The growing use of digital devices and massive increase in information flows in the modern world brought about new sources of information concerning our everyday life. These sources, collectively known as Big Data, provide unique opportunities for researchers to analyze quantitative data on various social and economic developments. This paper provides a brief review of research and pilot projects, which have been carried out recently by national and international statistical organizations to analyze prospects of using Big Data sources in the official statistics. Results of surveyed projects indicate that use of Big Data in the official statistics is hindered by several serious impediments, such as unresolved questions concerning methodological approaches to analysis of new data sources, data access and protection of data confidentiality, technical requirements to new IT infrastructure. Therefore, the paper concludes that examples of statistical indicators calculated solely on the basis of Big Data sources are rare, and national statistical offices efforts are currently concentrated mainly on integration of Big Data sources with traditional sources of information.

Combining forecasts has already proven to be a simple and practical method for improving the quality of forecasting. The use of all available information from various prediction methods makes it possible to increase the accuracy of forecasting even when individual methods are not accurate enough.

This article is a continuation of previously published reviews of different approaches and methods for constructing weighting factors to combine forecasts. In the last several decades, in the foreign literature a lot of studies have been published on comparative analysis and establishing various methods of combining forecasts. Unfortunately, Russian authors pay little attention to this trend in improving forecasting methodology, e.g. they do not compare various methods of combining forecasts. Those studies that do make such comparisons are limited only to a few simple methods. Many methods of combining forecasts that are widely used abroad are disregarded by the Russian scientists. This paper along with the previous reviews have been tasked with examining different approaches to obtaining weighting factors in combining forecasts, and introducing them to Russian scientists and researchers.

As a result, this article adds to the previously proposed classification (Voprosy statistiki, 2015, № 8) of basic and most frequently used methods of combining forecasts with a description of the obtained results and indicating scientific publications authored by both Russian and foreign researchers on the issue in question, with those studies that were excluded from the authors’ previous reviews.

Monitoring Sustainable Development Goals (SDG), consensus on which was achieved until 2030 by countries led by the United Nations, is based on a set of global indicators. Their development required considering both the commonalities and the uniqueness of the participating countries in terms of information support of their development.

The international community is tasked with creating a multilevel «sustainable development statistics» that shall provide the necessary information platform for coordinating actions of all components of the national economic systems (governments, public organizations, businesses, population) with universal human priorities. In this respect, the author comments on the methodological requirements for ensuring coherence, consistency, complementarity at different levels of aggregation used by the United Nations as indicators for achieving the Sustainable Development Goals. The article identifies new tasks and formulates directions to complete them, using labour productivity dynamics as an example.

This article presents results of mathematical and statistical modeling research of sustainability of banks’ functioning that is relevant in light of banking system reforms. Basic concept for these constructions lies in the understanding of stability (or reliability) of business operation, financial state of which in normal circumstances ensures the fulfillment of all its obligations to the employees, other organizations and the State due to sufficient income and the matching of cost with revenue. Some of the most significant for the analysis of banks’ stability (reliability) indicators are available for all users (individuals as well as organizations) in the form of monthly financial reports submitted by the majority of banks, which are uploaded to the official site of the Bank of Russia and are duly and timely updated. In the authors’ opinion these indicators include some for which the Bank of Russia defines normative values. Although, monitoring the degree of consistency between actual indicators and regulations is far from being an ideal instrument to analyze bank reliability (i.e. to determine the situation that does not lead to license revocation), values of these indicators in the authors’ opinion are sufficient characteristics to establish financial health of a credit institution.

This article studies the nature of statistical relationship between trends of quantity values of the most vital aspects of banks’ activities and the probability of license revocation. The specific feature of models’ construction is the automatic choice of the functional entry form for bank characteristics. For this purpose, the authors use the generalized polynomials, which makes it possible to select model specification most comparable to data attributes.

The study was conducted on the basis of the public records of 887 banks for the period from 01.01.2013 to 01.12.2015. For this analysis, the authors used reports from already operating banks, in particular those that have just started operation, as well as banks that were liquidated in that period (some of the banks that operated during the research period later, in 2016 and 2017, also lost their licenses). The results of the model evaluation demonstrated a high degree of comparability of the entry form for bank characteristics with their economic substance and the Bank’s ofRussiaregulations. Quality comparison between the classical binary choice model for panel data and a model based on generalized polynomials shows a distinct advantage of the latter.

The article outlines methodological approaches to measuring the level of convergence in the Eurasian Economic Union (EEU) countries. The authors introduce the concept of conditional cyclical convergence of national economies as a convergence of short-term growth cycles in the overall macroeconomic dynamics. Decomposition of the cyclical macroeconomic dynamics in the countries is performed through identifying a long-term sustainable profile and short-term growth cycles. For this purpose, the double-pass through the Hodrick-Prescott statistical filter is applied for the original time series. Identified short-term growth cycles are visualized for each country using tracers of cyclic profiles. The conditional convergence of economies is determined on the basis of statistically significant cross-correlation coefficients.

The results of the calculations also make it possible to estimate the degree of synchronism in short-term growth cycles in the dynamics of the gross domestic product (GDP) index, the stable trajectories of the countries economic development, and draw a number of conclusions, in particular, to assess the slowdown in long-term macroeconomic growth in recent years and the growth volatility.

A distinctive feature of recent convergence in the economic integration is not the gaps reduction between the countries potentials, but the convergence of short-term growth cycles in the macroeconomic development. The growth slowdown in the Russian economy did not contribute to the positive prospects in the EEA countries. The strongest correlation in the analyzed period describing over 90% of the entire variation in time series was observed in the short-term GDP growth profiles inRussia,Kazakhstan, andBelarus.

Based on the calculations performed, the conclusions are drawn about the large-scale slowdown of long-term stable profiles in the dynamics of macroeconomic growth in all countries of the integration, the noticeable volatility of growth, almost simultaneously observed two years of recession with a clear predominance of major crisis events in 2015.

The growing use of digital devices and a massive increase in information flows in the modern world brought about new sources of information concerning our everyday life. These sources, collectively known as Big Data, provide unique opportunities for researchers to analyze quantitative data on various social and economic developments.

This paper provides a brief review of research and pilot projects, which have been carried out recently by national and international statistical organizations to analyze prospects of using Big Data sources in the official statistics. Results of surveyed projects indicate that use of Big Data in the official statistics is hindered by several serious impediments, such as unresolved questions concerning methodological approaches to new data collection and processing, data access and protection of data confidentiality, technical requirements to new IT infrastructure.

Therefore, the paper concludes that examples of statistical indicators calculated solely on the basis of Big Data sources are rare, and national statistical offices efforts are currently concentrated mainly on integration ofBig Data sources with traditional sources of information.

This article proposes methodological approaches to conducting qualitative evaluation (ranking) of customs authorities on collecting customs payments. The author points out that today reporting on such payments (in a form of absolute data and as series of basic relative indicators) is used to assess performance of customs authorities without regard to their specifics. This paper considers the idea of including rating algorithm, adjusting factor, based on the ratio of the indicators examined, namely, customs payments and the volume of the foreign trade that are registered by the customs services, or are related to territories in their respective jurisdictions.

The author put together a list of five indicators of the rating evaluation that served as basis for the integral indicator of performance of customs office in collecting customs payments. There are experimental calculations, according to which the regional customs authorities were ranked. It was revealed that all the major customs offices in the Volga Federal District (Samara, Tatarstan andNizhny Novgorod) were at the bottom of the rating, although they were leaders in customs payments and execution of plans.