Intangible assets and the efficiency of manufacturing firms in the age of digitalisation: the Russian case

Intellectual capital endowment becomes a fundamental prerequisite for technological advancements across countries and industries. Intangible assets (IA) have been considered a main source of productivity on an aggregate level during the last decades (Aghion & Howitt, 2006; Ramirez & Hachia, 2008; Chun & Nadiri, 2016; Montresor &Vezzani, 2013).

In developed economies, the marginal contribution of intangible capital to output growth already exceeds the physical one in high-tech production industries (Marrocu et al., 2012). According to Dal Borgo et al. (2013), manufacturing is among the sectors most heavily invested in intangible assets in the UK (manufacturing accounts for 51% of intangibles contribution to growth). In the French production sector, the growth in the share of intangibles also contributes to its enlargement in other industries (Delbecque et al., 2015). In Germany, the investment in intangible capital grew by 80–89% of the physical capital's level during 1995–2006 and half of the overall investment in intangibles accounted for manufacturing firms (Crass et al., 2014). Similar tendencies manifest in China, where sectors with a higher share of investment in intangible assets were the most productive between 1999 and 2007 (Fleisher et al., 2015).

Such upswing was largely driven by information and communication technologies (ICT), which clearly manifested in the U.S. where after 1995, the contribution of intangibles to the GDP growth was equal to that of physical assets (van Ark et al., 2008; Corrado et al., 2009; Nakamura, 2010). The IT revolution of 1994–2005 saw the most significant impact of intangibles on economic growth (Brynjolfsson et al., 2017). The famous Solow paradox, addressing the absence of the effect made by new technologies on productivity, has been widely discussed in the literature that offers a set of explanations and evidence (David, 1991; Brynjolfsson, 1993; Hatzius & Kris Dawsey, 2015).

Currently, economies undergo changes due to the new generation of ICT, induced by a drastic advancement in computing power (Furman & Seamans, 2018). Digitalisation is interpreted as an introduction or significant expansion of digital technologies in an organisation, a sector or the whole economy, leading to changes in business processes and significant socio-economic effects. This is expected to result in productivity gains (Tambe & Hitt, 2014; Dedrick et al., 2013; Aboal & Tacsir, 2018), structural changes (Bogliacino & Pianta, 2016; Rasel, 2017; Neirotti et al., 2018), new business models (Teece, 2018) as well as innovation intensification (Kleis et al., 2013; Sun & Li, 2017).

National governments encourage companies to adopt digital technologies and heavily support such initiatives. Russia represents a good example of such a policy. The national programme “Digital Economy of the Russian Federation” has been introduced in 2019 to secure the digital transformation in main sectors. Will these measures lead to gains? As intangibles become the core of the industrial process, it is important to consider its current role, patterns of influence on Russian enterprises, and industry-specific and idiosyncratic differences of the firms.

Due to the different nature compared to the physical capital, the IA impact on firms’ performance shows distinct mechanisms and channels, which are widely discussed in the literature. They may result in technological change, efficiency improvement, production factor reallocation or capital deepening (Bresnahan & Trajtenberg, 1995; Kumbhakar & Lovell, 2000; Chun & Ishaq, 2016; Nwaiwu et al., 2020). From the innovation side, they foster within and across industries spillovers (Bontempi & Mairesse, 2015; Thum-Thysen et al., 2017; Pieri et al., 2018) and provide an interplay across different types of intangibles.

A large strand of literature is dedicated to analysing different IA types on macro, sectoral and micro levels. On a country-scale, authors use a growth accounting framework to measure the contribution of intangibles to labour productivity, the total factor productivity and the economic growth calculation (Corrado et al., 2009; Fukao et al., 2009; van Ark et al., 2009; Borras & Edquist, 2013; Adarov & Stehrer, 2009; Apokin & Ipatova, 2017; Corrado et al., 2013; Thum-Thysen et al., 2017; Chen & Krumwiede, 2017; Rylková & Šebestová, 2019; Soltysova & Bednar, 2015).

Among BRICS economies, the Russian case is the least researched. Shahabadi et al. (2018) used the Solow residual (Solow, 1957) to estimate the impact of different types of intellectual capital on total factor productivity (TFP) in emerging economies, including Russia, and conclude that this group of countries acquired existing new technologies rather than developed them.

