Towards socio-economic theory and practice of regulation. Evidence from OECD countries and Bangladesh

Abstract This research analyses the impact of social factors on regulation. The institutional theory of regulation introduces additional principles of independence, accountability and transparency of regulatory agencies, however, the paper argues here for further extending the theoretical and practical scope of regulation by introducing additional social aspects such as the notions of participation, inclusion, credibility and ultimately, legitimacy. The theoretical framework is illustrated with an empirical research conducted using the existing infrastructure regulatory indicators for energy, transport and telecommunications developed by OECD and specially reconstructed infrastructure regulatory indicators for Bangladesh for the period of 1975–2013 to exactly mirror the OECD’s infrastructure indicators datasets. The empirical results appear to generally support the theoretical assumptions made in this paper and the argument for broadening the theory and practice of regulation by inclusion of social factors in addition to economic and technical aspects of the traditional theory of regulation. More specifically, the impact on regulation of various social factors such as trust, access to information and absence of corruption, that promote participation, inclusion, sense of ownership and consequently and ultimately legitimacy appear to be statistically significant for both OECD countries and Bangladesh.


PUBLIC INTEREST STATEMENT
The diminishing role of the state in regulation brought about by creation of independent regulatory agencies has caused a removal of the citizens' oversight-that is tested at electionsaway from governments and has created a notion of professional accountability of regulatory agencies to replace such oversight. The socioeconomic theory of regulation is a response to the need to make this process more legitimate, credible and hence sustainable by re-introducing the notions of participation, inclusion and for pursuing a quest for legitimacy into all aspects of regulatory governance system, starting from its design of institutions, processes, policies and instruments. In the recent development paradigm shift, the regulation is increasingly being seen as the one that also has to support broader societal goals such as, quality of life and social cohesion.

Features of the contemporary theory and practice of regulation
This paper proposes a theoretical concept of socioeconomic theory of regulation and tests the theoretical assumptions for the need of inclusion of social factors, norms and informal institutions into the theory and practice of regulation with empirical analysis of OECD countries and Bangladesh as a developing country for the years between 1975 and 2013. Historically, the process of industrialization, the rise of modern capitalism and consequently the wide adoption of the Washington Consensus, gave rise and dominance to the deregulation paradigm. However, the last decades in an increasingly global world we see the return of regulation and rise of its significance in both economic and social areas with the key trends being allowing competition and private investment, and, creating markets through unbundling of the previously vertically integrated sectors of national economies. Mitnik (1980) defines regulation by independent agencies as state "within the frame of a law delegating certain decision-making regulatory powers to independent regulatory agencies and monitoring their application". Another defining feature is the proliferation of the number of independent regulatory agencies that have stripped away from governments important functions hence de-politicizing the role of the regulations.
Such a de-centered paradigm of regulation through independent regulatory agencies (IRA) has become a universal model that spread globally in the last several decades moving the concerned sectors outside parliamentary control and oversight since the decisions are made quickly and often at expert level as they demand specialized information and skills (Bevir & Richards, 2009). Such an autonomy-initially given to sectors such as electricity, gas, transport (economic regulation-Appendix 2)-has in an increasing manner spread to number of societal segments that it governs and regulates such as health and education (social regulation). These new modes of policy-making, including the importance of experts and trans-nationalization of policy-making, suggest that the democratic political control through elected officials is removed thus necessitating other forms of social control or oversight over the independent regulatory agencies.
These regulatory changes have been converging globally across many countries towards this proposed model based on the success of such transfer of regulatory powers from governments to independent agencies originally in the developed countries and then into developing countries too. As Moran (2001) explains, this shift reflects a fundamental alteration in the balance of state responsibilities such as movement away from attempting to manage the whole economy through command-and-control hierarchies towards disaggregated and loosely coordinated regulatory agencies. In addition to this Moran (2001) further links these developments to the rise of the liberalization, privatization and the rise of the New Public Management concept.
The traditional theories of regulation ranging from public interest to private interest theories and various in between sub-variations focused solely on the economic and political analysis of regulations as instruments for setting rules and norms for areas that affect both the economic and social life (Coase 1937;Peltzman, 1989;Stigler, 1971). This shortcoming is partly addressed by the consequently introduced institutional theory of regulation that enshrines into the framework the principles of independence and accountability of regulatory agencies (Gilardi, 2002;Levi-Faur, 2010;Majone, 1997;. Building on both the institutional theory of regulation and current practical experience of the regulation in both developed and developing countries, the recent thoughts and empirical research is converging towards attempting to capture the empirical and theoretical aspects of inclusion of additional key societal principles that are enabling good regulation such as participation, inclusion, credibility, and ultimately legitimacy. The current theoretical thinking and practical approaches to regulation already ventures beyond the basic principles of efficiency, regulatory independence, and accountability and into the wider area of participation, inclusion, legitimacy and credibility of both the content and processes of regulation (d 'Aquino, 2007;Zak & Knack, 2001). Such attempts have led to introducing social aspects notions into regulation embedding them in all its aspects from design and institutions to the processes and content of regulatory policy. This has given an impetus to conducting deeper analysis of the impact of social capital on regulation in order to explore what bestows the regulation with sustainability and credibility that are conducive to higher investment and higher output in infrastructure, beneficial for both businesses and citizens. The recent theoretical and empirical research have been focusing on the impact of social capital on particular aspects of development such as productivity, innovation, compliance at the firms' level (Granovetter, 2005) and on Weber's (1947) observation of the need for orders of public administration to be legitimate in order to be accepted by the key stakeholders. Important argument to the institutional economics theory for capturing social aspects, norms and informal institutions through social capital is made by Knowles (2005) who says that in terms of its definition and the arguments as to why social capital is likely to affect economic performance, social capital is a very similar concept to what North (1990) defines as informal institutions. This is important since it suggests that social capital can indeed be empirically modeled as a deep determinant of economic development conceptualized as a set of informally institutionalized expectations that other social actors will "reciprocate cooperative overtures, making otherwise uncooperative actors willing to undertake those overtures in the first place" (Boix & Posner, 1998).
The theoretical framework for inclusion of social factors in theory and practice of regulation is illustrated with an empirical research conducted using the existing infrastructure regulatory indicators for energy, transport and telecommunications developed by OECD and specially reconstructed infrastructure regulatory indicators for Bangladesh for the period of 1975-2013 to exactly mirror the OECD's infrastructure indicators datasets. The empirical analysis will test the argument for broadening the theory and practice of regulation by inclusion of social factors by analyzing the impact on regulation of trust, access to information, political activism, perception of safety, corruption that promote participation, inclusion, sense of ownership and consequently and ultimately legitimacy.

