Infrastructure System Obstacles and Technology Adoption by Firms in Transition Countries

John E. Anderson; Muazzam Toshmatova

doi:10.1515/soeu-2023-0005

Article Open Access

Infrastructure System Obstacles and Technology Adoption by Firms in Transition Countries

John E. Anderson
John E. Anderson is the Baird Family Professor of Economics and the Executive Director of the Central Plains Federal Statistical Research Data Center at the University of Nebraska-Lincoln. His fields of expertise include public finance, urban economics, and transition economics. He has served as an advisor at the President’s Council of Economic Advisers in the United States as well as an advisor to governments in Eastern Europe, the Balkans, and Central Asia.
and Muazzam Toshmatova
Muazzam Toshmatova is a Postdoctoral Associate at the Heldrich Center for Workforce Development. She holds an M.S. degree in economics from Texas A&M University and her Ph.D. in economics from the University of Nebraska-Lincoln. Her research is focused on labor, public, regional, and transitional economics.

Published/Copyright: March 14, 2024

Published by

Become an author with De Gruyter Brill

Submit Manuscript Author Information

From the journal Comparative Southeast European Studies Volume 72 Issue 1

Abstract

This study uses business survey data to analyze firm perceptions of infrastructure systems in transition countries as obstacles to conducting business. Factors that contribute to reported perceptions are modeled, including country macroeconomic conditions, firm characteristics, and measures of overall reforms and infrastructure system reforms. Identified obstacles arise from binding constraints that limit economic growth, reflect implicit shadow costs, and signal areas in need of policy reform or infrastructure investment. The analysis covers electric power, telecommunications, transportation, and water and wastewater infrastructure systems. Also examined is whether transition indicators of reforms, intended to reflect improvements, are associated with perceptions of obstacles. Beyond identifying infrastructure obstacles and the effects of reforms during the transition process, we also explore firm adoptions of new electronic technologies, essentially a new infrastructure system, as they use email, websites, cell phones, and high-speed Internet in their business activity.

Keywords: infrastructure; electricity; water and wastewater; telecommunications; business obstacles

Introduction

This study uses business survey data to determine the extent to which infrastructure systems in transition countries are perceived to be obstacles to business enterprises. We then model the factors that contribute to those reported perceptions, including country macroeconomic conditions, firm characteristics, and measures of overall reforms and infrastructure system reforms. Those obstacles arise from binding constraints that limit economic growth. They reflect implicit shadow costs and signal areas in need of policy reform or infrastructure investment.

We focus on how firms view their experience with the electric power, telecommunications, transportation, and water and wastewater infrastructure systems of their countries as obstacles to doing business. Results of the analysis indicate that there can be infrastructure bottlenecks in richer or faster developing economies that contribute to firms reporting more obstacles. Firm size, competition with informal market firms, and international credentials also have negative impacts on perceived obstacles.

We also examine whether transition indicators of reforms, intended to reflect improvements, are associated with perceptions of obstacles. Our analysis reveals that those indicators are often associated with more reported obstacles, not fewer. Those reform indicators are often reflective of liberalization and rationalization of tariff rates, leading to higher costs for firms, which is ultimately perceived as impeding business by firms in transition countries.

Beyond identifying infrastructure obstacles and the effects of reforms during the transition process, we go on to explore firm adoptions of new electronic technologies in their business activities, using email, websites, cell phones, and high-speed Internet. We find that, in richer or faster developing countries, bottlenecks arise due to inadequacies of electric, telecommunications, and transportation systems. Access to connections, delays in getting service, and inadequate capacity in legacy systems are all potential limitations. Firms in these countries are more likely to adopt the use of email, websites, high-speed Internet, and cell phones in their operations.

This research contributes to the literature first by confirming what perceived obstacles infrastructure systems present for firms. Second, the analysis provides insight into how the transition indicators, meant to measure progress in the transition process, may in fact reflect additional obstacles for firms. Finally, the analysis addresses the adoption of new electronic technologies, essentially a new infrastructure system essential for doing business in the transition process. In all three of these ways, new insights are provided that can motivate policy changes and/or infrastructure investments needed to support the success of firms in transition countries.

We make use of both cross-sectional and panel survey data collected by the European Bank for Reconstruction and Development (EBRD) to estimate firm-specific effects, country-specific effects, and time trends. The EBRD Business Environment and Enterprise Performance Survey (BEEPS) data are used in the estimation of reported business obstacles. The cross-sectional 2012–2016 (BEEPS V) data are used along with the 2009–2016 (BEEPS IV-V) panel data for applicable countries in this study.^[1]

Survey responses regarding the electric power system, telecommunications system, transportation system, water and wastewater systems are analyzed using appropriate empirical models. Technology adoption, which is related to the electricity and telecommunications systems is also analyzed, including the use of email, websites, high-speed Internet services, and cell phone use in firms. For transition indicators, we use EBRD assessments of the state of transition, reported its annual transition report. We use infrastructure system indicators at both a point in time and changes in the indicators over time to assess whether improvements in these indicators are associated with reductions in reported obstacles by enterprises.

Background and Literature

Perceived infrastructure obstacles may arise from both demand and supply side conditions. On the demand side, in a growing economy new firms arise, and existing firms expand their activities. Both aspects may put pressure on infrastructure systems that prove to be inadequate, thereby resulting in firms reporting obstacles. On the supply side, antiquated or deteriorating infrastructure systems may result in inadequate services leading to a firm perception of obstacles. In the transition countries we study, the legacy systems may be outdated or may have deteriorated during the transition period with inadequate maintenance. Policy reforms that liberalize or rationalize tariff rates for infrastructure services can also have the effect of raising costs for firms, perceived as presenting additional obstacles. We are not able in this article to disentangle the demand and supply side causes of firm perceptions of obstacles, but we are able to model the extent of those perceptions as they are reflected in the BEEPS survey data.

A recent comprehensive review of the effects of infrastructure systems on firm performance is provided by Trang and Hong (2021). Their evidence indicates that there are several key influences by infrastructure systems on business activities. First, quality infrastructure systems are essential for firms’ growth and competitiveness. Second, transportation infrastructure, specifically roads, plays a role in regional patterns of absolute advantage thereby explaining observed efficiency differences. Third, the quality of power supply and telecommunications infrastructure is important in explaining variations in comparative advantage and firm product differentiation. These results point to the importance of infrastructure systems as they affect firm performance and implicitly indicate the ways that those systems may present obstacles for firms. Their evidence includes firms in transition countries, as in the present study, going back to the origins of the transition process. In their review of the literature covering metrics of infrastructure quality and firm performance, Trang and Hong (2021) report that, at the country level, the EBRD BEEPS dataset is broad. Accordingly, in our empirical estimations we use the BEEPS metrics of infrastructure quality.

