In Russia in 2023, 18.3% of the employed were engaged in the informal sector. As in many other countries, the informal economy is important for the social and economic situation of the population.

This paper aims to identify the impact of specific key factors, such as unemployment rate, wages, poverty level, number of formal sector actors, proportion of the rural population, on the proportion of the informally employed.

The Two-Stage Least Squares Method estimates allow to conclude that wages and the number of formal sector actors had a negative effect on the proportion of the informally employed in Russia in 2009–2021, while the unemployment rate had a positive effect. The paper shows a lack of statistical significance of the impact of the proportion of the rural population on the proportion of the informally employed.

informal sector, formal sector, the employed, labour demand, labour supply, poverty level, unemployment rate, number of formal sector actors, wages, proportion of rural population

The literature reveals various aspects of informal employment. K. Hart (Hart 1973) pioneered research of the informal sector. (Gimpel’son and Zudina 2011) produced a definition and scope of informal employment. (Glinskaya 2018) specified the reasons encouraging people to engage in informal employment (economic, legal, motivational, administrative). Many studies (Dubravskaya 2020; Karpushkina and Voronina 2019) link informal employment to a low level of socio-economic development of the region concerned.

The regional levels of income and unemployment are widely considered to be the key factors contributing to a high share of informal employment (Bojko 2021). This means that informal employment directly correlates with the unemployment rate and inversely, with the level of income. (Nefedova 2019) holds a similar view. However, this point is not entirely clear and requires further verification.

This paper aims to identify the impact of the unemployment rate, wages, number of formal sector actors and the proportion of rural population on the proportion of informal employment by using econometric models.

There are two main approaches to defining the informal sector.

The production approach focuses on the activity pattern of informal sector actors (or production units). Their distinction from the formal actors is based on their legal form, i.e. whether an entity is registered as a legal entity. If not, it is considered to be part of the informal sector. Whether or not an entity is engaged in illegal activity is not a priority. The advantage of this approach is that it allows to keep statistical records of informal sector actors and employment on a country-wide scale (see Annex 1) (Barsukova 2004). This approach is in line with the provisions of the 15^th International Conference of Labour Statisticians¹. It is this approach that this paper follows.

Another is the legal approach that associates informal employment and the informal sector with a lack of business and tax accounting in place. This can be considered in the context of illegality, a lack of registration, non-payment of social security contributions (Gimpel’son and Zudina 2011). Due to the difficulty of data collection, this approach is only appropriate in narrowly focused studies, without the possibility of aggregating the data on the country-wide scale (Barsukova 2004). Weeks (1975), Leonova and Shushunina (2011), Belikova et al. (2022) support the legal approach to research on informal employment and the informal sector.

There are a number of microdata-based studies to be found in literature of the factors that affect informal employment in Russia (Lehmann and Zaiceva 2013; Lukiyanova 2015; Kim et al. 2019). (Nureev and Akhmadeev 2021) point to the dependence of informal employment primarily on the level of income of the population, rather than the condition of the formal sector. (Karpushkina and Voronina 2019) showed a significant negative impact of the unemployment rate on informal employment. The results of these studies need further verification on the panel data using an instrumental variable. A possible relationship between informal employment and the number of entities (actors) in the formal sector and the proportion of rural population has been insufficiently studied.

The novelty of this study is the testing of the causality between informal employment and the unemployment rate, wages, the number of formal sector actors, and the proportion of the rural population. The analysis aims at removing the endogeneity of the unemployment rate in the models by using the poverty rate as an instrumental variable. This technique is not to be found in the literature with respect to these indicators. Its practical significance lies in a more detailed consideration of the problem of endogeneity and the way it can be addressed with respect to informal employment.

A distinction is made between the formal and informal sectors of the economy in which the employed are engaged (Figure 1).

Figure 1.

Typology of the employed by their engagement in the formal vs. informal sector of the economy. Source: compiled by the authors based on (Methodological provisions on statistics… 2003). Note: For the purposes of this analysis, formal sector actors, enterprises and organisations are treated as a single concept.