Shortcomings of the contemporary theory and practice of regulation
The need for establishing wider socio-economic foundations of the regulation stems from number of different reasons such as the one identified as a major challenge in producing a sustainable regulatory governance systems, that of legitimacy (Levi-Faur, 2010;Maggetti, 2009;Majone, 1997;. Such socio-economic approach to regulation should also solve the problems of regulatory capture by the elite that can be addressed by social participation as a balancing act against the risks of capture (Stigler, 1971). Another subtle shift that has taken place in the semantics from "government" to "governance" in itself marks a move away from the old command and control modes of "hierarchical, classic bureaucracy to a world of negotiation within, and between, selfsteering networks" (Black, 2001;Coffe & Geys, 2005;Coglianese, 2002Coglianese, , 2011Gilardi, 2002;Jordana, Levi-Faur, & Fernandez, 2011;Levi-Faur, 2010;Moran, 2001). Hence, if this is a move towards governance then this new paradigm has to also encompass the key issues of good governance such as participation, inclusion, and thus legitimacy.
Moran's explanation of regulatory model as being modernizing from "above" and based on specialized expert knowledge is too narrow and in order to address the need for obtaining legitimacy from "below" it should be broadened by introducing the requirement for participation and inclusion in regulation. The expert approach to regulation, meaning that such issues should be solved by experts since they are too complex for "common" citizens to comprehend, risks creating a new mystifying black-box of regulation that can backfire down the line if citizens do not feel included, consulted or being part of the process and of the monitoring and evaluation of the regulatory outcomes. The regulatory experts should indeed be able and should also have a duty to convey such complex notions in a clear and understandable manner to both professional and non-expert audiences in order to get the stakeholders fully informed, on board, consulted and included as integral part of the regulatory process.
The diminishing role of the state in regulation brought about by creation of independent regulatory agencies has caused a removal of the citizens' oversight-that is tested at elections-away from governments and has created a notion of professional accountability of regulatory agencies to replace such oversight. However this concept still lacks mechanisms for similarly testing its success beyond that of agencies being obliged to deliver results. The understanding of the systemic nature of such sustainable regulatory reforms explains why there is a progression in many countries towards more holistic concept of regulation.
The socio-economic theory of regulation is a response to the need to make this process more legitimate, credible and hence sustainable by re-introducing the notions of participation, inclusion and for pursuing a quest for legitimacy into all aspects of regulatory governance system, starting from its design of institutions, processes, policies and instruments. In the recent development paradigm shift, the regulation is increasingly being seen as the one that also has to support broader societal goals such as quality of life and social cohesion.