In the 1990s, at the beginning of the transition process for countries of Central and Eastern Europe and the former Soviet Union, there was a flurry of infrastructure studies for transition countries as they began the transformation of their systems from highly centralized modalities to newer technologies and institutional arrangements. Researchers examined reforms in the electricity industries (Stern and Davis 1998; Lampietti 2004; Levai and Jaszay 1991; Newberry 1994; Krishnaswamy 1999), and technology innovations and policy options for various transition countries (Choi 1995; Davies et al. 1996; Sallai et al. 1996; Welfens 1995). Water system studies were generally conducted by the Organisation for Economic Cooperation and Development (OECD) and the World Bank, as they served as funders for infrastructure improvements.^[2] Later studies in the 2000s examine the diffusion of mobile telecommunications in Central and Eastern Europe (Gruber 2001). Pissarides et al. (2003) explore the effects of constraints reported by small and medium entrepreneurs (SMEs) in Bulgaria and Russia, finding that chief executive officers (CEOs) reported securing infrastructure was the most important problem they faced (51 % of CEOs in Russia and 59 % in Bulgaria).

The EBRD conducts regular assessments of the state of transition, reporting its assessments in its annual transition report. The EBRD also reports the methodology used in its survey data collection.^[3] As part of that effort, the EBRD produces transition indicators for aspects of the economies of transition countries, as well as for infrastructure systems. While these indicators have been used extensively in the transition literature, they are not without critics as in Myant and Drahokoupil (2012). In what follows, we will use these indicators in our analysis and discuss their potential strengths and weaknesses for use in the assessment of infrastructure systems.

Three studies use BEEPS data, of which we employ a newer version. Among more recent analytical studies, Iimi (2011) uses the 2005 (BEEPS III) data to analyze the impact of improving the quality of several infrastructure systems. The results for the electricity sector indicate that firm costs significantly increase with electrical outages, in cases of both frequency and duration. Water supply interruptions are also shown to weaken firm competitiveness. While the effects observed vary by industry, the firms hardest hit are in manufacturing, construction, and hotel-restaurant sectors. While informative, this study has severe limitations due to lack of data provided in the BEEPS survey. Most notably, there are no physical output or input measures, prices of inputs, measures of electricity and fuel consumption, among others. To address the BEEPS data limitation, the author uses various proxies. Given these severe data limitations, the estimated results of cost functions and stochastic frontiers are likely to be highly unreliable.

Anderson (2019) uses the 2012–2016 (BEEPS V) data for firms in transition countries together with country characteristics to explain reported obstacles due to access to land and permits. Access to land is a fundamental issue related to complementarity with infrastructure systems. Results indicate that government restrictions that effectively limit access to land and permits result in significant effects on reported obstacles. He also finds that limited access to land and permits is associated with firms making informal payments to “get things done.” In that analysis, the restrictions on land and permits are viewed as implicit constraints that impose shadow costs on the firm. In a related study, Carlin et al. (2007) examine a variety of bottlenecks firms face that increase the cost of operations. Consistent with their approach, we view the survey responses by firms as indicating aspects of their external environment that affect firm performance. Firm reports of obstacles can reflect the shadow cost of various constraints faced while doing business, including infrastructure system constraints.

The Correa et al. (2010) study uses the 2002 (BEEPS II) and the 2005 (BEEPS III) cross-sectional data as well as the panel composed of those two cross sections to investigate technology adoption among firms in Eastern Europe and Central Asia. They use ISO 9000 certification and web use by firm as proxies for technology adoption. Their results show a negative and statistically signification association between ISO certification and infrastructure for the 2002 cross-sectional sample. In the web use models, however, the correlation between web use and infrastructure index is positive and significant in the 2002 and 2005 cross-sectional samples but negative in the panel sample. They speculate that a negative effect of improved infrastructure on technology adoption may be due to the effect of infrastructure operating through other variables included in the model. They also speculate that their infrastructure index does not account for the costs of poor accessibility and associated risks of loss in transit which they consider to be the most important factor in infrastructure-related ISO certification.

Our paper advances the literature by examining the driving forces of firm views of infrastructure and technology obstacles. We do so by analyzing a rich set of the 2009 (BEEPS IV) and the 2012–2016 (BEEPS V) responses. Our model accounts for whether a country’s EBRD transition indicators are associated with firm perceptions of infrastructure and technology obstacles.

This study proceeds as follows. The next section describes the data used in the analysis of key infrastructure systems, both the BEEPS V cross-section data and the BEEPS IV-V panel data. That is followed by an explanation of the modeling approach used in the analysis and presentation of the main empirical results together with extended evidence based on panel data. The final section concludes with a summary and recommendations.

Empirical Methods

In our empirical models we use the BEEPS survey metrics on firm perceptions of infrastructure quality as the dependent variable. Explanatory variables include controls for the macroeconomic conditions of each country, individual firm characteristics, EBRD indicators of reform in each sector, and industry and regional controls.

For each of the infrastructure systems, electric, transportation, and telecommunications, the BEEPS survey asks firms to report the severity of obstacles they face. Responses include no obstacle, minor obstacle, moderate obstacle, major obstacle, and very severe obstacle (along with “don’t know” and “does not apply” responses, which are omitted from the data). Responses related to email, website, cell phone, and high-speed usage by firms are dichotomous (1 = “yes” and 0 = “no”). Water system obstacles include two sets of questions. The first question asks whether a firm applied for a water connection. Further, the second question collects answers on how many days it took from application to water connection.

Consequently, the estimation method used for electricity, transportation, and telecommunication obstacles is to estimate ordered logit models. The dependent variable takes on the five discrete values corresponding to survey responses, ordered from least to greatest degree of obstacle. For the technology adoption analysis (email, website, high-speed Internet, and cell phones), we use logit models where the dependent variable is dichotomous (0/1). In particular, the model takes the following general form:

(1) Y ic = β 0 + β 1 macro c + β 2 firm i + β 3 EBRD c + μ s + γ r + ε ic ,

where Y _ic is the survey response by firm i in country c regarding an infrastructure obstacle. We control for country macroeconomic indicators and firm-specific characteristics. The next section and Tables A1 and A2 in the supplementary online appendix describe variables included in the regression model. Additionally, we estimate equation (1) with regional and industry (sector) dummies, denoted as γ _r and μ _s, respectively. The error term is ε _ic.

Using equation (1), we estimate two models the first of which, Model (1), follows equation (1). A second variant, Model (2), adds the EBRD transition indicators of the relevant infrastructure system for each country in 2010, and the change in that indicator over the period 2005 to 2010. This second model variant allows us to distinguish the effects of the EBRD infrastructure system transition indicators as influences on reported obstacles, conditional on macroeconomic, firm, and industry characteristics.

In our analysis of water and wastewater system connections, there is additional data available, so we use the Heckman sample selection model as described in Maddala (1983). That approach simultaneously estimates a logit model that shows whether the enterprise applied for a water connection and a model of the number of days it took from application to connection. This approach is needed to account for the fact that enterprises applying for water connections may be systematically different from those not applying. Conditional on an enterprise applying for a connection, we then want to know how long it took to make the connection.

Data

The main analysis uses data from the fifth round of the EBRD-World Bank BEEPS survey. In the robustness check, we also report results using the panel data which combines the 2012–2016 (BEEPS V) with the 2009 (BEEPS IV) cross sections. The panel data covers fewer enterprises, but the matched data allow for comparisons of changing infrastructure effects for enterprises over time (from 2009 through 2012–2016). In the following subsections, we describe each dataset.