Statistics on informal employment is based on the data from the national labour force survey conducted by Rosstat. The production approach to defining informal employment is used. The categories of individuals considered to be informally employed are listed in Figure 2.

Figure 2.

Structure of informal employment. Source: compiled by the authors based on Order of the Federal State Statistics Service No. 445 of 30 June 2017 “On approval of the basic methodological and organisational provisions for conducting a sample labour force survey”.

This study follows the interpretation of Rosstat², which defines informal sector workers as “individuals who, during the survey period, were employed in a minimum of one of informal production units, irrespective of their employment status or whether that was their primary or secondary job”³. Absence of registration as a legal entity is used as a criterion for defining informal production units.

Data on the proportions of informal employment (InfEmp_i,t) by region is a dependent variable in the panel regression analysis.

The regressors are as follows:

1. Unemployment rate (according to the ILO guidelines) – Unemp _i,t , %. A high unemployment rate could encourage individuals to take alternative informal jobs instead of those in the formal sector. However, a different situation is also possible. During a crisis with its unemployment, there is a lower demand for labour, such as housekeepers, nurses, drivers etc. In that case, the proportion of the informally employed could decline as the unemployment rate increases. The econometric modelling shows which of these effects is more pronounced in Russia.
2. Average monthly accrued wages of employees across entities in the entire economy in 2009 prices, Wage _i,t , rubles. This regressor takes into account the price effect in the form of the labour price. For example, individuals can find it attractive to work in organisations legal entities where wages are usually higher than in the informal economy (Uzyakova 2022).
3. Number of enterprises / organisations (according to state registration data) per 1,000 population, Org _i,t , units. This is the same as the number of formal sector actors, in contrast to the informal sector. A high value of this regressor would show a high employment potential, based on availability of many job vacancies, and could have a positive impact on individuals’ decision to seek employment in the formal sector. Conversely, a lack of jobs could force people into the informal sector (Sanghi et al. 2019).
4. Proportion of rural population in the total population as of 1 ^st January, AgriPopul _i,t , %. As a rule, this regressor correlates positively with the regional level of unemployment and poverty. However, the question remains open as to how a high proportion of rural population can influence the proportion of the informally employed. The effective demand for hired labour in the informal sector is higher in urban than in rural areas. However, cities have more legal entities, which expands the potential for formal employment.

This regressor may be endogenous due to bi-directional causality. It may not be the case that high unemployment leads to a lower proportion of informal employment. Rather, it is the opposite, that a low proportion of informal employment causes high unemployment.

Endogeneity can also be caused by omission of significant variables. Measurement errors can be another reason for endogeneity of the regressor.

To eliminate endogeneity due to bi-directional causality, a tool is needed that correlates with the unemployment rate (i.e., is relevant), but not with the proportion of the informally employed (i.e., is exogenous). We use the poverty rate as such a tool, i.e., a proportion of the population with monetary incomes below the poverty line (subsistence minimum) in the total population (Poverty_i,t, %).

The poverty rate should correlate positively to the unemployment rate. An increase in the latter, in turn, should lead to an increased proportion of the informally employed. Individuals who lose their jobs could subsequently move into the informal sector.

We apply the two-stage least squares method (2SLS).

First-Stage Regression. We estimate the regression of lnUnemp_i,t on the tool (lnPoverty_i,t) and on the logarithms of all exogenous variables to obtain the predicted values of the unemployment rate:

$\ln U nem{p}_{i,t}=a_0+a_1\ln Povert{y}_{i,t}+a_2\ln Wag{e}_{i,t}+a_3\ln Or{g}_{i,t}+a_4\ln AgriPopu{l}_{i,t}.$ (1)

Second-Stage Regression. We regress lnInfEmp_i,t on lnU nemp_i,tand the logarithms of all exogenous variables:

$\ln In fEm{p}_{i,t}={\stackrel{⏜}{b}}_0+{\stackrel{⏜}{b}}_1\ln U nem{p}_{i,t}+{\stackrel{⏜}{b}}_2\ln Wag{e}_{i,t}+{\stackrel{⏜}{b}}_3\ln Or{g}_{i,t}+{\stackrel{⏜}{b}}_4\ln AgriPopu{l}_{i,t}.$ (2)

There is a problem of potential endogeneity of the remaining regressors, but no appropriate tools have been found to address it.