The rationale for a broadened socio-economic theory of regulation
This socio-economic model of regulation integrates the economic and regulatory theory with aspects from theories of sociology to account for social norms in addition to the traditional economic and technical handles that are now key components of the regulation. This extended model explains how existing regulation and regulatory compliance are expected to improve the regulatory outcomes by introducing the requirements of participation and inclusion in the theory and practice of regulation. The social contract for good regulation also necessitates going beyond the traditional command and control model in trying to promote regulatory compliance, legitimacy and acceptance in a complex interaction between stakeholders such as citizens, businesses, and, governments and agencies (Feld & Frey, 2007). This is even more important since if the citizens perceive their interests and preferences to be properly represented in the regulatory process and outcomes they will be more likely to legitimize them and comply with them thus increasing the chances for establishing and maintaining a sustainable and successful regulatory governance system. Or as Bird et al. (2007) argues, having societal institutions for better regulatory compliance represents having a meaningful voice of citizens in influencing the affairs of the state. Regulatory reforms are more likely to be accepted by citizens when the regulatory process is perceived to be not only fair but also inclusive and participatory making the regulatory policy outcomes legitimate. In such cases citizens perceive their acceptance of regulatory regimes as a "greater good" and comply with it even if they do not personally receive a full public good equivalent from their compliance.
Public choice theory and psychological theories also argue that legitimacy depends not only on the regulatory agency ability to provide favorable outcomes but that stakeholders place great importance on the processes it employs (Tyler, 1990). Tyler (1990) in his study finds that the perception of legitimacy and regulatory compliance is closely linked to people's views of the fairness of the procedures used and demonstrates that the people comply more with the law if the procedures (more important) and distributive justice (less important) employed by the regulatory authority are perceived to be fair. The effectiveness of the regulatory outcome is the extent to which regulation is implemented and an individual is made better off and the perceived distributive justice of the outcome is the perceived fairness of how the benefits or losses are shared in the society. In regards to the efficiency of the regulatory process, this involves the participation and inclusion through that people perceive that the regulatory authority is giving to them in the process of delivering on its mandate, and, how fairly regulatory agency treats citizens and businesses that are affected by the regulatory processes and outcomes. Tyler (1990) concludes that the positive outcomes matter less in promoting legitimacy than procedural justice and distributive justice.
The willingness to comply stemming from social norms is based on the perceived legitimacy of the regulatory agencies that are mandated to implement regulations and evidence suggests that a key determinant of perceived legitimacy is the principle of fairness and inclusion being built into the procedures for developing and implementing regulation. Hence regulatory authorities should determine what processes and practices are judged fair and inclusive by those segments of the population affected by the regulations. Additionally, the regulatory compliance is affected by legitimacy of regulation that is in turn impacted by the outcomes achieved and by processes employed in the regulatory governance. Once the regulatory system obtains legitimacy an individual compliance is expected to be affected by the behavior of others via the nature and extent of social influence exerted in the community that depends on the larger community's perception of the institution's legitimacy (Berg, 2000a, 200b;Sutinen & Kuperan, 1999;Young, 1979).

Social factors, informal institution and regulation
Social factors and informal institutions represented here by social capital promote better regulation by increasing the trust, participation and communication of policies hence improving regulatory compliance, acceptance and legitimacy. It does so through increased participation, inclusion and equity in regulatory processes and through improved channels of communication. Social capital represents a powerful additional explanatory variable of societies' abilities to regulate its mutual internal relations in a generally beneficial and sustainable manner. However, the theoretical underpinnings of how social capital improves regulatory effectiveness of outcomes and processes are still under-researched. This leaves us with a task to search for more explicit articulation of mechanisms by which the ability of people in society to cooperate affects the performance of their institutions and consequently regulation.
Why some states are able to produce more efficient and better regulation that is both economically beneficial and socially accepted as legitimate than other states? The links between social capital and regulation have not been extensively researched and only recently there have been studies conducted into various aspects of exploring whether such links exist, what are the channels and how these links can be captured and operationalized for practical research and policy purposes. Granovetter (2005) for example recommends opening of the economic black-box and introducing linkages between social phenomena and development. Examining the links between social capital and community governance Bowles and Gintis (2001) argue that realizing that market failures are the rule rather than the exception and that governments are neither smart enough nor good enough to make things entirely right cleared the intellectual stage for social capital's entry. The social capital notion brought an increased focus in policy and academic circles on peoples' values and not simply the utility functions of homo economicus.
One channel that explains the impact of social capital on regulation is that it facilitates cooperation that makes it possible for all actors to achieve superior social and economic outcomes. The notion of community seems to better captures the aspects of good governance and explains the popularity of social capital by focusing on what groups do rather than on what individuals have. Hence, social capital is generally interpreted as a degree of trust, cooperative norms and networks and/or associations within a society. Ostrom (1990) for example defines social capital as "an attribute of individuals and of their relationships that enhance their ability to solve collective-action problems" and defines three types of social capital for such collective actions: (1) trustworthiness, (2) networks, and (3) formal and informal rules or institutions and regards the second-generation collective-action theories as the organizing tool for social capital discourse. Bernstein (2015) researches the master agreements that nominally govern the transactions between Mid-Western original equipment manufacturers and their suppliers which are not designed to create legal obligations. The relational contracts between manufacturers and their suppliers are not informal and they rely for their effectiveness on numerous formal contract administration mechanisms that are expensive to create and administer. Nevertheless, because the non-legal sanctions that are built into these relationships can be imposed without filing a lawsuit and ending the contracting relationship, they create conditions that promote the growth of trust-based and relationship-specific social capital that can, over time, be used to bond ever more complex undertakings, concludes Bernstein (2015).
The social capital in Coleman's view (Coleman, 1988(Coleman, , 1990 is seen as structure of relations among people able to "facilitate exchange and build reciprocity and trust in ways that can reduce transaction costs through enhanced communication and information". The shortcoming of the current theory and practice of regulation is that it promotes the advantages of the trust-centered perspective of regulatory governance through "trust-to model" emphasizing the role of expertise and policy credibility in the context of government and business relations instead through the issue of legitimacy that the regulatory policies and processes earn in the eyes of the citizens, or "trust-from model". Such trust-centered explanations focused on social acceptance might be more persuasive than explanations centered on expertise and policy credibility when one considers their applicability to the creation of autonomous regulatory agencies (Levi-Faur, 2010). The level of trust in the society affects the design, implementation and continuation of regulatory reforms by reducing the enforcement and compliance cost and by bestowing legitimacy and social acceptance to such complex undertakings as regulatory reforms.
The social networks play two separate roles such as: transmitting the social custom from one person to another; and, helping individuals infer its views of certain policy partly from their own experience and partly by receiving information about the experiences of others. Posner (2000) defines social norm as equilibrium-signaling behavior such as when people shake hands, give money to charities, exchange gifts or engage in similar ritualized activities, they are sending signals that once started, tend to repeat themselves. Posner (2000) argues that if others conform to a socially accepted mode of behavior then the individual will also behave similarly. This means that individuals will comply with new regulations "as long as they believe that compliance is a social norm" (Alm, McClelland, & Schulze, 1999). Alm et al. (1999) analyze the effect of group communication and voting behaviour on tax compliance and find that communication combined with voting possibilities increases compliance and suggest that the social norm of tax compliance can be influenced by group communication. They claim that discussion in the society gives the opportunity to clarify benefits and costs of more expensive enforcement and increases the cooperation among group members.
Similarly, research into the public good notions indicates that the communication positively affects the cooperation (Keser & van Winden, 2000). Granovetter (2005) argues that novel information is more likely to flow through the weak ties than strong in his attempt to identify the channels through which social structure has an impact on development and economic outcomes. He reviews how the areas of labor markets, prices, productivity, compliance and innovation are impacted by social structures. Dessi and Ogilvie (2003) in their study of the merchant guilds as social networks that generated beneficial "social capital" through promoting and maintaining shared norms, find this to be effectively transmitting information and correspondingly enabling undertaking successful collective action. Such social capital benefited society as a whole by enabling the rulers to commit to providing a secure trading environment for alien merchants.
The dimensions of heterogeneity studied for their effect on social capital have included the race, ethnicity, income, education, religion, origin and mother language. Alesina, Baqir, and Easterly (1999) find that increased racial heterogeneity reduces local government's provision of core public goods such as roads, sewerage and education. Putnam (2000) using various number of pair-wise relations between the index of social capital and a number of important social and economic outcomes argues that social capital has powerful effects on the composite measure of educational performance (SAT scores, test scores, high school dropout rate); and, on the composite measure of child welfare (it includes teen pregnancy, infant mortality). Putnam (2000) also finds crime to be strongly negatively predicted by social capital with murder rates being lower in states where social capital is higher and also shows that variance in the percentage of tax evasion in the US states is strongly related to differences in social capital. In Chicago neighborhoods studied by Sampson, Morenoff, and Earls (1999) where neighbors express a high level of collective efficacy the violent crime was markedly lower illustrating the informal enforcement of community norms. Jankauskas and Šeputienė (2007) examining the relations between social capital (trust, social norms and networks), government effectiveness and economic performance for 23 European countries in their empirical investigations indicate positive correlation of two social capital dimensions, trust and networks, with economic performance indicators and governance indicators. In their analysis the civism dimensions do not correlate either with economic performance or with governance indicators. Ritter (2013) elaborates that since 1990s people are neglecting the volunteer work across all regions of Europe giving rise to self-centered and selfish behavioral manners as well as growing mistrust towards people. Using social capital indicators from survey conducted in EU in 2005 and good governance WGI indicators: Government Effectiveness, Rule of Law, Voice & Accountability, Regulatory Quality, Ritter's (2013) empirical analysis of 27 EU member states finds that higher level of social capital is associated with better governance.