Cross-Sectional Data (BEEPS V)

The BEEPS V data were collected at the firm level over the period from 2011 to 2014, depending on the country. A total of 16,566 enterprises in 32 countries are included in the survey data. The data cover a broad range of topics related to the business environment and performance of firms. The survey also includes numerous questions on infrastructure issues. Data for 28 transition countries are used in this analysis. The countries included in this study are the East European and Caucasus (EEC) countries of Armenia, Azerbaijan, Belarus, Georgia, Moldova, and Ukraine, and the Southeast European (SEE) countries of Albania, Bosnia and Herzegovina, Bulgaria, former Yugoslav Republic (FYR) countries of Macedonia, Kosovo, Montenegro, Romania, and Serbia, the Central European and Baltic (CEB) countries of Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, Slovenia, the Central Asian (CA) countries of Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan, and the Russian Federation. Sample sizes vary among the countries in the data set. Generally, there are approximately 250 to 600 survey responses per country. The exceptions are Ukraine with approximately 1000 responses and Russia with approximately 4000.^[4]

The BEEPS V dataset is augmented with country-level data on economic conditions. The country variables are taken from World Bank World Development Indicators (WDI) data and include Gross Domestic Product (GDP) per capita (in the 2010 constant US dollars), the GDP growth rate (%), the size of the agricultural sector of the economy (%), and a total tax rate applied to corporations (%). In addition, an EBRD measure of the private ownership share of the economy (%), and a broad measure of overall economic reforms (computed from EBRD indices) are included. In order to capture the historical and institutional contexts of the disparate transition countries, dichotomous (0/1) regional indicators of whether the country is part of the Commonwealth of Independent States (CIS), EEC, SEE, CEB, and CA regions are included. Russia is the country left out, not included in the EEC, SEE, CEB, or CA variables. So, estimated coefficients on these variables in the models to follow can be interpreted relative to Russia.

Table A1 in the supplementary online appendix reports summary statistics for the cross-sectional variables. These statistics are computed using the BEEPS V sample weights (wstrict), not the raw unweighted data. The top section of the table reports country characteristics. The weighted mean GDP per capita is $9065 (PPP 2010). GDP per capita ranges from a minimum of $815 in Tajikistan to a maximum of $22,860 in Slovenia, revealing wide variation among the countries included in the data set. GDP growth rates average 2.24 %, ranging from a low of −2.72 % in Montenegro to a high of 12.32 % in Mongolia. Agricultural value-added averages 6.99 % of GDP, ranging from a low of 2.07 % in Slovenia to a high of 26.6 % in Tajikistan. The EBRD overall reform index (ORI) has a mean of 3.19 on a four-point scale and ranges from a minimum of 2.2 in Belarus to a maximum of 4.0 in Estonia. CIS countries account for 28 % of the firms; FYR countries for 13.5 %; and the Baltics for 5.6 %. While not reported in the table, it is notable that the average private ownership share of the economy is 65.81 %, varying from a low of 30 % in Belarus to a high of 80 % in Estonia, Hungary, and the Slovak Republic. Additionally, total tax rates applied to business profits average 44.07 %, ranging from a low of 7.4 % in Macedonia to a high of 98.7 % in Uzbekistan.

The second panel of Table A1 reports summary statistics for the firm characteristics used in the analysis. Firm size is a modest 39.56 fulltime employees on average, ranging from a minimum of one employee to a maximum of 11,000 employees. Managers of those firms generally have a good deal of experience, averaging approximately 16 years, and a minority of managers are female (20.7 %). The percentage of firms that report that they compete with informal firms is 35 %. Only 9 % of the firms indicate that they are subsidized by the government. Most firms are privately held or traded firms (86 %). Sole proprietors account for 7.2 % of the firms. Most of the firms have been private from the start (88.3 %), with only 1.4 % being private subsidiaries or former state-owned enterprises (SOE). Only 1.5 % are current SOEs.

Reported infrastructure obstacles are summarized in the third panel of Table A1. For electricity, the mean score is 0.92 on the four-point scale where zero indicates no obstacle and four indicates a very severe obstacle. Telecommunications and transportation have somewhat lower mean scores –0.76 each. Dichotomous indicator variables for firm use of email, websites, and high-speed Internet are also included in the analysis. The percentage of firms reporting using email stands at 86 %, while 61.4 % indicate use of a website, and 83.1 % report use of high-speed Internet service.

Industry control variables are reported on the bottom panel of the table. Firms that are manufacturing firms in the survey data account for 36 %. The most frequent industries represented include retail establishments that account for nearly 16 % of the firms in the data, while wholesalers account for another 22.2 %, and construction firms account for 13.8 %. The remaining industries represented generally account for 2 % or less.

Panel Data (BEEPS IV–V)

The panel data set is based on the BEEPS IV and BEEPS V cross-sectional data sets. The dataset contains a long list of indicators describing firms’ performance and business environment, including firms’ characteristics, location, and the industries in which they operate. Private firms operating in all types of industries were covered in the surveys. The full dataset covers thirty countries, but three were dropped from our analysis (Kosovo, Czech Republic, and Turkey) due to the unavailability of transition indexes data that is described later. The resulting sample covers firms located in 27 transition countries located in Central Eastern Europe and Baltic states, Southeastern Europe and the Commonwealth of Independent States, and Mongolia. The date of survey data collection varies by country. Firms in 25 countries were interviewed in 2008 and 2012. Russian Federation firms were interviewed in 2008 and 2011 and those in the Slovak Republic were interviewed in 2008 and 2013. On average, 95 firms were interviewed in each country with the smallest response of ten firms in the Slovak Republic and the largest response of 192 firms in Ukraine. The final sample is a balanced panel of 2504 firms across two waves of data with a total of 5008 observations.

Table A2 in the supplementary online appendix reports summary statistics for the BEEPS IV–V panel data. For brevity, we will only comment on the clear differences with respect to the cross-sectional data. In terms of macroeconomic indicators, the panel data set has a lower mean GDP per capita ($7073 vs. $9065 in PPP 2010), likely due to the different time periods covered: from 2011 to 2014 for the cross-sectional data and from 2008 to 2012 for the panel data. GDP growth rates also differ substantially with the mean growth rate in the panel data being 7.16 % versus 2.24 % in the cross-sectional data set. Again, it is most likely that the difference reflects the earlier time frame which included the recovery from the global Great Recession. Among the firms’ characteristics on average, somewhat larger firms are covered in the panel data set (62 fulltime employees vs. 39.5). It is also evident that the firms in the panel are less likely to be privately held or traded (63 % vs. 86 %). Furthermore, firms in the panel data are more likely to be privatized SOEs (17.3 % vs. 6.9 %) and less likely to be originally private firms (76.4 % vs. 88.3 %).

All the infrastructure obstacle indicators reveal higher degrees of problems in the panel data set (electricity, telecommunications, and transport) and less use of technology (email, websites, and high-speed Internet). These differences reveal that progress was made over time as the later cross-sectional indicators reveal reduced obstacles and greater use of technology by firms. Industries represented by firms in the surveys also changed over time. The panel data set covers more retail firms, fewer wholesalers, and fewer construction firms compared to the cross-sectional data set.