Based on the existing macro estimates for Russia (Voskoboynikov et al., 2020), the effects of intangibles are not as large in comparison with developed countries. ICT, as the main asset in the age of digitalisation, contributes little to the TFP growth of in comparison to other physical elements (machines, buildings, etc.). On the other hand, ICT-growth rate in 2002–2007 was the largest in manufacturing as in most dynamic finance and services (Voskoboynikov et al., 2020). Arguably, the capacity of ICT and other intangibles in Russian manufacturing was not fully exploited opposite to developed economies, did not achieve a threshold and might serve as a productivity driver in the next decades.

Fig. 1

Only several studies exist regarding the role of intangibles on Russian microdata, which support evidence in the macrolevel. According to Shakina et al. (2014), a gap in intangibles is responsible for more than 26% of the variation in the gap in the economic value-added of Russian companies. Overall, the performance heterogeneity has a different scale of intangible capital in the firms (Molodchik et al., 2019). Several papers consider particular intangible assets and get similar patterns with other countries, particularly related to R&D and its link with technological change (Apokin & Ipatova, 2017; Pieri et al., 2018). Russian authors focus on company strategies for using IA (Shakina et al., 2016) and the taxonomy of firms based on this (Podmetina et al., 2011; Paklina et al., 2017), and addressing research and development in detail (Dezhina & Ponomarev, 2014; Simachev & Kuzyk, 2014; Gershman et al., 2018; Simachev & Kuzyk, 2019; Zemtsov et al., 2019).

According to previous results, efficiency was the main component of TFP, which has been affecting the productivity of industries since the end of 2000 (Ipatova, 2015) compared to the economic boom of 1998–2007 with a predominant role of the technological progress channel (Brock & Oglobin, 2018). Papers that refer to intangibles as an efficiency determinant of the level of firms are scarce. To close this gap, considering the evidence from the academic and empirical literature, this paper applied the stochastic frontier model (SFM) on the panel data for 2009–2018 of more than 300 public Russian companies from manufacturing industries. The following hypotheses were formulated: 1) intangibles positively affect technical efficiency and this effect increases over time; 2) IA with time become a major source of efficiency; 3) intangibles are more important for high-tech industries; and 4) its effects are reduced due to the crisis in 2014.

The paper is organised as follows. The first section briefly reviews the theoretical background of the paper. The second section reports on the empirical background. The third section describes the data of the study. The fourth section represents and discusses the results. The final fifth section gives policy implications and makes concluding remarks.

This part overviews extant papers on IA and their relationship with productivity using microdata. This data-level enables using a broader set of intangibles concepts, variables and estimation techniques (Roth, 2019). To start, the outline of used IA definitions and that of the current study are given.

In financial accounting, an intangible asset is an object without a physical form that can bring economic benefits, when used in production activities for a long time. The academic literature considers intangible assets more broadly. They have key features, such as high value, a rarity for an organisation, complexity for imitation or substitution (Bontempi, 2016; Paklina et al., 2017). A comprehensive IA framework was proposed by Corrado et al. (2005) and is most often used in comparisons of countries and cross-countries. It includes research and development results, computerised information (software and databases), and economic competencies (particular characteristics of an individual firm, including personnel, trade names, etc.). An analysis is often stipulated by the availability and consistency of the data.

Most of the empirical results indicate a strong impact of IA on the efficiency of industrial companies regardless of country affiliation (Marrocu et al., 2012; Dal Borgo et al., 2013; Corrado et al., 2013; Goldar & Parida, 2017; Piekkola, 2020). Based on data of 1523 industrial enterprises located in key Chinese cities, Yang et al. (2018) found a significant positive relationship between the IA types (software, research and development, and organisational investment), and the performance of firms demonstrated differences in the relative importance of particular IA types compared with developed economies.

However, high investments in IA do not always lead to productivity growth: if a certain threshold is exceeded, further investments fail to generate positive effects. This relationship was found in public companies in Japan in 1991–2001 (Ramirez & Hachia, 2008). Companies from the industries of non-ferrous metallurgy and transport and telecommunications that reduced the volume of research and development, expressed in terms of capital stock, showed higher productivity. Two explanations are feasible: the IA distinct nature and lags to fully deploy and achieve effects. In general, IA act as a main source of productivity regardless of features particular to sectors and firms.