Measurement of social capital
Measurement in the social sciences is in general a complex issue with the difficulties often rooted in the problem of measurement of social concepts and capturing complex multifaceted notions with a single or set of variables or indicators. Here is a brief overview of empirical research focused on capturing and measuring social capital. The measurements of social capital are often constructs that are both abstract and require subjective interpretation and translation into operational measures. This makes any operational measures of social capital inherently surrogates of their associated constructs.
One proposed step in defining social capital Cassidy (1997, 2001) is to unbundle the theory into its dimensions such as membership in formal groups, community participation, trust and fairness norms, crime and safety, political engagement etc., and measure each of them separately. These social capital dimensions are then illustrated by Cassidy (1997, 2001) with their observable variables of governance and political engagement, safety, empowerment for practical purposes of measuring of social capital. Krishna (2002Krishna ( , 2008 provides detailed review of several empirical studies where social capital has been used as an independent variable with social capital measures used in these studies, mostly based on household surveys, centered on membership of networks and norms. Evers (2003) also gives an overview of data collection methods used in this field of research by using surveys, official statistics or observations and lists the indicators of social capital used as networks, trust, civic engagement, norms and values. Evers (2003) argues that it is difficult to observe actual relations for connections and networks and suggests that researchers should develop other approaches focusing on the structure of social capital beyond the roles, rules and networks. Woolcock and Narayan (2000) review several innovative studies that have attempted to quantify social capital and its impact on economic development with number of measures employed such as trust, confidence in government, voting trends, social mobility, modern outlook and hours spent volunteering. They examine some of these measures evaluating their usefulness and their limitations for development theory and policy and conclude that there is no single "true" measure of social capital. Putnam (2000) also advocates is that no single source of data for quantifying social capital for the purposes of deriving measures that can be aggregated beyond the community level is flawless and proposed three criteria for selecting national level indicators such as comparability, continuity, and comprehensiveness. Hence he proposes what amounts to triangulation of data sources around any significant component of social capital. Catts and Ozga (2005) also suggests that multiple sources allow to search for congruity among outcomes across components of social capital to secure evidence of current status and more importantly, of change over time. If various indicators provide seemingly conflicting data, post hoc explanations of difference require judgment which should be documented to guard against rationalization or wishful thinking.
In an attempt to delineate social capital concepts at macro and micro level Krishna and Shrader (1999) present the conceptual framework in their Social Capital Assessment Tool. The macro-level refers to the institutional context in which organizations operate includes formal relationships and structures, such as the rule of law, legal frameworks, the political regime, the level of decentralization and the level of participation in the policy formulation process while the micro level includes cognitive and structural social capital (Bain & Hicks, 1998;North, 1990;Olson, 1982). Easterly (2006) predicts that societies with a lower initial inequality as proxied by larger share for the middle class and more linguistic homogeneity have higher social cohesion and thus better institutions that in turn lead to higher growth.
The tax compliance or tax evasion has also been identifies as a proxy of social capital. Slemrod (1998) argues that the social capital can be measured and derived from the willingness to "voluntarily" pay taxes lowers the cost of operating government and of equitably assigning its cost to citizens. He finds positive association between a country's tax to GDP ratio and the level of affluence for over 70 countries. Knack and Keefer (1997) also propose tax rule obedience as one dimension of civic cooperation as a social capital variable that exhibits a strong and statistically significant positive relationship on economic growth. Bjørnskov (2003) in his work "Corruption and Social Capital" finds the social capital impact on corruption to be robust to the inclusion of a number of other variables and suggests that it is possible to build social capital through investing in education, interest in society and some level of income redistribution, which in turn reduces corruption. He confirms two findings from previous studies: economic development is a strong cause of less corruption, as is a proper regulatory environment, measured by an index of civil liberties.
Other studies have measured social capital using indicators such as propensity to volunteer, vote, be members of organizations, give to charity, trust others, or support welfare programs (Alesina & La Ferrara, 2000;Clark & Kim, 2009;Foxton & Jones, 2011;Putnam, 1990Putnam, , 2007.