Empirical Results

In what follows we report the main results of firm responses related to (1) the electric power system, (2) the telecommunications system, (3) the transportation system, and (4) the water and wastewater system that rely on BEEPS V cross-sectional data. In addition, we also estimate models of email, cell phone, website, and high-speed Internet usage by firms.^[5] Adoption of these technologies by firms indicates firm innovation or adaptation, likely depending on the state of other infrastructure systems, especially electricity and telecommunications systems.

Electric Power System

The first question we investigate is the extent to which firms report the electric power system as an obstacle. Figure 1 presents the summary of key estimates, revealing correlates for firm responses on how much of an obstacle the electric power system is. The dependent variable is discrete, taking on the values from zero to 4, where zero indicates no obstacle and 4 indicates a very severe obstacle. Model (1) includes macroeconomic variables as controls for the economic conditions of the countries in which firms operate, firm-specific characteristics, industry sectors, and regional indicators. The model also includes the country’s Getting Electricity score from the World Bank Doing Business data as an indicator of system quality. Model (2) adds two variables that are indicators of electricity system reform: (i) the EBRD indicator of the electricity sector for each country in 2010 and (ii) the change in that indicator over the period 2005 to 2010.^[6]

Figure 1:

Electricity system, BEEPS V. Source: Authors’ elaboration. Notes: Point estimates indicated by triangles are from Model 1, whereas those marked with circles are from Model 2. Horizontal bars indicate 95 % confidence intervals. Variable name abbreviations used in these and other figures are defined as follows: Overall Reform Index (ORI); International Quality Certification (IQS); Commonwealth of Independent States (CIS); Eastern Europe and the Caucasus (EEC); Southeastern Europe (SEE); Central Europe and the Baltics (CEB); Central Asia (CA).

Macroeconomic conditions are significantly associated with reported obstacles. GDP per capita is associated positively with reported obstacles, indicating that, in more advanced economies, firms report more electricity system obstacles. The GDP growth rate variable has a positive estimated coefficient indicating that, in faster growing economies, firms also report more obstacles. These results point to potential bottlenecks for firms caused by inadequacies of the electricity system, consistent with the evidence on bottlenecks in Carlin et al. (2007). The agricultural value-added variable of the country also has a positive and significant effect, indicating that firms operating in more agricultural countries reported more obstacles. Additionally, the EBRD ORI variable has a positive and significant estimated coefficient, indicating that firms operating in countries with more advanced reforms report more electricity system obstacles. This result may reflect higher expectations of firms in more advanced reform countries so that when there are problems with the electricity system, they are more likely to be disappointed and report obstacles. This result is similar to that of the EBRD electricity system indicator discussed below and may reflect the tendency to report an obstacle rather than the obstacle itself.

Several firm-specific characteristics are also correlated with reported electricity system obstacles. Firms headed by a female manager report greater electricity system obstacles, which may be due to sample selection (i.e., female-managed firms tend to be in specific sectors more prone to such obstacles) or may be due to gender-specific tendencies in reporting obstacles. Firms competing with informal market firms also report greater obstacles. This may be due to a general sense of frustration that they are not competing on a level playing field, and any obstacle they confront in that context is sensed as more severe. However, firms with a longer presence report fewer electricity-related obstacles. This may be because more experienced firms tend to adjust better to the electricity outages and limitations. Firms that are subsidized by the government also report greater obstacles. This effect may be because subsidized firms tend to be pre-reform enterprises that are depending on older electricity system infrastructures that are prone to break down more frequently or have limited capacity. It is also consistent with the Carlin et al. (2007) bottleneck finding that state-owned firms are less productive.

The Getting Electricity score from the Doing Business data has a negative and statistically significant estimated coefficient indicating that the more advanced a country’s electricity system, the fewer the obstacles are reported by firms. The result confirms that an objective indicator of electricity system quality is associated with fewer obstacles that firms confront.^[7]

Regional dummy variables have significant estimated coefficients that are positive for CIS countries and negative for EEC countries. Firms operating in CIS countries report greater electricity system obstacles, and those operating in EEC countries report fewer. Firms in SEE countries report greater electricity system obstacles, but the estimated positive coefficients are marginally significant. The CEB coefficient is negative and significant in Model (1) but not significant in Model (2). Conversely, the CA coefficient is not significant in Model (1) but becomes significant in Model (2).

The second question we investigate is whether advances in the electricity system reduce obstacles. This is done in Model (2) with the inclusion of the EBRD indicator for the electricity system, and its change over time, to highlight their effects. The EBRD electricity indicator coefficient is positive and significant, as is the coefficient for the change in that indicator from 2005 to 2010. As the electricity system of a country is rated as more advanced by the EBRD, firms report more electricity system obstacles, not fewer. This apparently counterintuitive result drives us to examine more deeply what the EBRD indicator is capturing. Table A4 in the supplementary online appendix reports the elements considered in the EBRD indicators. Several factors used in that indicator reflect liberalization of electricity prices. Firms in countries with higher EBRD indicator ratings may be reacting to higher electricity tariffs, indicating that they face greater obstacles. Similarly, positive changes in the EBRD indicator may reflect greater liberalization and higher tariffs to which firms are reacting. The inclusion of these indicators in Model (2) also has the effect of reducing the significance of the ORI which was positive and significant in Model (1) but loses its significance in Model (2).^[8]

Telecommunications Systems

Cross-sectional estimates for the telecommunications systems are reported in Figure 2. The dependent variable in these models is discrete, taking on the values zero to 4, where zero indicates no obstacle and 4 indicates a very severe obstacle.

Figure 2:

Telecommunication system, BEEPS V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates indicated by triangles are from Model 1, whereas those marked with circles are from Model 2. Horizontal bars indicate 95 % confidence intervals.

Among the macroeconomic variables, four of the variables have positive and significant coefficients: GDP per capita, GDP growth rate, agricultural value-added, and the ORI. Firms operating in countries with higher GDP per capita report greater telecommunications system obstacles. Similarly, firms in countries with faster rates of GDP growth report greater obstacles. Additionally, firms in countries with a strong agricultural sector report having more obstacles related to telecommunication systems. Finally, the ORI variable has a positive and significant coefficient. These results may reflect the fact that, with greater economic development and advancement, bottlenecks may arise in the telecommunications systems causing greater obstacles for firms.

Among firm characteristics that are significant, the international quality certification variable, competition with informal market firms, and subsidies from the government all have positive and significant coefficient estimates. Firms with these characteristics report greater obstacles regarding the telecommunications systems in their countries. The only firm characteristic with a significant negative coefficient is establishment age, indicating that older establishments are less likely to report telecommunications system obstacles. As will be seen later in the article, older establishments are more likely to be technology adopters using email, websites, high-speed Internet, and cell phones in the conduct of their business. Hence, they are more likely to be confronted with limitations that arise in the telecommunications systems. Finally, industry indicators are negative and significant for firms in the textile and garment industries, and positive and significant for firms in publishing, IT, and post and telecommunications industries.

Four of the regional dummies are statistically significant, positive for firms in CIS countries and negative for firms in EEC, SEE, and CEB countries. Firms operating in these regions report lower levels of telecommunications system obstacles relative to Russia, which is the omitted region, other things being equal.