Long lags may lead to a negative impact on efficiency and productivity in the short term. Chappell & Jaffe (2018) found that IA investments lead to a decrease TFP caused by the time lag and cost growth of its implementation. Basu & Fernald (2007) obtained similar results when modelling the impact of ICT on industry productivity in the United States for 1987–2004. Short term investments may diminish TFP, as it needs time and resources for reorganisation and training. This lag can extend from 5 to 15 years. It also takes time to gain experience with a new production process. Over the long term, intangibles become particularly important for firms with initially low levels of productivity due to the catch-up effect (Heshmati et al., 1995; Castiglione & Infante, 2014).

The propensity to invest and the volume of investments in IA depend on internal characteristics of a particular firm, such as age and size, sector type and others (Marrocu et al. 2012; Goldar & Parida, 2017; Chappell & Jaffe, 2018; Yang et al., 2018). However, productivity is also strongly affected by external shocks. During these periods, even in the absence of significant changes in company strategies regarding IA, the rate of productivity growth may decrease (Tang & Wang, 2020).

Recent papers increasingly focus on the combination between different types of IA and its impact on performance through company innovation (Ramirez & Hachia, 2008; Kleis et al., 2012; Gómez & Vargas, 2012; Chun & Ishaq, 2016). Again, there is variation in the results. Montresor & Vezzani (2016) showed that IA is more important for the industry than internal research and development, which in turn is more important for the service sector. On the contrary, Ramirez & Hachia (2008) argued for a higher significance of internal R&D in manufacturing.

Thus, intangibles serve as an innovation factor (Hall et al., 2013), production factor (Corrado et al., 2009) or both (Pieri et al., 2018). Different types of IA may act as the first or the second category. Research is more important for innovation, while ICT — for productivity and efficiency (Hall et al., 2013); however, the former serves as a prerequisite of the latter two. Several papers also confirmed the role of R&D for efficiency gains (Ramirez & Hachia, 2008; Añón Higón et al., 2017; Shahiduzzaman et al., 2017). New waves of the literature suggest intangible assets contribute to service innovation in light of servitisation (Cheng & Krumwiede, 2017; Kozłowska, 2020) and business model transformation. Heterogeneity of results is often explained by individual characteristics of a firm (age, size, historical base of intangible assets, financial status, ownership, technology intensity, export status, and trade issues).

Based on the brief analysis, the IA assessment findings are rather diverse and depend on a large set of characteristics. This paper represents the first step to a wide analysis of IA features and trends in emerging countries on the example of Russian production companies.

This paper uses data from the Ruslana database for 340 public companies belonging to the economic activities listed under codes 10–33 OKVED2 (synchronised with the NACE classification). The time span covers 2009–2018 and includes 3310 observations. This category of companies makes a crucial contribution to productivity and overall investment. Previously, Paklina et al. (2017) also studied listed companies and assessed their strategic choices regarding intellectual capital.

Output as a dependent variable is presented by the operating revenue of companies. The number of employees (l) is measured in persons, while other explanatory variables, including fixed assets (fa), other assets (asset), intangible assets (ita), in thousands of Russian roubles. All monetary variables are nominated in constant prices of 2009 and deflated using the GDP index-deflators, which are calculated by the national statistical office and available on the website of the statistical office.

The main limitation of the study is the absence of data on ICT-capital at the level of firms. The afore-mentioned database contains data on R&D capital, but with numerous omissions that hamper estimation. To assess the intellectual capital of companies as a whole, the aggregate indicator of intangible assets is presented in the annual financial statements of the firms. IA has been previously shown as an adequate variable in the stochastic frontier exercises for Russian firms (Ayvazyan et al., 2012). According to the national accounting system, intangible assets comprise patent and other intellectual property on inventions, licenses on software and databases, trade names, as well as goodwill (KPMG, 2012).

As cited earlier, the Cobb-Douglas specification with logged values is chosen for Russian data. It is expressed as follows: (2)lnyit=β0+β1t+β2lnf ait+β3ln assetit++β4ln itait+β5lnlit+vit−uit\matrix{{\ln {y_{it}}} \hfill & { = {\beta _0} + {\beta _1}t + {\beta _2}\ln f\,{a_{it}} + {\beta _3}\,{\ln}\,asse{t_{it}} + } \hfill \cr {} \hfill & { + {\beta _4}\,{\ln}\,it{a_{it}} + {\beta _5}\,{\ln}\,{l_{it}} + {v_{it}} - {u_{it}}} \hfill \cr } where ln y_it — turnover of the company i in period t, t — years, ln f a_it — fixed assets as proxy for capital, ln ita_it — intangible assets, ln l_it — number of employees, v_it — stochastic noise, u_it — technical inefficiency. To include fixed capital that is rented by a company, and based on Ipatova & Peresetskiy (2013) and Shchetynin (2017), other assets (ln asset_it) are inserted in the model.