Measurement of regulation
Regulation here refers to government control of firms' decisions over price, quantity, entry and exit into markets such as whether governments get involved into setting rates for electricity service or impose restrictions on entry into markets, which are examples of government exerting control over firms' decisions and thus engaging in economic regulation. Regulation is the use of the government's power to coerce for the purpose of restricting the decisions of economic agents (Viscusi, Vernon, & Harrington, 1995). Hertogden (2010) sets a framework for analyzing an efficient regulatory management system, its governance, and its policies and outcomes. From the two of main forms of regulation: economic regulation as governing the infrastructure sectors that were previously viewed as natural monopolies (utilities, telecommunications, transport etc.) and markets with either imperfect or excessive competition with an aim to counter the effect of dominant firm or firms and social regulation governing sectors such as health, education etc. We will explore the social aspects such as social capital that affect the economic regulations outcome in terms of lending to the regulatory content and process credibility, legitimacy, hence sustainability.
The database for infrastructure regulation is the OECD ETRC infrastructure regulatory indicators sets for seven infrastructure sectors (Gas, Electricity, Airlines, Roads, Railways, Post and Telecoms) that are available and maintained by the OECD for its member states for the period between 1975 and 2013. Key characteristics (Table 1) that the OECD-style sector regulations capture are: (1) allowing competition by allowing entry, (2) creating equal playing ground for public and private investors, and, (3) creating markets by unbundling vertically integrated monopolies. OECD infrastructure regulatory indicators range on the scale from 0 (least regulatory restrictions in the sector) to 6 (most restrictions) and aggregate indicators for sectors are constructed from the sub-indicators that capture the following notions: Entry, Public/Private Ownership, Vertical Integration and Market Structure. Sub-indicators are measured and an aggregate indicator for sectors is constructed according to the weighting formulae (Appendix 2). Key characteristics that OECD-style infrastructure regulation indicators capture are: (1) allowing competition and creating equal playing ground for all investors, and, (2) creating markets by unbundling vertically integrated monopolies. The OECD infrastructure indicators while capturing regulatory policy, do not capture firstly, the regulatory processes and institutions, and secondly, they are also de jure indicators with lack of insight of the enforcement of the adopted policies which is particularly important for a developing country such as Bangladesh where enforcement of regulation is imperfect.
Due to their nature in the research these seven sectors are often combined in three general infrastructure sectors: Energy (Gas and Electricity), Transport (Airlines, Railways and Roads) and Telecommunications (Post and Telecom). There should be another distinction made that some of them are concerning with both the building and the usage/operation of the infrastructure (Energy, Telecommunications) while the Transport has a shortcoming of only analyzing the usage of existing transport infrastructure and not the building of the transport infrastructure itself.
For the purpose of extending this research beyond the developed OECD countries a same set of indicators have been uniquely reconstructed for Bangladesh for the infrastructure sectors (airlines, railways, roads, gas, electricity, telecommunications post) and for the same period between 1975 and 2013 in order to match the scope, depth and breadth of the current OECD infrastructure indicators. The OECD ETRC indicators are fully replicated for Bangladesh by collecting data from the primary sources such as ministries and agencies in charge of regulating infrastructure in Bangladesh.
In order to gain some insight from a historic perspective of how the infrastructure sector regulation developed over time and across countries we draw a timeline of how the ETRC infrastructure regulation indicators in these sectors developed worldwide. If we compare the curves for the same sectors Energy, Transport and Telecommunications for Bangladesh (Appendix 1: Figures 1-3) we see a pattern that is common for many developing countries of starting this infrastructure regulation simplification process at a much later stage but having much sharper curve in an obvious effort to catch up with developed world. The graphs highlight the fact that the United States was the first country to begin reforming infrastructure regulation in the 1970.
A number of other countries followed-notably the United Kingdom, Canada, New Zealand, while Japan and Sweden-commenced their reforms slightly later, starting from the late 1980s and early 1990s. In other European countries such as France and Italy the regulatory reforms in infrastructure began only in the mid 1990s. The regulatory improvements in Bangladesh in all three infrastructure sectors here are obvious however they are still lagging beyond those of the developed countries.
Compounding this problem is that even these regulatory improvements in Bangladesh are also to a significant extent de jure and to a much lesser extent fully enforced or implemented as yet in practice than it is the case in the OECD countries.