The inclusion of EBRD indicators in Model (2) changes the estimated coefficients very little. Neither of those variables have coefficient estimates that are statistically significantly different from zero. Hence, the EBRD indicators of telecommunications systems in those countries appear to have no influence on firm responses. Neither the level of the indicator nor its change affect firm responses. Among the macro indicators in telecommunications systems analysis, the agricultural value-added variable is negative and significant. This result shows that firms in countries with strong agricultural sector report having a higher degree of obstacles related to telecommunication systems. Conversely, firms in countries with fast economic growth have fewer telecommunications obstacles.

Transportation Systems

Cross-sectional estimates for the transportation systems are reported in Figure 3, providing correlates for firm responses on how much of an obstacle this system is to the current operations of the establishment. The dependent variable is discrete, taking on the values zero to 4, where zero indicates no obstacle and 4 indicates a very severe obstacle. Once again, Model (1) includes macroeconomic variables to control for the economic conditions of the countries in which firms operate, firm-specific characteristics, industry sectors, and regional indicators.

Figure 3:

Transportation system, BEEPS V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates indicated by triangles are from Model 1, whereas those marked with circles are from Model 2. Horizontal bars indicate 95 % confidence intervals.

Among the macroeconomic variables, GDP per capita has a positive and significant coefficient, indicating that, in more advanced economies, firms report greater transportation system obstacles. In addition, the ORI variable also has a positive and significant coefficient indicating that firms in countries with more advanced reforms also report greater obstacles. In Model (2), there is the additional significance of GDP per capita growth and agricultural added value. Firms in faster-growing economies report significantly greater obstacles with the transportation systems, but those in more agriculturally oriented economies report fewer obstacles.

Among firm-specific characteristics, we see that firms with more employees, competition with informal market firms, and those subsidized by the government report greater transportation system obstacles. Conversely, firms with more experienced managers, or female managers, report less severe transportation system obstacles. Apparently, managerial experience helps to overcome such obstacles. Firms operating in several industry sectors report greater transportation system obstacles: food, wood, publishing, chemicals, plastics and rubber, nonmetallic mineral products, furniture, retail, and wholesale sectors. In all these sectors, transportation obstacles are greater. Only in the IT industry sector do firms report significantly less severe obstacles with the transportation system. Four of the regional dummies are statistically significant: EEC, SEE, CEB, and CA. Firms operating in those regions all report fewer transportation system obstacles (relative to Russia).

On the question of whether advances in the transportation systems reduce obstacles from firms’ point of view, the results for the EBRD transportation system indicators reflect opposing effects. The roads indicators, for both the level and change in road system quality, are positive, indicating that the more advanced the country’s road system the more severe the obstacles reported by firms. Contrary to that finding, the indicators for railroads have negative signs indicating that the better the rail system the less severe the reported transportation system obstacles. These results reveal a striking difference as roads and railroads have opposite impacts on firm reports of transportation system obstacles. In this section we investigate the relationship between a country’s transportation system and private sector operation. The agricultural value-added variable is negative and highly significant in that both models indicate that firms in more agricultural economies report that transportation is less likely to be an obstacle.

Water and Wastewater Systems

For water and wastewater systems we estimate two types of models. First, we estimate a logit model of firms’ water system connections. Then we estimate a Heckman model explaining the number of days reported by firms for water connections to be completed, conditional on their having applied for a water connection. The BEEPS V survey does not ask about water and wastewater system obstacles, so the panel data do not include this system: hence the alternative modeling approach used in this case. Table A3 in the supplementary online appendix reports the results of both models: (1) a firm’s application for a water connection and (2) conditional on application, the number of days waiting for a water connection.

In the water connection model, the dependent variable is a dichotomous indicator (0/1) of whether the firm applied for a water connection. In that model, none of the macroeconomic variables have discernible effects. That is, macroeconomic conditions do not appear to have any impact on firm applications for water connections. The ORI coefficient is negative and significant, however, indicating that, in countries with more substantial reforms, firms are somewhat less likely to have applied for a water connection.^[9]

Several firm-specific characteristics affect the likelihood of firm applications for water connections, however. The number of employees has a strong positive effect on the likelihood of water connection applications. Larger firms are more likely to apply for water connections. This may be because larger firms have more establishments requiring such connections. Firms with more experienced managers are also more likely to apply for a water connection. More experienced managers may be more familiar with the application process and more inclined to apply when a connection is needed, compared to a less experienced manager. It may also be that a more experienced manager has relational capital with government officials on which to rely when applying for a connection, increasing the likelihood of success and thereby lowering the perceived application barriers. Firms with international quality certification are more likely to apply for a water connection. External recognition of the firm thus gives the firm formal standing, enabling it to apply for a connection. Firms operating in competition with informal market firms report a greater likelihood of applying for a water connection. Since these firms are operating legally, they are more likely to apply for a connection relative to informal market firms that cannot apply for a connection but must rely on illegal means of obtaining a connection (e.g., tapping into an existing water line). Government subsidized firms are also more likely to apply for water connections, which may reflect the need for updating water connections because subsidized firms tend to be older firms.

The nature of the firm also matters, with privatized SOEs, originally private firms, private subsidiaries of former SOEs, and sole proprietors all reporting a lower likelihood of applying for a water connection, relative to SOEs for which the estimated coefficient is not significantly different from zero. Only joint ventures with foreign firms are more likely to apply for a water connection. Firms located in larger cities are also less likely to apply for a water connection. This is likely because in a larger city there is already a well-established water system connecting most buildings and other facilities. Only three of the industry indicators are significant: wood, machinery and equipment, and IT. Firms in those industry sectors are less likely to apply for water connections. Regionally, firms located in CIS and CEB countries are significantly less likely to apply for water connections.

Inclusion of the EBRD water and wastewater reform indicator and its change is significant only for the 2010 level of the indicator. The estimated coefficient on that indicator is positive and significant indicating that firms operating in countries with more advanced reforms in their water systems are more likely to apply for a water connection. The ratings for that indicator, provided in the Appendix, reveal that the higher the rating the more commercialized the water system and the more rationalized the tariffs charged. Consequently, it is not surprising that firms report greater likelihood of applying for a water connection in countries with more highly reformed water systems.

Conditional on applying for a water connection, the second model estimates determinants of the number of days the firm must wait for that connection.^[10] Three firm characteristics have statistically significant positive effects on the number of days for a water connection: (1) firms with more fulltime employees, (2) firms with more experienced managers, and (3) firms subsidized by the government. These characteristics are indicative of captive firms that are more closely associated with the government, for which the water monopoly can be less responsive. The nature of firms also matters as privatized SOEs, private subsidiaries of former SOEs, joint ventures with foreign firms, privately held firms, sole proprietors, and partnerships all report that they wait significantly fewer days for water connections. Location also influences the waiting time for a connection. Firms in larger cities, along with those in CIS, EEC, SEE, and CEB countries report shorter waiting times.

Email, Website, Cell Phones, and High-Speed Internet Use

Firms adapt to conditions in their countries and industries, including the infrastructure systems they use, by implementing new technologies. These forms of electronic technology essentially represent a new infrastructure system that is critical for the conduct of contemporary business. In this section we analyze patterns of technology adoption including email, websites, cell phones, and high-speed Internet use.

