Construction statistics or constructing statistics?

Development of Russian construction statistics

Written by Andrey Vakulenko – MACON Realty Group, EECFA Russia

Assessing the development of construction industry on national scale is practically impossible without high quality statistical data that allow us to draw conclusions on industry trends and create any forecast model. The quality of Russian official statistics and its reliability have increasingly been becoming the subject of public discussion and the work quality of statistical service has been questioned by independent experts and economists. To overcome the problems, at end 2018, a comprehensive plan was developed for the reform and modernization of the Russian Federal State Statistics Service (Rosstat).

In 2018 the Russian economy seemed to have registered the highest growth over the past 6 years. According to Rosstat, GDP grew from 1.6% in 2017 to 2.3% in 2018; the highest value since 2012 (+3.7%). Such pronounced growth came as a surprise since all official and unofficial forecasts were much less optimistic: an average of 1.5%-1.8%. To a large extent, the successes of the Russian economy in 2018 derived from artificial manipulations, i.e. Rosstat’s review of the growth rate of the construction sector in 2017-2018. The indicator of the volume of construction works completed over 12 months has drastically changed: 2018 was to amount to RUB 8.4 trillion, 5.3% higher, or RUB 422 billion higher (at current prices) than in 2017. It was astonishing as previously Rosstat estimated construction works for 11 months of 2018 to post a modest growth of 0.5%. The 2018 growth in construction was a record for the last 10 years: it was only in 2008 when the sector grew at a higher rate (by 12.8% per year). On the contrary, between 2014 and 2017, construction industry saw a steady decline, which, according to official statistical calculations, gave way to a rather sharp increase in 2018. The final contribution of the construction sector to Russian GDP in 2018 was 0.3pp, although in 2017 it was previously negative (-0.1 pp). Such drastic changes caused a wide discussion for the following reasons:

• Weak argumentation for revising statistics. The Ministry of Economic Development and Rosstat recalculated construction data in late 2018 and early 2019 on grounds of clarifying previously submitted information by respondents at the end of the year. (This is due to the peculiarities of statistical accounting in construction in Russia: the peak of completions is at the very end of the year and then statistics are updated for a long time. Final data for the reporting year are published in spring and some figures may be adjusted retrospectively for a longer period). However, in this case, Rosstat adjusted the data by RUB 565bln, referring to only one project (Yamal LNG), which adds only RUB 241bln. The artificial increase in the indicator couldn’t be explained by only one project in one region, but Rosstat did not voice other official explanations.
• Growth of indirect construction indicators. Volume of completed construction works posted a massive rise against the backdrop of a decline in many industries related to construction, for example, in the production and transportation of building materials. In 2018, rail transportation of building materials for the year decreased by 6.8%, cement transportation also fell by 6.5%, cement production shrank by 2%, brick production dropped by 4.8%, and the construction of metal structures saw a 1.5% slump. Thus, according to Rosstat, production and transportation of building materials dipped, while more construction works were carried out. An important indicator here is also growth in the volume of housing completion, the most capacious segment of the Russian construction industry, which at end 2018 showed a steady decline by more than 4% (and by 6% in the multi-unit segment).
• Administrative reasons. In 2017, Rosstat, previously a fully independent agency under the Government, became subordinate to the Ministry of Economic Development. This created an internal conflict of interest since Rosstat data directly or indirectly indicate the effectiveness of the Ministry and the reliability of its forecasts.

EECFA’s Russia Construction Forecast Report with detailed analysis and forecast for the construction market of Russia can be ordered on http://eecfa.com/

Over the last year, official statistics was at the center of public discussion in the scientific community due to regular adjustments and revisions. And construction is not the only area of statistics affected by data manipulation, there are examples for other important macroeconomic indicators being revised:

• At end December 2018, Rosstat significantly improved data on Russia’s GDP growth rate in 2015-2017. The new estimate showed that in 2016 the economy expanded by 0.3% despite the previous drop of 0.2%. GDP growth in 2017 also turned out to be adjusted, although less: +1.6% instead of +1.5%. Decline in 2015 was also less than originally indicated: -2.3% instead of -2.8% (the first estimate by Rosstat was -3.7%). The recalculation was associated with obtaining newly revised data.
• In October 2018, public attention focused on published data on the real income of the population for January-June 2018, which, as per Rosstat, in the whole country rose by 2.4%. However, 6 out of 8 federal districts registered negative growth (from -1.6% to -0.4%), and the income growth of the population in the remaining 2 districts was +0.5% and +2.0%. The apparent contradiction in statistics was not explained in any way, and from early 2019, Rosstat switched to a new methodology for calculating population income and recalculated all data on this indicator from 2013. As a result, it turned out that in 2013-2018 real income decline was 8.3%, instead of 10.9% (previous estimate), and in 2018, the initial drop of 0.2% was replaced by a rise of 0.2%. Thus, growth rate of the real income indicator has been revised upward.
• Rosstat’s recent upward revision of industrial production data for 2016–2018 also raised many questions. Instead of stagnation in the industry in recent years, new statistics began to show moderate growth. For example, at end 2017, Rosstat estimated growth in industry at +1.0%, but after the revision at the level of +2.1%. Similarly, data for 2016 were revised upward. It was an interesting coincidence that Rosstat was fully in line with the forecast of the Ministry of Economic Development published even before the final results of 2017 became known.