Regulatory environment in Bangladesh
Despite the continuous economic growth, the administrative and political apparatus of Bangladesh is still overtly bureaucratic, clientilistic, prone to capture by interest groups and marked by major discrepancies between the de jure and de facto regulatory environment due to not fully enforcing the adopted regulations. Institutions including the IRA are usually staffed by generalists with little interest in specializing in the area of their work beyond the mastering general administrative and bureaucratic procedures since that is knowledge that is replicable in other institutions where they will be rotated to. At the level of IRAs governing the infrastructure in Bangladesh, the reasons for desired outcomes in the infrastructure provision not being produced as yet are both of institutional and governance nature as well as of technical nature such as existing capacity and expertize levels.
The IRAs' institutional and governance level problems stem from the limited independence both from government and from special interests and lack of ability or political will to commercialize the regulated state-owned enterprises. This is compounded by the failure to provide de facto competitive environment and lack of property rights protection. Lack of adequately providing and considering the public feedback on regulatory decisions similarly leads to lack of credibility of the regulator and mistrust of regulatory reforms. At capacity level the issues hampering the IRA effectiveness are the limited resources, technical expertise, and capacity to fully use their regulatory powers. This is illustrated by lack of systematic data collection and analysis required for good quality regulatory decisions resulting in applying inappropriate benchmarks or standards, poor use of regulatory methodologies and instruments for pricing, tariff, penalties and existence of unjustifiable cross-subsidies. There is also need to better articulate the social and other non-economic obligations being imposed on regulated enterprises and to provide openness and transparency for the general public on important documents such as power purchase agreements.

Empricial analysis
In order to explore the validity of the theoretical concept of the socio-economic theory of regulation elaborated here, we build an empirical model that will test the impact of these social aspects on the states' ability to provide good regulation in the age of the de-centered mode of regulatory governance characterized by rise of independent regulatory agencies. The theoretical models of the impact of the social capital on regulation is illustrated here with an empirical research of the ability of states to provide quality regulation at national level, and, to regulate infrastructure sectors in particular in a manner that is conducive to efficient outcomes and processes of provision of goods and services, while at the same time being accepted by the citizens and businesses as legitimate. The social capital concept will be captured through number of instruments based on the conducted literature review and drawn from the relatively novel concept of the post-regulatory state theories of regulation that extend the focus of interest into participation, inclusiveness, and legitimacy.
Methodological issues of analyzing the impact social and economic concepts and institutional, governance and policy instruments are many and Kanbur (2005) identifies three types of problems with it: data and econometrics; variations around the average regression; and, interpretation of the results in terms of mechanisms. Rodrik, Subramanian, and Trebbi (2002) reviews econometric problems with such regressions such as parameter heterogeneity, outliers, omitted variables, model uncertainty, measurement error, endogeneity, etc. While reviewing the post-disaster social capital changes at community levels, Minamoto (2010) highlights the difficulties of capturing the cognitive social capital as opposed to the structural social capital. Similarly, the social capital is inherently easier to capture at micro and community level and is more difficult to do so at national level due to lack of measurements and indicators. In order to address these methodological concerns and based on recommendations that not a single variable that we want to study can capture social concepts at national level we will explore the use of several variables in order to be able to "triangulate" the results. Hypothesis that the empirical research will test is: Ho: Social factors have positive impact on states' regulatory and bureaucratic quality. To test our assumptions about the impact of social capital on regulation we build the following empirical model: Both IVs and DV will be logged variables with natural logarithm ln used which will mean that any % increase in IV will lead to intercept β % increase/decrease of DV.

Data and variables
Following a literature review of the empirical studies of social capital and their links with regulation we select data for analyzing its impact on states' ability to produce regulation that are conducive to development in general, and, for ability of government to produce regulation.

Dependent variables
3.1.1.1. Quality of regulations. ICRG indicator Bureaucratic Quality will be used to proxy for states' ability to provide good regulatory quality at national/macro level.

Infrastructure regulatory index. In our study OECD ETRC indicators are fully replicated for
Bangladesh for all infrastructure sectors for the period between 1975 and 2013 from the primary sources such as ministries and agencies in charge of regulating infrastructure in Bangladesh in order to match the scope, depth and breadth of the current OECD infrastructure indicators (Appendix 3).