Figure 4 reports the results of logit model estimation for email and cell phone usage by firms, estimated using cross-sectional data. Figure 5 illustrates model estimation for high-speed Internet and website use, again using cross-sectional data. In each case, the BEEPS questions on firm use of these technologies are used in analyzing the yes/no responses. A common set of regressors is used in all four models estimated, including macroeconomic variables, firm-specific characteristics, industry indicators, regional location dummies, and two EBRD indicators for the telecommunications systems.

Figure 4:

Email and cell phone, BEEPS V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates marked by triangles are obtained from logistic regression with the outcome variable “Email.” Point estimates marked by circles are from similar regression, but the dependent variable is “Cell phone.” Horizontal bars indicate 95 % confidence intervals.

Figure 5:

High-speed internet, BEEPS V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates marked by triangles are obtained from logistic regression with the outcome variable “Website.” Point estimates marked by circles are from similar regression but with the dependent variable “High-speed Internet.” Horizontal bars indicate 95 % confidence intervals.

GDP per capita has a positive and significant estimated coefficient for all four technologies. Firms operating in more advanced economies are more likely to use websites. Curiously, the GDP growth rate is negative and significant for three of the models: email, cell phone, and high-speed Internet. This result implies that in faster-growing economies firm technology adoption is lower. However, recall that the GDP growth rate is greater in the panel data because it was collected earlier than the cross-sectional data, which may partially explain this result. Firms operating in faster growing countries report that they are less likely to use these technologies. The agricultural value-added variable is only significant in the website model, with a positive coefficient indicating that firms in more agricultural economies are more likely to use websites, all other things being equal. Also, the ORI is only significant in the email model, indicating that firms operating in countries with more advanced reforms are less likely to use email communications.

Many of the firm characteristics are associated with these forms of technology adoption. Firms with more employees are more likely to adopt all four forms of technology, as are firms with international quality certifications. Competition with informal market firms increases the likelihood of email and websites. Subsidies from the government increase the likelihood of website, high-speed Internet, and cell phone use. The manager’s experience increases the likelihood of website and cell phone use but reduces the likelihood of high-speed Internet use. Curiously, a female manager reduces the likelihood of all four forms of technology adoption. Finally, firms located in larger cities are more likely to adopt email, websites, and high-speed Internet use. This set of results may reflect greater availability in the larger cities.

Patterns of adoption also vary across industry sectors. Firms in publishing, chemicals, IT, and post and telecommunications are more likely to adopt these technologies. Firms in the garment industry and retail are less likely to adopt. As for regional effects, the patterns vary. Among CIS countries, firms are less likely to adopt email and high-speed Internet. Among EEC countries firms are more likely to adopt websites and high-speed Internet but less likely to adopt cell phone use. Among SEE countries, firms are less likely to adopt email but more likely to adopt websites and high-speed Internet services (which are clearly complementary). Finally, among CA countries, firms are less likely to adopt email and websites.

Inclusion of the EBRD indicator for the telecommunications systems and the rate of change in that indicator have mixed results. The level indicator is positive and significant in the website adoption model, but negative and significant in the cell phone adoption model. The rate of change in the EBRD telecommunications indicator is positive and significant in the email and high-speed Internet models. These results indicate that firms in faster growing economies, or those with more advanced reforms, are less likely to report use of email and high-speed Internet. Similarly, firms in more agricultural countries are less likely to use email and high-speed Internet. Recalling that Mongolia has the highest GDP growth rate in the data set, it may be that these results simply reflect the fact that fast-growing economies and those with less overall reform are less developed economies with low base levels of these indicators at the start.

Robustness Checks

This section provides results using panel data that combine BEEPS V and BEEPS VI. We estimate equation (1) for all types of infrastructure obstacles, except for the water and wastewater variable. The water and wastewater data are not available in the panel data, so this type of infrastructure obstacle is analyzed only using cross-sectional data. In what follows, we will mainly discuss how our results using panel data differ from the cross-sectional data results.

Electricity

Figure 6 presents estimates for the electric power system, revealing correlates for firm responses on how much of an obstacle this system is to the current operations of the establishment. Among the macroeconomic variables, the EBRD ORI variable has a positive and significant estimated coefficient, indicating that firms operating in countries with more advanced reforms report greater degrees of electricity system obstacles. This result may reflect higher expectations of firms in more advanced reform countries so that when there are problems with the electricity system, they are more likely to be disappointed and report obstacles. This result is similar to that for the EBRD electricity system indicator discussed below and may reflect the tendency to report an obstacle rather than the obstacle itself. Firm-specific characteristics display similar signs of correlation with obstacles as the results reported using cross-sectional data. As for regional dummy variables, we do observe some differences between the estimates reported in cross-sectional results and those in this section. Our panel data analysis shows that firms operating in SEE and CA countries face significantly fewer electricity system obstacles. With the advantage of panel data, we are able to identify an improved trend in obstacles due to the electricity system in these countries.

Figure 6:

Electricity system, BEEPS IV–V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates marked by triangles are from Model 1, whereas those marked by circles are from Model 2. Horizontal bars indicate 95 % confidence intervals.

The second question we investigate is whether advances in electricity system reform reduce obstacles from firms’ point of view. The difference between Model (1) and Model (2) is the inclusion of the EBRD indicator for the electricity system, and its change. The estimated coefficient for the change in that indicator is negative and significant.^[11] Apparently, when the electricity system of a country is rated as more advanced by the EBRD, firms report fewer electricity system obstacles. This result is a reassuring indicator that the EBRD metric is capturing meaningful information about the reliability of the electricity system in each country.

Telecommunications Systems

Figure 7 provides correlates for firm responses on how much of an obstacle this system is to the current operations of the establishment. The macroeconomic, firm-specific, and regional variables display similar signs of correlates as those reported in cross-sectional results. The inclusion of EBRD indicators in Model (2) changes the estimated coefficients modestly. While the change in the EBRD SICI telecommunication indicator variable does not have a significant coefficient, the level indicator has a positive and significant estimated coefficient. The greater the telecommunication indicator, the greater the reported telecommunication system obstacles, all other things being equal. While this result appears anomalous, it may be that in countries with higher-quality telecommunication systems, firms experiencing difficulties have high expectations and are more likely to report obstacles when they occur. Hence, this result may primarily reflect the reporting behavior of firms.

Figure 7:

Telecommunication system, BEEPS IV–V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates marked by triangles are from Model 1, whereas those marked by circles are from Model 2. Horizontal bars indicate 95 % confidence intervals.

Transportation Systems

The first panel model estimates for the transport system reported in Figure 8 provides correlates for firm responses on how much of an obstacle this system is to the current operations of the establishment. Again, the macroeconomic and firm-specific characteristics display similar correlates as those reported in the cross-sectional results. As for regional dummies, two estimates are significant in both Model (1) and Model (2): SEE and CEB. Regional dummies for CA are negative, while the same estimates are positive using cross-sectional BEEPS V. However, it is worth noting that the panel data estimates for CA countries are not statistically significant in both model (1) and model (2).