Annex 2 explains the variables with their mean values and standard deviations.

More than 1,000 observations in 78 Russian federal constituents from 2009 through 2021 were used to model the factors⁴. Observations for 2022 and 2023 were not included in the regression analysis due to a lack of regional statistical data on individual regressors at the time of writing.

To compare results, modelling was also conducted on the panel data with fixed-effects estimators⁵.

According to Rosstat data, in Russia in 2023 the proportion of the informally employed was 18.63%, and the number of such workers was about 13.4 million. For comparison, in Germany in 2021, informal employment was at 3.8%, in France, 4.4%, in Italy, 11.2%, in Australia, 26.1%, in Turkey, 29.9%⁶. Although the number of the informally employed in Russia decreased in absolute terms in 2017, 2022, 2023, this was not an improvement in the labour market situation. The trend was related to an overall unfavourable economic situation in Russia, especially in 2022, due to sanctions imposed against Russia. In 2010 and 2017⁷, the number of the informally employed decreased as the country recovered from the economic crises of 2008-2009 and 2015-2016, correspondingly, with a higher demand from the formal economy, while in 2020, due to the impact of the COVID-19 pandemic (Kunitsyna and Dzhioev 2023). In our view, the number of people employed in the informal sector declined in 2022 due to the partial mobilisation, increased demand for labour in the formal sector and a significant reduction in migration growth.

Figure 3.

Number of the informally employed, by economic activity in Russia in 2009-2023, million. Source: compiled by the authors on the basis of data for 2009-2013 from the statistical compendia “Ekonomicheskaya aktivnost’ naseleniya Rossii” [Economic Activity of the Population of Russia] for 2010, 2012, 2014, as well as on the basis of data for 2014-2023 from the statistical compendia “Rabochaya sila, zanyatost’ i bezrabotitsa v Rossii” [Labour Force, Employment and Unemployment in Russia] for 2016, 2018, 2020, 2022, 2024 / Federal State Statistics Service. URL: https://rosstat.gov.ru/folder/210/document/13211 (accessed: 07 October 2024). Notes: 1. Before 2017, Rosstat conducted the survey for persons 15-72 years old, and as from January 2017, for persons aged 15 years and older. 2. Since 2017, only those who produce mainly for sale have been counted as the informally employed. Before 2017, the informally employed number included both those who produce mainly for sale and those who sell their surplus products* (*Labor Force, Employment and Unemployment in Russia (based on the results of Sample Labor Force Surveys) (2024) Stat. collection / Rosstat. Moscow, 2024. p. 92. URL: https://rosstat.gov.ru/storage/mediabank/Rab_sila_2024.pdf (accessed: 26 September 2024.)

In a crisis, the demand for goods and services other than essentials decreases. Usually, the informal sector is a source of non-essentials. As a result, the number of the informally employed decreases. In difficult economic circumstances, the formal sector features a higher resilience, as it is able to keep more jobs or, in extreme cases, to transfer workers to part-time or hidden employment.

Starting from 2018, the proportion of the informally employed in agriculture has stabilised in the range of 50-52%. In previous years, this figure was higher (Figure 4). In 2023, the proportion of the informally employed in agriculture, forestry, hunting, fishing and fish farming reached 51.4% and the number, 2.1 million.

Figure 4.