In 2019, Rosstat conducted a radical revision of macroeconomic statistics since 2014. The losses of the economy from the “sanction war” and the slump in world oil prices were exaggerated and the economic recession was slight and short-lived. According to newly recalculated data, there was neither a long economic downturn, nor a big recession in industry and construction, and 2015 was the only crisis year.

Large-scale revisions by Rosstat, the wide range of indicators that they affect, their upbeat nature (indicators are only revised upward) and the often insufficient or unconvincing argument behind raise doubts in all who use these data. Refinement of statistics and revision itself is a normal practice taking place in any country, any revision though should have a clear and understandable explanation, and if such adjustments frequently occur, the question of the quality of applied methodology for collecting and analyzing statistical data arises.

Periodic revisions of statistical data in construction and other sectors of the economy are not the only difficulties. There are weaknesses not only in the statistical office itself, but also in the whole system of collecting and publishing statistical information in Russia such as:

• Delays in publishing data, irregular data revisions. Many macroeconomic and other major indicators can be revised even several years after their publication. Moreover, often up-to-date information appears with a long delay. The usual explanation is the insufficient quality of primary statistical information, namely monthly or quarterly statistical forms filled by enterprises. If respondents delay the submission of information, the accuracy of calculating statistical indicators is greatly reduced. For example, if in assessing the growth of industry, many companies did not submit information about themselves in a timely manner, Rosstat is forced to independently evaluate industrial production index, and these estimates can greatly differ from actual data. As the necessary amount of primary information is accumulated, Rosstat adjusts the data, and this can happen after a rather long period of time.
• Outdated format for publishing statistics. Rosstat publishes its data in press releases (“statistical publications”) on its website. These are separate pdf files or excel sheets, and statistics for many key indicators is a set of data files that are either not combined into one system within the site, or only partially are. Currently all statistical information available through the website, except for individual publications, is grouped under three information systems: The Central Database of Statistical Data (CBSD), the database of indicators of municipalities and the Unified Interdepartmental Information and Statistical System or (EMISS). Despite having been developed and implemented recently, these systems do not provide an application program interface (API), are not integrated and are inconvenient for end users.
• Inconsistency of data and of official statistics in various sources. As noted above, all official statistics in Russia are primarily published on Rosstat’s website as separate files. This information can be adjusted or refined, but is also published in press releases, and any such revisions lead to previous files being deleted, making it impossible to restore adjustments. Rosstat issues additional supporting material with explanations on some adjustments, but not always. Official statistics are published with a major delay in EMISS portal, but only after the publication of press releases. And it can happen that Rosstat website is adjusted, but EMISS is left intact, so in different publications and information systems different values of the same indicator might be given. And it is not clear for the user which number is correct (and why) as often no explanations are given.
• Confidentiality of certain data and lack of access to ministries’ statistics. Many indicators monitored on a regular basis and information collected by ministries are not combined with statistical systems. Ministries collect huge arrays of useful information but have no obligation to publicly disclose them. For instance, Rosstat does not have such an important indicator as the number of issued building permits. The Ministry of Construction maintains a specific database of issued building permits. This consists of projects that are being built with state participation, or projects that are significant nationwide. This information, however, is not transferred to official statistics and is not published in open sources. In addition, many indicators, their components and specific breakdowns observed by Rosstat, for some inexplicable reason, are not published in public domain. For example, a part of the statistical information necessary for the EECFA forecast model and for our regular reports is not published by Rosstat openly and is available only upon special official request.
• Frequent and weakly reasoned change in methodology. Many statistical revisions are explained to be the result of a change in methodology and in methods of calculating an indicator. Quite often these methodological changes are chaotic, inconsistent and are aimed at improving individual indicators. This is aggravated by the almost complete non-transparency of the methods used; usually general explanations are given, and detailed information or methods of specific calculations are not announced.
• The value of certain published information. Much of the collected and processed information is difficult to use in practice. For example, when publishing information on the age of housing stock, Rosstat uses 25-year intervals, having of practically no analytical value. At the same time, they have more detailed breakdowns, but for some reason do not publish them.

Although Russian statistics has several flaws creating a sense of chaos and disorder, negatively affecting its image and reducing overall confidence in state statistics, the situation is not critical. Rosstat seeks to adopt the best practices of foreign countries in statistics and is constantly improving. At end 2018, a comprehensive plan was developed for the reform and modernization of Rosstat to overcome the problems described above. Key areas of development of official statistics in Russia are as follows:

• no reporting in press releases and individual files, but publishing immediately digital data in information systems and bringing the reporting system to uniform standards;
• systemic revision and improvement of methodology for calculating the most problematic indicators where adjustments to historical data most often occur;
• use of data from alternative sources and big data to improve the accuracy of statistical information;
• increasing transparency, creating a clearer and more detailed explanation of calculation methods, as well as creating a platform to discuss and improve the current methodology by involving experts from different fields;
• easing the work with statistics and focusing on end users who are not only government agencies, but to a greater extent private businesses or citizens.

Positive shifts are already visible. In September 2019, the updated Rosstat website was launched. It has become significantly more user-friendly, while retaining its wide functionality. And Rosstat’s mobile application is also in preliminary testing mode.