Figure 8:

Transportation system, BEEPS IV–V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates indicated by triangles are from Model 1, whereas those marked by circles are from Model 2. Horizontal bars indicate 95 % confidence intervals.

In Model (2) the EBRD transportation system indicators for railroads and roads are included in both levels and changes. None of the variables have significant estimated coefficients. In this way, the EBRD indicators appear to have no association with reported transportation system obstacles.

Email, Website, Cell phone, and High-Speed Internet Adoption by Firms

Figure 9 reports estimates of logit models for email and Figure 10 reports similar set of estimates from logistical regression for websites and high-speed Internet usage by firms. Cell phone usage is not available in the panel data set.

Figure 9:

Email, BEEPS IV–V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates marked by triangles are from logistic regression with the binary outcome “Email.” Horizontal bars indicate 95 % confidence intervals.

Figure 10:

Website and high-speed Internet, BEEPS IV–V. Source: Authors’ elaboration. Notes: Variable name abbreviations are specified in the notes in Figure 1. Point estimates marked by circles are from logistic regression with the binary outcome variable “Website.” Point estimates marked by triangles are from a similar regression with the dependent variable “High-speed Internet.” Horizontal bars indicate 95 % confidence intervals.

Looking at macroeconomic variables, it can be seen that all estimates exhibit similar signs of correlation as estimates presented in Figures 4 and 5, except for the agricultural value-added variable. Apparently, firms in more agricultural countries are less likely to use email and high-speed Internet as well, which is understandable given the nature of agriculture in these countries.

The panel data correlation pattern of the firm characteristics and technology adoption aligns with cross-sectional correlation results. We found that larger firms, firms with internationally recognized quality certification, firms subsidized by the government, and older firms are more likely to adopt all three forms of technology. In contrast, managers with greater experience are significantly less likely to adopt high-speed Internet, which perhaps reveals that older managers are generally less inclined to use advanced technology. Being a publicly held or traded firm also increases the probability of using all three forms of technology. But sole proprietors and limited partnership companies are less likely to use email for their operations.

Patterns of adoption also vary by region. Email adoption is more likely in CEB countries, relative to Russia (the omitted region), based on the panel estimation, whereas it was not significant in the cross-sectional analysis. It is less likely in CA countries, consistent with the cross-sectional result. Based on the panel estimation, website adoption is less likely in EEC and CA countries. While the negative estimated coefficient for firms in CA countries is consistent with the cross-sectional estimate, the negative estimated coefficient for firms in EEC countries based on the panel model is the opposite sign compared to the positive coefficient in the cross-sectional analysis, although it is marginally significant. The panel estimates provide a more consistent negative pattern for website obstacles. High-speed Internet adoption is more likely in EEC, SEE, CEB, and CA countries.

For the second research question, the EBRD telecommunication reform index variable has a positive and significant coefficient for both email and high-speed Internet adoption in contrast to the cross-sectional results. This result indicates that firms operating in countries with higher-rated telecommunications system reforms are more likely to be adopters of these technologies. The EBRD telecommunications index change variable, however, has a negative and significant coefficient in the models of email and high-speed Internet adoption, indicating that firms in countries where there is an improvement in the reform index are less likely to use these technologies in their business operations. Finally, it should be noted that the year fixed effect indicator variable has a positive and significant estimated coefficient indicating that technology adoption in all three cases is increasing over time.

We observe some differences in signs for the estimated coefficients of the EBRD change indicators. A primary factor may be the differences in samples. Firms sampled in BEEPS IV–V represent 12 industries, while firms sampled in BEEPS V represent 19 industries. Moreover, the panel sample lacks firms that are more relevant to telecommunication infrastructures such as publishing, electronics, information technologies, and post and telecommunications. When comparing the results of cross-sectional and panel estimates, the different coefficient signs occur primarily in models with changes in EBRD telecommunication indicators. Those differences are likely due to differences in industry composition.^[12]

Summary and Conclusions

The results presented here provide insights on macroeconomic conditions, firm characteristics, industry sectors, and regions that affect reported infrastructure system obstacles for firms and their patterns of technology adoption. Our results indicate that macroeconomic conditions affect reported infrastructure obstacles. Firms operating in countries with higher GDP per capita, i.e., more advanced economies, report greater obstacles with electricity, telecommunications, and transportation systems. It appears that the older technology involved with electricity systems may provide more stable service, while the demands of a more advanced economy provide challenges for telecommunications and transport systems to provide adequate services. Firms in countries with higher GDP are clearly much more likely to adopt the use of email, websites, high-speed Internet, and cell phones in their business operations. Firm characteristics also have impacts on reported infrastructure system obstacles. Depending on the infrastructure system, firm size, competition with informal market firms, and having an international quality certification all have impacts on reported obstacles.

While we anticipated that higher values of the EBRD transition indicators would be correlated with fewer reported infrastructure obstacles, the empirical results generally indicate that they are associated with more reported obstacles. This finding may reflect the fact that higher ratings by the EBRD are often reflective of the liberalization and rationalization of tariff rates, so firms may be reacting to the reality that the infrastructure system services are costing them more, even as they improve. They report more severe obstacles as the cost of the service is higher. It may also be that as the infrastructure system is improved, firm expectations rise and when a system has a failure of some kind, firms are more likely to report their disappointment. In this case it is higher expectations for the system’s service that is driving the result. As a policy matter, transition governments pursuing infrastructure reforms must be aware that those reforms can elicit negative responses from firms that perceive the reforms as obstacles to doing business. While the reforms may be essential to long-run economic development they can be perceived in the short run as detrimental. To obtain support for reforms, policy makers must be able to frame the motivation for reforms in ways that communicate the ultimate benefits to firms.

Regional variation is clear in the obstacles reported by firms and the patterns of technology adoption. Firms operating in CIS countries report greater obstacles with the electricity and telecommunications systems in their countries. They are also less likely to adopt email and use high-speed Internet in their business operations. This result points to the legacy of the former Soviet system which continues to impede business activity. Those firms operating in Eastern Europe and the Caucasus countries report less severe obstacles with the electricity, telecommunications, and transportation systems. They are also more likely to adopt websites and high-speed Internet in their business operations, but they are less likely to use cell phones. Firms in Southeastern Europe are less likely to report obstacles with the telecommunications and transportation systems. They are less likely to use email but more likely to use websites and high-speed Internet in their business operations. Firms in Central Europe and the Baltics report less severe obstacles with the electricity, telecommunications, and transportation systems. Their adoption of technology in business operations is not significantly different from Russia, the reference region. Finally, firms operating in Central Asia are more likely to report electricity system obstacles but less likely to report transportation system obstacles. They are also less likely to use email and websites but more likely to use high-speed Internet services. Reported obstacles like those identified in this analysis reflect implicit shadow costs that signal areas to be addressed in policy reforms. The obstacles indicate where there are binding constraints that limit economic growth, as described in Carlin et al. (2007) and Hausmann et al. (2005). Once identified, reported obstacles can then be used for policy analysis and planning for infrastructure investments to eliminate bottlenecks and facilitate economic growth.