In this class of stochastic models, technical efficiency changes under determinants specified in the heteroskedasticity equation (Pieri et al., 2018). There are two determinants in the current study: intangible assets and the time trend. Intangibles also contribute to TFP due to the accumulation with time and technical change, which is embedded in the time trend t in the production frontier equation. The heteroscedasticity equation is defined as: (3)log(σuit)=δ0+δ1ln itait+δ2t\log \left( {\sigma _{uit}} \right) = {\delta _0} + {\delta _1}\,{\ln}\,it{a_{it}} + {\delta _2}t where δ_n — estimated coefficients of technical efficiency determinants.

Following Pieri et al. (2018), it is assumed that TFP is influenced by the trend (t), intangible assets (ln ita_it), its evolution in time (ρ₂t × ln ita_it) and technical inefficiency (u_it): (4)tfpit=ρ0t+ρ1ln itait+ρ2t×ln itait−uittf{p_{it}} = {\rho _0}t + {\rho _1}\,{\ln}\,it{a_{it}} + {\rho _2}t \times \,{\ln}\,it{a_{it}} - {u_{it}} where ρ_n — estimated coefficients of TFP determinants.

Along with this, the time trend is usually interpreted as TFP in more common models for panel analysis (the time-variant model). To compare the trend attitude, two other specifications are applied: time-variant and time-invariant models. The first one estimates technical efficiency for each year separately. Such a model focuses on persistent inefficiency and does not require distributional assumptions (Kumbhakar et al., 2014). It has the following form: (5)ln yit=β0+β1ln fait+β2ln assetit++β3 ln itait+ln4 lit+d yearit+vi−ui\matrix{{\ln \,{y_{it}} = {\beta _0} + {\beta _1}\ln \,f\,{a_{it}} + {\beta _2}\,{\ln}\,asse{t_{it}} + } \hfill \cr { + {\beta _3}\,\,{\ln}\,it{a_{it}} + {{ln}_4}\,{l_{it}} + d\,yea{r_{it}} + {v_i} - {u_i}} \hfill \cr } where d year_ti means dummy variables for each of ten years, other variables are the same as stated earlier. In contrast to the specification indicated in equation (2), the time-variant decay model assumes v_i and u_i to be independent. It also implies that there is a trend in inefficiency error, which is estimated in the following way (Kumbhakar & Lovell, 2000): (6)uit=(−exp(−η(t−Ti)))ui{u_{it}} = \left( { - \exp \left( { - \eta \left( {t - {T_i}} \right)} \right)} \right){u_i} where Ti — the last year in the panel, η — the decay parameter, errors v_it are independent and identically distributed with zero mean and constant variance (vit∼N(0,σv2)\left( {{v_{it}} \sim N\left( {0,\sigma _v^2} \right.} \right) , u_it is the base-level inefficiency (the level of inefficiency for firm i in the last period T_i) that follows truncated normal distribution (ui∼N+(μ, σu2))\left( {{u_i} \sim {N^ + }\left( {{\rm{\mu }},\,\sigma _u^2} \right)} \right) , v_it and u_i are distributed independently. This helps to distinguish different patterns in trend and further discuss its possible reasons.

How to stimulate firms to invest in intangibles? Company incentives to invest in transformation and implement related complementary changes are largely affected by the policy to promote technology adoption, and this trend is stable across developed and emerging economies (Teece, 2018). Despite differences in the scope and direction of policies, almost all governments offer such support. It is not surprising that the industry receives attention, especially in times of crisis, when the modernisation of production becomes a factor of survival (Shakina & Barajas, 2016; Polder et al., 2018).