Independent variables of social capital
Based on the review of empirical research from this study we find the following commonly used indicators of social capital at macro, national, level: trust, civic engagement, norms and values, trust, confidence in government, voting trends, initial inequality, tax compliance, murder rates, corruption, voice and law and order (Alesina & La Ferrara, 2000;Alm, Clark, & Leibel, 2011;Clark & Kim, 2009;Easterly, 2006;Evers, 2003;Knack, 2002;Knack & Keefer, 1997;Krishna, 2002Krishna, , 2008Minamoto, 2010;Putnam, 2000Putnam, , 2007Slemrod, 1998;Woolcock & Narayan, 2000). We select for this research the variables: INCTAX-voluntary tax compliance as Income tax collection-WDI database, POLACT-level of political activity voting participation as Election turnout-UID database, LAWORD-perception of safety and security-ICRG database, MEDIA-existence of quality information channels-Freedom House, CORR-level of corruption in society-ICRG database.

Control variables
GOVSPN, EDU and HEALTH-from WDI database.
Total Energy Sector Gas + Electricity = Entry + Public Ownership + Vertical Integration.
Total Transport Sector Rail, Air, Roads = Entry + Public Ownership + Vertical Integration.

Results and discussion
This empirical research tests the premises of the proposed broadened, socio-economic theory of regulation, and testing the arguments for inclusion of the impact of social factors on the ability of states to produce good regulation through independent regulatory agencies that have become a predominant mode of regulation globally in the last several decades. First the impact of the social capital on regulation in the infrastructure through REGINFRA indicators maintained by OECD and the same set specifically reconstructed for Bangladesh. Secondly, the ability of regulatory agencies to produce good regulatory outcomes at national level is proxied by the bureaucratic quality indicator from ICRG database. The empirical model is testing the assumption that participation and inclusion-facilitated by trust, access to information, willingness to engage in political processes, perception of safety and lack of corruption-in the regulation are key social factors that enable, promote and ensure good regulatory outcomes by providing the necessary legitimacy to the regulatory systems that are now independent and outside of governments mandate and control, hence outside of the political accountability through elections. In the following table we show the results of testing the channels through which social capital promotes participation and inclusion such as: trust in the society through the level of voluntary income tax payments, INCTAX; level of political activity through elections turnout; POLACT, existence of quality information channels, MEDIA; perception of safety and security, LAWORD, and, level of corruption in society, CORR. We use the direct year to year comparison of data. The VIF results are showing no multicolinearity in most of the different regressions. Results are robust to several sensitivity checks that we show further in this paper. Visual review of the scatter plot for each regression does not show presence of heteroskedasticity and further test such as Levene and Breusch-Pagan were performed and did not show presence of heteroskedasticity. Both IVs and DV logged variables. We use the direct year to year comparison of data. Endogeneity is addressed through performing additional 2SLS regressions with instrumental variables that confirmed the robustness of the original regressions. Both IVs and DV logged variables. Natural logarithm ln is used where 1% increase in IV leads to intercept %β increase/decrease of DV. The cross-section time-series (longitudinal) regression is performed for the 25 OECD countries and the results are given in Table 1.
The empirical analysis of impact of the social capital on regulation for OECD countries shows that access to information captured through MEDIA and absence of corruption captured through CORR as statistically significant and conducive to both regulatory and bureaucratic quality. And while the access to information-MEDIA has most immediate effect the impact of the absence of corruption-CORR on bureaucratic and regulatory quality shows both immediate and long term effect in a 3-year lag. In addition to this, POLACT do not show significant impact on regulatory quality in OECD while INCTAX and LAWORD show some statistically significant impact on BIRQUAL and on REGINFRA in either year on year and a 3-year lagged version. The results of the empirical research for developed OECD countries appear to suggest that citizens, businesses and government officials in OECD countries value the openness and absence of corruption most strongly.