Proportion of the informally employed, by economic activity in Russia in 2009-2023, % of the total employed population by economic activity. Source: compiled and calculated by the authors based on the 2009-2013 data from the statistical compendia “Ekonomicheskaya aktivnost’ naseleniya Rossii” [Economic activity of the population of Russia] for 2010, 2012, 2014 and the 2014-2023 data from the statistical compendia “Rabochaya sila, zanyatost’ i bezrabotitsa v Rossii” [Labour force, employment and unemployment in Russia] / Federal State Statistics Service. URL: https://rosstat.gov.ru/folder/210/document/13211 (accessed: 07 October 2024). Notes: 1. Before 2017, Rosstat conducted the survey for persons 15-72 years old, as from January 2017, for persons aged 15 years and older. 2. Since 2017, only those who produce mainly for sale are counted as the informally employed. Before 2017, the informally employed number included both those who produce mainly for sale and those who sell their surplus products* (*Labor force, employment and unemployment in Russia (based on the results of Sample Labor Force Surveys) (2024) Stat. collection / Rosstat. Moscow, 2024. p. 92. URL: https://rosstat.gov.ru/storage/mediabank/Rab_sila_2024.pdf (accessed: 26 September 2024).

Due to restrictions during the COVID-19 pandemic, individuals were losing their jobs across the informal economy, which led to a decline in the proportion of the informally employed. The lower proportion of the informally employed in 2017 was caused by a higher demand for labour in the formal sector. Also to be considered is the change in the methodology used by Rosstat to account for the employed (see the notes to Figure 4).

In our view, the labour market will continue to be characterised by low unemployment and high demand for labour in the formal sector in the near future.

It is important to consider the problem of informal sector employment not only at the national level but also at the regional level.

The highest proportions of the informally employed in 2023 were observed in the constituents of the North Caucasus Federal District (Fig. 5). Agriculture is key for these regional economies. The dominance of small-scale farming is usually associated with a high proportion of the informally employed.

Figure 5.

Proportion of the informally employed aged 15 years and older by Russian regions in 2023, % of the total employed population. Source: compiled by the authors based on the 2023 data from the statistical compendium “Rabochaya sila, zanyatost’ i bezrabotitsa v Rossii” [Labour force, employment and unemployment in Russia] for 2024 / Federal State Statistics Service. URL: https://rosstat.gov.ru/folder/210/document/13211 (accessed: 29 September 2024). Note: The new territories of the Russian Federation [incorporated] after the 2022 referendum are indicated by hatching; they were not included in the analysis due to lack of data.

In 2023, the Arkhangelsk Oblast (18.3%), the Bryansk Oblast (18.5%), the Sakha Republic (18.8%) and the Novgorod Oblast (19.1%) were close to the national average (18.3%). The Udmurt Republic (17.2%), the Republic of Mordovia (17.5%) and the Tver Oblast (17.9%) were slightly below the Russian average.

The lowest proportions of the informally employed in Russia were observed in the Chukotka Autonomous Okrug (2.7%), Moscow (4.9%), St. Petersburg (6.2%), Murmansk Oblast (7.3%), Nenets Autonomous Okrug (5.4%), and Khanty-Mansi Autonomous Okrug – Yugra (9.0%). These are industrial regions where agriculture does not play a major role (Fig. 5).

Prior to econometric modelling, it is necessary to determine the correlation coefficients between the variables. A strong correlation is characteristic between the average monthly accrued wages and average monthly cash income of the population (r_corr. = 0,85) (Table 1). The Belsley-Kuh-Welsch test confirms the absence of multicollinearity between the variables, providing that the average monthly cash income of the population (Income) is excluded.

Notable correlation is inherent in the pairs of variables “Share of rural population – Number of enterprises and organisations” (r_corr. = -0.54); “Share of employed in the informal sector – Unemployment rate” (r_corr. = 0.53); “Share of rural population – Unemployment rate” (r_corr = 0.53); “Share of employed in the informal sector – Average monthly cash income per capita” (r_corr. = -0.51); “Share of employed in the informal sector – Number of enterprises and organisations” (r_corr. = -0.51) (Table 1).

