier population. It is a challenge, but the national statistics authorities should undertake it, in cooperation with academic and research institutions, in order to bring their methodologies in accord with the internationally recognized ones. The IMF’s and other concerned international organizations’ assistance should be welcome.

Where the exclusion of data on rural population, dubious price indices for potatoes, fruit and vegetables lead to the underestimation of the general CPI, the distortion of its composition results in overestimations, albeit minor. Once amended, the CPI composition will show an increase in the weights of non-foods whose price rise indices have lately been lower than those of foodstuffs. For instance, in 2005 prices for foodstuffs grew 2.7 times more than for non-foods - by 10.7% and 4.0%, respectively.

Housing and municipal subsidies have a twofold impact on CPI.

A notable rise in rent and utilities tariffs (the last one took place in 2000, the next is expected after the parliamentary elections in March 2006) leads to a considerable overestimation of general CPI. These services make up a great part of consumer expenditures (over 11%) and are calculated, for CPI purposes, together with housing and municipal subsidies provided to numerous households. At the same time, the actual level of the increased tariff payment by the population is always lower. Therefore, if we want CPI to reflect the actual situation, its component relating to rent and utilities should be calculated exclusive of housing and municipal subsidies.

In 2000, when Viktor Yushchenko was Prime Minister, the price index for rent and utilities increased by 39.2%, while the real payment for these services by households grew only by 17.3%, as more than half of the tariff increase was covered from budgetary subsidies. Given this, the official 2000 CPI was overestimated by 2.2 percentage points: it should have been 123.6% instead of 125.8%.

In subsequent years, however, when there were no sharp increases in rent and utilities tariffs and the subsidies remained in force, coupled with a noticeable rise in nominal wages and salaries, the number of households eligible for subsidies decreased. Accordingly, the share of subsidies decreased while the share of the population’s expenditures for rent and utilities increased. In other words, these expenditures grew even with no rises in tariffs, and the overall CPI proved to be underestimated. As a result, in 2000-2005 the total amount of rent and utilities payments by the population increased by 2.13 times (given the decrease in Ukraine’s population by 2 million people); the number of households eligible for subsidies reduced by 3.4 times, and the total amount of subsidies fell by 5.4 times.

Unfortunately, economics-related ministries and governmental agencies, as well as leading NBU analysts, have to rely on the current CPI for their economic reporting and projections. Hopefully, the ZN publications will serve as eye-openers for Ukrainian officials alerting them to the controversial situation with the national CPI.

About 80 years ago, the State Planning Committee of the Ukrainian SSR - grandparent of today’s Ministry of the Economy - was more cautious and critical in dealing with the Ukrainian CPI calculations (made twice a month due to high inflation rates in the Ukraine of the mid-1920s). Referring to the then CPI in “Target Economic Indicators of the Ukrainian SSR for 1928-1929”, the State Planning Committee argued: “Our statistic analysts should pay special and urgent attention to improving the methodology and procedures for its computation by the end of this year”…