This statement is supported by the recent global crisis in 2008–2009. A trend was observed in developed economies to establish specialised institutions for the development and dissemination of advanced manufacturing technologies. Such initiatives were launched in the United States (sine 2011, the programme “Manufacturing USA”), the United Kingdom (since 2013, catapult centres), Australia (“Industry 4.0 Testlabs for Australia”), Canada (the Advanced Manufacturing Supercluster), Japan (Smart Manufacturing — Smart Monozukuri initiative), South Korea (manufacturing innovation centres 3.0) (GOV.UK, 2017; Australia Prime Minister's Industry 4.0 Taskforce, 2017; METI, 2017; Next Generation Manufacturing Canada, 2018; GAO, 2019). Most of these initiatives promote an advanced class of technologies, including computer modelling, new material development, production systems etc. that are expensive and need business restructuring (Nazarko, 2017).

Current trends in sectoral technological development are induced by the next wave of information and communication technologies (ITU, 2017; Brynjolfsson et al., 2017). Though the channels of technology dissemination and influence on production performance are common, digitalisation varies due to differences in countries and characteristics of firms. In emerging economies, productivity is frequently driven by the acquired and imported technologies, embodied in machinery and software. In general, they serve as a leading mechanism to promote innovation activities compared to domestic R&D in developed countries (Shahabadi et al., 2018). Recent studies show that the growth of ICT plays a key role in the TFP increase in emerging economies due to larger investments compared to other intellectual assets and primarily R&D (Shahabadi et al., 2018). In search of new innovation sources, digitalisation may play a role as a factor for production efficiency and the development of new products (Paklina et al., 2017).

To find appropriate triggers in sectors, a range of new policy mechanisms arises with the implementation of traditional ones. They contribute to narrowing the digital gap across and within sectors (Spiezia, 2011; Polder et al., 2018). The set of new tools comprises “living labs” (e.g., for driverless cars in Germany), testbeds (for blockchain technologies in the Republic of Korea) or platforms for joint research. Regulatory sandboxes is a relatively new tool that plays a particular role in industry absorbing new solutions. For example, special regimes help to test unmanned aerial vehicles in the US, or unmanned road vehicles in Germany (Federal Aviation Administration, 2018; BMWi, 2020). Many such initiatives address SMEs, including technology transfer, assistance with finding partners, and financial support (BMWi, 2019). Specialised platforms for small firms from different sectors provide an opportunity to choose an appropriate financial tool and receive professional consultation on digitalisation (France NUM, 2020).

Instruments aimed at promoting the demand for digital technologies also differ. Flexible fiscal mechanisms are applied to promote the mass adoption of technologies among companies. They cover a wide range of economic agents and include an accelerated depreciation or tax credits for investments in information technologies etc. Along with soft loans for buying digital products and services, various vouchers were actively used to support SMEs, including those focused on innovation (European Commission, 2018). Standardisation and certification is another area of interest to support the technology dissemination on a massive scale. Along with it, the entrepreneurship infrastructure, methodological recommendations for digital transformation, market regulation and other existing tools represent a large area for policymakers (OECD, 2017).

Aiming to maximise efforts of different decisions, requires them to be targeted in terms of sectoral problems and features, including efficiency and productivity issues. This is especially critical in the current times marked by the coronavirus crisis and challenges faced by countries. The national programme “Digital Economy of the Russian Federation” goes in line with the foreign initiatives and provides many mechanisms to support the adoption and use of digital technologies. The initiatives resemble those in other countries, where manufacturing is among the priority industries.

Results of the previous section suggest that intangibles do play an important role in decreasing the sectoral technical inefficiency. It is expected that intangibles in the short and medium run will secure an efficient production process, and later, it will contribute to the channel of innovation via the technological shift. Based on this reasoning, an analytical framework to choose policy instruments is introduced (Table 4).

Policy initiatives to promote digitalisation should be different due to sectoral R&D intensity and strategies to adopt digital technology, i.e., to develop or acquire. From this perspective, companies may introduce existing or new technologies. Several sets of policy tools can be distinguished. The number of R&D support tools is limited because of a risky nature; however, they imply the most extensive effects in terms of efficiency and technical change. Low-tech firms that provide R&D obtain the same results, but they are less significant than high-tech. Again, a less intensive impact resulted from the acquisition of technological assets in low-tech firms. On average, it results in more considerable efficiency gains and is supported by instruments (tax incentives, preferential loans, transfer centres, etc.) that are expected to spread technologies in a large group of companies.