Table 2 presents the coefficients of the regressors (both significant and insignificant) for the six models.

The fundamental equation of the model is as follows:

$$\begin{gathered} \ln In fEm{p}_{i,t}={\underset{(1.73)}{6.17}}^{***}+{\underset{(0.10)}{0.25}}^{\text{'}}\ln U nem{p_{i,t}}^{**}-{\underset{(0.16)}{0.37}}^{\text{'}}\ln Wag{e_{i,t}}^{***}-{\underset{(0.08)}{0.22}}^{\text{'}}\ln Or{g_{i,t}}^{***} \\ +{\underset{(0.14)}{0.22}}^{\text{'}}\ln AgriPopu{l}_{i,t}. \end{gathered}$$ (3)

Let us explain the choice of the 2SLS(1) model. The Hausman test confirms the appropriateness of the 2SLS(1) and 2SLS(3) models at the 5% significance level, as opposed to 2SLS(2). The 2SLS(1) model is preferable, with a higher corrected R² value (0.55) than in 2SLS(3) (Table 2).

In the 2SLS(1) model, the F-statistic for the weak tool test⁸ is 468.5, which is greater than 10. Thus, we can conclude that the tool is relevant.

The 2SLS estimate of the unemployment rate in the 2SLS(1) model was significantly larger in absolute value than the estimates in FE (2) and FE (3) models (Table 2). This implies that measurement errors were likely the predominant cause of the endogeneity of the unemployment rate. In the presence of measurement errors, the coefficient estimates for this regressor are biased towards zero. 2SLS eliminated this problem, and the coefficient estimate is larger. This suggests the superiority of 2SLS over the fixed-effects estimation models (FE (1), FE (2) and FE (3)).

Other things being equal, a 1% increase in the average monthly accrued wages of employees across organisations decreases the proportion of the informally employed by an average of 0.4 percentage points (p.p.). This relationship seems quite logical: an increase in wages makes working in organisations more attractive for both their existing employees and those informally employed.

An increase in the unemployment rate by 1 p.p. contributes to an increase in the proportion of the informally employed by an average of 0.3 p.p. An increase in the unemployment rate can indicate difficulty in finding a job in the formal sector, which can encourage people to take jobs in the informal sector instead.

The increase in the number of enterprises / organisations leads to a decrease in the proportion of the informally employed by 0.2 p.p. This speaks to the attractiveness of working in organisations legal entities, and some of the workers could leave the informal sector for formal jobs. However, there are barriers to finding employment in enterprises / organisations (due to lack of necessary qualifications and knowledge) (Ohnsorge and Yu 2022). As is shown in the section “Discussion of results”, the prevailing part of the informally employed have secondary general or basic general education. It can be concluded that the transition of applicants to organisations legal entities is possible providing they have appropriate knowledge and skills.

This paper evaluates a possible impact of socio-economic factors on the proportion of the informally employed in Russia in 2009-2021. The significant regressors, which indicate the statistical significance of the impact on the dependent variable, include average monthly wages and the number of enterprises / organisations (a negative impact), as well as the unemployment rate (a positive impact).

The results obtained seem quite logical and do not contradict the theoretical assumptions.

Higher wages in the formal sector make working in organisations more attractive. As a consequence, workers can become better anchored in the formal sector and informal workers can take jobs there too.

In general, an increase in the unemployment rate can indicate difficulties in finding formal employment, which may cause people to take informal jobs. Declining and low unemployment rates indicate a high demand for labour, substantially exceeding its supply.

Growth in the number of enterprises / organisations may come along with a decrease in the proportion of the informally employed. This can indicate the attractiveness of working in organisations legal entities, and some of the workers may leave the informal sector for a formal job.

Department of Economic and Social Affairs, Statistics Division. International Classification of Status in Employment 93 (ICSE-93). URL: https://unstats.un.org/unsd/classifications/Family/Detail/222 [Accessed on 12.06.2023].