Future research on this topic will benefit from further updates in the data, including the next version of the BEEPS data and updated EBRD transition indicators. It would also be insightful to create a retrospective view of firms over time in the BEEPS data. While cross sections are available going back to 1999 (BEEPS I), the panel data set only includes the 2009 (BEEPS IV) and 2012–2016 (BEEPS V) cross sections. Exploration of the feasibility of creating a synthetic panel going back to 1999 would be helpful in covering a longer period of the transition process experienced by firms. Furthermore, research can be advanced by exploring the ways that infrastructure quality improvements over time enhance firm performance. Regional variations observed in the analysis deserve further research to identify specific conditions by country groups that may be responsible. Finally, as countries make substantial investments in public infrastructure, contributions to our knowledge of the return on those infrastructure projects will be enhanced as analysis moves to include improvements in the business environment.

Corresponding author: John E. Anderson, Department of Economics, 242612 University of Nebraska-Lincoln , Lincoln, NE, USA, E-mail: janderson4@unl.edu

About the authors

John E. Anderson

John E. Anderson is the Baird Family Professor of Economics and the Executive Director of the Central Plains Federal Statistical Research Data Center at the University of Nebraska-Lincoln. His fields of expertise include public finance, urban economics, and transition economics. He has served as an advisor at the President’s Council of Economic Advisers in the United States as well as an advisor to governments in Eastern Europe, the Balkans, and Central Asia.

Muazzam Toshmatova

Muazzam Toshmatova is a Postdoctoral Associate at the Heldrich Center for Workforce Development. She holds an M.S. degree in economics from Texas A&M University and her Ph.D. in economics from the University of Nebraska-Lincoln. Her research is focused on labor, public, regional, and transitional economics.

References

Anderson, John E. 2019. “Access to Land and Permits: Firm-Level Evidence of Impediments to Development in Transition Countries.” Journal of Economics and Business 10: 38–57.10.1016/j.jeconbus.2018.09.001Search in Google Scholar

Carlin, Wendy, Mark E. Schaffer, and Paul, Seabright. 2007. Where Are the Real Bottlenecks? Evidence from 20,000 Firms in 60 Countries about the Shadow Costs of Constraints to Firm Performance. IZA Discussion Paper 3059. Bonn: Institute of Labor Economics (IZA).10.2139/ssrn.1022970Search in Google Scholar

Choi, Jay Pil. 1995. Telecommunications Reform in Central and Eastern Europe. East European Series Working Paper 28. Vienna: Institute for Advanced Studies.Search in Google Scholar

Correa, Paulo G., Anna M. Fernandez, and Chris J. Uregian. 2010. “Technology Adoption and the Investment Climate: Firm-Level Evidence for Eastern Europe and Central Asia.” The World Bank Economic Review 24 (1): 121–47.10.1093/wber/lhp021Search in Google Scholar

Davies, Gareth, Steve Carter, Stuart McIntoch, and Dan Stefanescu. 1996. “Technology and Policy Options for the Telecommunications Sector: The Situation in Central and Eastern Europe and the Former Soviet Union.” Telecommunications Policy 20 (2): 101–23.10.1016/0308-5961(95)00057-7Search in Google Scholar

Gruber, Harald. 2001. “Competition and Innovation: The Diffusion of Mobile Telecommunications in Central and Eastern Europe.” Information Economics and Policy 13 (1): 19–34.10.1016/S0167-6245(00)00028-7Search in Google Scholar

Hausmann, Ricardo, Dani Rodrik, and Andrés Velaso. 2005. Growth Diagnostics. Cambridge MA: Harvard University, John F. Kennedy School of Government, https://growthlab.hks.harvard.edu/files/growthlab/files/growth-diagnostics.pdf (accessed 22 November 2023).Search in Google Scholar

Iimi, Atsushi. 2011. “Effects of Improving Infrastructure Quality on Business Costs: Evidence from Firm-Level Data in Eastern Europe and Central Asia.” Developing Economies 49 (2): 121–47.10.1111/j.1746-1049.2011.00126.xSearch in Google Scholar

Krishnaswamy, Venkataraman. 1999. Non-Payment in the Electricity Sector in Eastern Europe and the Former Soviet Union. Technical Paper 423. Washington DC: World Bank.Search in Google Scholar

Lampietti, Julian. 2004. Power’s Promise: Electricity Reforms in Eastern Europe and Central Asia. Working Paper 40. Washington DC: World Bank.Search in Google Scholar

Levai, Andras, and Tamas Jaszay. 1991. “Present and Future Electric Power Systems in Eastern Europe: The Possibilities of a Broader Cooperation.” In Energy and the Environment in the 21st Century, edited by Jefferson W. Tester, David O. Wood, and Nancy A. Ferrari, 595–600. Cambridge and London, UK: Cambridge University Press.Search in Google Scholar

Maddala, Gangadharrao S. 1983. Limited Dependent Variable and Qualitative Variables in Econometrics. Cambridge, UK: Cambridge University Press.10.1017/CBO9780511810176Search in Google Scholar

Myant, Martin, and Jan, Drahokoupil. 2012. “Transition Indicators of the European Bank for Reconstruction and Development: A Doubtful Guide to Economic Success.” Competition and Change 16 (1): 69–75.10.1179/1024529411Z.0000000005Search in Google Scholar

Newberry, David M. 1994. “Restructuring and Privatizing Electric Utilities in Eastern Europe.” Economics of Transition 2 (3): 291–316.10.1111/j.1468-0351.1994.tb00119.xSearch in Google Scholar

Pissarides, Francesca, Miroslav Singer, and Jan, Svenjar. 2003. “Objectives and Constraints of Entrepreneurs: Evidence from Small and Medium Size Enterprises in Russia and Bulgaria.” Journal of Comparative Economics 31: 503–31.10.1016/S0147-5967(03)00054-4Search in Google Scholar

Sallai, Gyula, Ivan Schmideg, and George, Lajtha. 1996. “Telecommunications in Central and Eastern Europe: Similarities, Peculiarities and Trends of Change in the Countries of Transition.” Telecommunications Policy 20 (5): 325–40.10.1016/0308-5961(96)00015-8Search in Google Scholar

Stern, John, and Junior R. Davis. 1998. “Economic Reform of the Electric Industries of Central and Eastern Europe.” Economics of Transition 6 (2): 427–60.10.1111/j.1468-0351.1998.tb00058.xSearch in Google Scholar

Trang, Ha Thi Thu, and Pham, Thi Thanh Hong. 2021. “Measuring the Impact of Infrastructure on Firm Performance: A Review of the Literature, Metrics, and Evidence.” Advances in Economics, Business and Management Research 196: 211–20.10.2991/aebmr.k.211119.022Search in Google Scholar

Welfens, Paul J. J. 1995. “Telecommunications and Transition in Central and Eastern Europe.” Telecommunications Policy 19 (7): 561–77.10.1016/0308-5961(95)00032-2Search in Google Scholar

Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/soeu-2023-0005).

Received: 2023-02-08

Accepted: 2023-11-13

Published Online: 2024-03-14

Published in Print: 2024-03-25

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Supplementary Material Details

Articles in the same Issue

https://doi.org/10.1515/soeu-2023-0005

Keywords for this article

infrastructure; electricity; water and wastewater; telecommunications; business obstacles

Creative Commons

BY-NC-ND 4.0