It is assumed that digital asset accumulation correlates with innovation capacity (Hall et al., 2013; Borgo et al., 2013; Añón Higón et al., 2017; Ejdys, 2020). Since then, an average firm has two alternatives: to develop a solution in-house or in cooperation with partners, including universities and scientific organisations. It may also choose acquisition from an external supplier. The previous studies found that high-tech firms more often adopting customised technological solutions or developing them in cooperation with external suppliers (Pieri et al., 2018). The opposite is true for low-tech firms, which frequently implement existing technologies. To support the development of new intangible assets, decision-makers may opt for more risky measures, such as venture capital, testbeds or regulatory sandboxes. The development of new technologies should reconcile with the elaboration of standards that offer new technical rules. When it comes to new solutions, the number of organisations is often limited, and in this case, grants or subsidies may be the most efficient way to stimulate innovation. As more firms get engaged in R&D, measures having a wider coverage are required, such as tax incentives (e.g., income tax relief for R&D activities). For all categories, regulation plays a central role as an enabler of legal conditions for technology adoption and use.

The acquisition has a relatively lower impact on productivity but may still result in the frontier shift. Policy instruments are less risky and aim to involve a larger number of firms. In the case of the purchase of new assets, a business often needs guidelines, frameworks and general information on new technological issues. Both for high-tech and low-tech groups, a similar set of instruments may be applied. The initial idea of such support is to smooth differences in technological capacities and stimulate within and across industry spillovers.

The reasoning presented in this section represents a starting point to a further, more detailed investigation of types of intangibles and the scale, to which they affect manufacturing companies in Russia. Here, only some general vision is developed. Such an approach enables better planning and assessment of technological development in organisations (Bieńkowska, 2020; Nazarko et al., 2020). Next research may address the empirical estimation of different types of intangibles and the impact associated with this policy instrument on the propensity of companies to accumulate intangibles. This may bring useful insights into fostering investments in intangibles and, particularly, digital technologies.

The role of intangibles in the digital economy is growing rapidly in emerging and developed countries. In Russia, they have not yielded productivity gains despite the rapid upsurge of the IT industry (ISSEK HSE, 2020). Several structural features may be responsible for such a situation. Studies in the area on a company level are scarce; however, they may shed light on some reasons for the current low impact of intangibles on productivity and growth.

The current paper enlarges the extant empirical literature by revealing the role of intangibles in emerging economies. In particular, it contributes to the strand of productivity analyses and efficiency as its key component, including patterns and its development over time. It also accounts for differences in sectors due to research intensity and gives special attention to the crisis period. This may be of significant interest in the discussion on post-pandemic economic development and appropriate tools for it.

Focusing on listed companies from the manufacturing sector, the stochastic frontier model is applied to estimate the role of aggregate intangible assets as a determinant of technical efficiency. Its role as a production driver is still modest due to low investments and the level of accumulation (Shakina & Molodchik, 2014; Shakina et al., 2016; Paklina et al., 2017). Firms from high-tech sectors enjoy more extensive effects of intangibles on inefficiency decrease. After 2014, this effect was lower than before.

The consequences of the crisis were significant for all groups and widened the gap within and across the high-tech and low-tech firms. Consequently, a small subgroup of most efficient units improved its level, while others worsened their position. The dynamics of the indicator in low-tech firms reflect the increase in inefficiency due to higher dispersion.

Such disproportions have a structural nature and should be addressed with appropriate policy tools. To secure systemic investments in intangibles and digital technologies as its major component, national governments have adopted sets of measures, especially during the last years (OECD, 2019). Sectoral and time features of firms, as well as the actual endowment with intellectual capital, should be considered while designing policies.

The paper offers an analytical framework to select relevant policy tools to foster corporate investments in the development or acquisition of intangible assets. In-house research implies targeted measures, while in the case of acquisition, instruments with large business coverage are required. Both types are important to accumulate the domestic intellectual endowment and on the other hand, to adopt existing frontier technological solutions.

This approach is reasonable to consider since the share of Russian organisations engaged in technological innovation remains low. To achieve a large scale of technology adoption, small and medium companies should largely implement and use different digital solutions in technological, organisational and other domains, and restructure all business processes. Along with the problems of underinvestment in innovation, firms do not fully see the advantages of digital technologies. It is important to provide communication and financial tools to scale up domestic technologies and contribute to their dissemination across industries.

The current study has several limitations. First, due to the lack of data, only the general aggregate effects of intangibles were considered. Second, it does not account for other determinants, which are captured in the time trend. The further elaboration on these problems represents a large area for investigation in the field of productivity analysis in Russian firms.