Table 1. OECD-regulation, bureaucratic quality and social capital
Notes: The β standardized coefficients are reported and marked with (***), (**) and (*) marking statistical significance at The OLS linear regression is performed for Bangladesh and the results are given in Table 2.
The results of the empirical analysis of social capital impact on regulation for Bangladesh show that voluntary income tax payments INCTAX has the most statistically significant positive impact on both REGINFRA and BIRQUAL for all variations, year on year and in a 3-year lagged analysis.
Similarly to OECD countries, empirical research results for Bangladesh also show that access to information-MEDIA has an immediate impact on both regulatory and bureaucratic quality. However, and unlike it is the case with corruption in OECD countries in Bangladesh it is the trust in government captured through INCTAX that has the same strongest immediate and long-term impact in a 3-yearlagged version on both regulatory and bureaucratic quality. In addition to this, POLACT and CORR do not shows any impact on regulatory quality in Bangladesh while LAWORD shows statistically significant impact on BIRQUAL for year on year and 3-year lagged version. The robustness check with additional control variables of education, health and government spending added to the OLS regression model confirms the results of the original regression and Media and INCTAX as having highest statistically significant effect on bureaucratic and regulatory quality ( Table 3).
The results shed light on the channels through which social capital influences good regulatory outcomes such as the improved trust and access to information that is conducive to higher participation and inclusion in the regulatory processes hence lending the required legitimacy to the regulatory processes and outcomes. The results of the empirical research support the theoretical assumptions of the research based on arguments that social factors that generate trust are conducive to better regulation (Bernstein, 2015;Coleman, 1990;Ostrom, 1990;Ostrom & Ahn, 2008;Putnam, 2000). The results more specifically support the arguments that social factors channels such as access to information and communication are key channel in generating such trust (Granovetter, 1973(Granovetter, , 2005Keser & van Winden, 2000). The results also lend credence to the claim that trust in government either expressed by lack of corruption (Bjørnskov, 2003) or by willingness to voluntary pay taxes (Alm et al., 1999;Knack & Keefer, 1997;Slemrod, 1998), providing further justification for the theoretical assumptions about inclusion of social factors in the theory and practice of regulation.  While in general the research results confirm that social factors are conducive to good regulatory and bureaucratic quality in both OECD countries and in a developing country such as Bangladesh, digging deeper into the specific variables yields further useful insights. While for example access to information is equally important for both, the diverging outcomes on corruption in OECD and trust in government in Bangladesh are equally informative. The possible conclusion cannot be that the trust in government in OECD countries or corruption in Bangladesh are not as important, but that the specific conditions for both these notions in two different settings produce data that is obscuring the impact. The duty to pay income tax for example in OECD countries is deeply engrained and adhered to leading to constantly high tax collection levels that makes income tax collection levels weak as a variable for capturing subtle changes over years. Similarly, data on corruption in Bangladesh paints a picture of constantly high and deeply embedded corruption that it also renders it a constant that is not very suitable for capturing its impact on various other institutional, economic or social phenomena. This lends support to the body of research that argues that it is difficult to capture social factors through a one single measure and that a triangulation of variables and results is required to obtain more detailed insight (Catts & Ozga, 2005;Evers, 2003;Putnam, 2000;Woolcock & Narayan, 2000).
Another implication of the empirical results on theory of regulation is that policy makers and regulatory agencies should also devote greater effort to developing legitimacy. It is also important to note that empirical research results support the argument that procedural equity as captured by lack of corruption and transparent communication appear to be more important than outcome equity suggesting that regulation in this new socio-economic framework can still aim for economically efficient outcomes and not being forced to aim for "equal" allocation of resources. The fact that a regulatory institution is established does not necessarily bestow legitimacy on it but the legitimacy of the regulatory governance should be earned through both provision of quality outcomes and through ensuring participation and inclusion in the regulatory process. This can be done by employing more equitable procedures when imposing regulatory obligations on the parts of economic sectors which according to this theory should strengthen legitimacy and voluntary compliance. Regulatory regimes in which participants are empowered to play a prominent role in decision-making may be the means of achieving this aim of legitimacy (Hanna, 1995). This would address, in part, the need to incorporate procedural justice in the regulatory institutions and, similarly, in the enforcement policies and practices that need to be seen as fair by participants.
Such legitimate regulatory authority will ensure high degree of acceptance and compliance on which it can draw in times of crisis. A collapse of an infrastructure outcome or environmental calamity for example can be then swiftly and efficiently dealt with by a legitimate authority imposing significant short-term sacrifices on the participants if they view the authority as legitimate enough to feel a strong obligation to comply even when the dictates of the authority are contrary to their selfinterest. In this aspect the efficiency and equity are seen as complementary instead of substituting or competing with each other.

Conclusion
The aim of this research is to contribute towards identifying the factors beyond institutional, such as norms, social factors and informal institutions that facilitate establishment of a good regulatory regime in infrastructure that is conducive to both investment and infrastructure delivery in both developed and developing countries. By testing the hypothesis that social factors are important for creating and maintain good regulatory environment in infrastructure the empirical analysis of the impact on regulation of trust, access to information, perception of safety and lack of corruption that promote participation, inclusion, sense of ownership and consequently and ultimately legitimacy appear to generally support the theoretical assumptions made in this paper. The results also appear to confirm the assumption that the most important feature of such socio-economic theory of regulation is that a top-down, command and control regulatory governance, is not likely to be perceived as legitimate, resulting in a regulatory outcome that is ineffective in achieving its stated goals, costly to ensure regulatory compliance, and maybe even more importantly, has a potential to be rejected, or, to produce popular protest.
The global regulatory agenda already encompasses a variety of goals important to the economic, social, and environmental development, promoting (or introducing) accountability, transparency and participation of citizens in regulation. It is also becoming increasingly clear that the economic growth depends on a stable regulatory environment that can be maintained on a sustainable basis only if seen and perceived as legitimate by all actors. Lack of information, inclusion and participation undermine the effectiveness, credibility and legitimacy of regulation and diminish the respect for the law thus encouraging non-compliance. Although the emphasis on these social aspects of the regulatory policy varies both across countries and across its forms, it has nevertheless emerged as a strong and key feature of regulatory governance in most countries mainly through requirements for accountability and independence. Hence, if the regulation is to become a truly societal, the socioeconomic approach to theory and practice of regulation will have to go further and beyond accountability, independence and transparency requirements and should further introduce the requirements of participation, inclusion, engagement, and, ultimately, achieving legitimacy, credibility and sustainability of the regulatory systems.

Funding
The author received no direct funding for this research.

Citation information
Cite this article as: Towards socio-economic theory and practice of regulation. Evidence from OECD countries and Bangladesh, Goran Sumkoski, Cogent Social Sciences (2016), 2: 1254840.

Cover image
Source: Author.