Labor force, employment and unemployment in Russia (based on the results of Sample Labor Force Surveys)(2024) Stat. collection / Rosstat. Moscow, 2024.

Methodological provisions on statistics. 7. Methodological provisions for measuring employment in the informal sector of the economy / Federal State Statistics Service. URL: https://rosstat.gov.ru/bgd/free/b99_10/isswww.exe/stg/d030/i030150r.htm [Accessed on 26.09.2024].

Methodological provisions on statistics. Vol. 4 (2003) . M.: Goskomstat Rossii. URL: https://rosstat.gov.ru/folder/210/document/13250 [Accessed on 26.09.2024].

Order of the Federal State Statistics Service “On approval of the Basic methodological and organizational provisions for conducting a sample survey of the labor force” dated 30.06.2017 No. 445 (ed. on 03.10.2023). URL: https://legalacts.ru/doc/prikaz-rosstata-ot-30062017-n-445-ob-utverzhdenii-osnovnykh/ [Accessed on 28.10.2024].

Federal State Statistics Service. Results of the sample survey of the labor force. URL: https://rosstat.gov.ru/compendium/document/13265 [Accessed on 07.10.2024].

Federal State Statistics Service. Labor force, employment and unemployment in Russia. URL: https://rosstat.gov.ru/folder/210/document/13211 [Accessed on 26.09.2024].

Fifteenth International Conference of Labour Statisticians (ICLS). URL: https://www.ilo.org/public/libdoc/ilo/1993/93B09_65_engl.pdf [Accessed on 28.10.2024].

ILO Data Explorer: Proportion of informal employment in total employment by sex and sector (%). URL: https://ilostat.ilo.org/topics/informality/ [Accessed on 28.10.2024].

In-depth review of measuring the non-observed/informal economy (2022) Prepared by Mexico, International Monetary Fund and United Nations Economic Commission for Europe. Seventieth plenary session, Geneva, 20-22 June 2022, 17 pp. URL: https://unece.org/sites/default/files/2022-03/CES.2022.10_In%20depth%20review%20on%20measuring%20non-observed%20economy.pdf [Accessed on 10.06.2023].

OECD (2021) Informality and COVID-19 in Eurasia: The Sudden Loss of a Social Buffer / Paris: OECD Publishing. 69 pp. https://doi.org/10.1787/b27abb06-en

OECD, International Labour Organization, International Monetary Fund and International Statistical Committee of the Commonwealth of Independent States (2002) Measuring the Non-Observed Economy: A Handbook / OECD Publishing, Paris. 233 pp. https://doi.org/10.1787/9789264175358-en

Thus, the proportions of the informally employed in the regions are closely related, among other things, to their sectoral structure¹⁹. Where the latter is dominated by economic activities with stricter government regulation (including registration as legal entities), such regions have a lower proportion of the informally employed. Such types of economic activities include, in particular, mining, financial and insurance, health care and social services, education, information and communication, provision of electricity, gas or steam, air conditioning, and, in some degree, manufacturing, etc.

In Russia in 2023, 7.8 million were employed by individuals, individual entrepreneurs or farmers, or 58.2% of the total informally employed.

Sergey Victorovich Kiselev – Doctor of Economics, Professor. Faculty of Economics of Lomonosov Moscow State University. Moscow, 119991, Russia. E-mail: servikis@yandex.ru

Sanat Kairgalievich Seitov – Candidate of Economics, 2nd Category Engineer. Faculty of Economics of Lomonosov Moscow State University. Moscow, 119991, Russia. E-mail: sanat_95@inbox.ru

Valery Albertovich Samsonov – 2nd Category Engineer. Faculty of Economics of Lomonosov Moscow State University. Moscow, 119991, Russia. E-mail: v.a.samsonov@mail.ru

Ilya Valerievich Filimonov – Candidate of Economics, Assistant. Faculty of Economics of Lomonosov Moscow State University. Moscow, 119991, Russia. E-mail: filimonov.i.v@mail.ru