EXPLAINING PARTICIPATION IN UNDECLARED WORK: A result of exit or exclusion?

Since the turn of the millennium, it has been widely recognised that undeclared work is a sizeable and expanding feature of the contemporary global economy (ILO 2002a ,b; Feige and Urban 2008; Schneider 2008; Charmes 2009; Jütting and Laiglesia 2009; Rodgers and Williams 2009). Indeed, a recent OECD report finds that out of a global working population of some 3 billion, nearly two-thirds (1.8 billion) are undeclared workers (Jütting and Laiglesia 2009). How, therefore, is this participation in undeclared work to be explained? The aim of this paper is to begin to evaluate critically the various competing explanations.

Firstly, therefore, the various rival explanations for participation in undeclared work are introduced, ranging from structuralists who read participation as driven by labourers' ‘exclusion’ from state benefits and the circuits of the modern economy (Sassen 1997; Gallin 2001; Hudson 2005; Davis 2006), to those depicting participation as driven by the decision to voluntarily ‘exit’ the formal economy (Cross 2000; Gerxhani 2004; Maloney 2004; Snyder 2004; Perry and Maloney 2007), including neo-liberals who portray undeclared workers as voluntarily opting out of the declared realm due to the costs of working legitimately and post-structuralists who depict this voluntary exit to be driven more by social and redistributive rationales than pure cost-benefit calculations. To evaluate critically these rival perspectives, the second section then introduces a survey conducted in Moscow during 2005/6 based on 313 face-to-face interviews. Reporting the findings, the third section will then display that participation is the result of neither solely exit nor purely exclusion but instead, how some is conducted due to exit, some due to exclusion and some for both reasons. The outcome in the concluding section will be a call for greater appreciation of the heterogeneous nature of undeclared work and a move beyond simplistic explanations and policy solutions.

Undeclared work throughout this paper will be defined as paid work that is not declared to the authorities for tax, social security and/or labour law purposes when it should be declared (European Commission 1998 ,2007; OECD 2002; Renooy et al. 2004; Schneider 2008). Indeed, there is a widespread consensus that this is the definition of undeclared work and therefore that the only difference between declared and undeclared work is that it is not declared for tax, social security or labour law purposes. If there are other differences, such as that the goods and/or services are illegal (e.g., drug-trafficking), for example, then this is not defined as undeclared work but ‘criminal’ activity. Of course, in practice, the boundaries between these other types of activity and undeclared work sometimes blur. Those in undeclared work and those engaged in criminal activities sometimes overlap in various spatial-temporal contexts.

To commence, the conventional but now largely refuted modernisation thesis will be reviewed followed by the structuralist, neo-liberal and post-structuralist perspectives that provide alternative competing explanations for participation in undeclared work.

For most of the twentieth century, undeclared work was portrayed as a leftover or residue from an earlier mode of production. As such, its persistence was seen as a sign of ‘under-development’, ‘traditionalism’ and ‘backwardness’ whilst the declared realm represented ‘progress’, ‘development’ and ‘advancement’ (Lewis 1959; Geertz 1963). Declared and undeclared work were therefore temporally ordered by depicting declared work as in the ascendancy and undeclared work as ‘the mere vestige of a disappearing past [or as] transitory or provisional’ (Latouche 1993: 49). Undeclared work was the traditional part of economies not yet absorbed and included in the modern economy. To have a relatively large undeclared realm, therefore, is to be underdeveloped, inferior, deficient and inadequate. Modernisation theory consequently reduced geographical diversity to a temporal sequence. It privileged declared work and positioned economies with large declared realms at the front of the development queue, and then benchmarked other economies with smaller declared realms as behind and lagging. The presence of undeclared work practices was consequently reduced to backwardness.

Recently, however, numerous studies have revealed that undeclared work is extensive, persistent and even growing relative to declared work in many populations (Schneider and Enste 2000; ILO 2002a ,b; OECD 2002; Feige and Urban 2008; Schneider 2008; Charmes 2009). The result is that explanations for the continuation and expansion of this sphere have been sought. Until now, two broad perspectives have been advocated. Either a structuralist perspective has been adopted which reads its persistence and growth as driven by labourers’ ‘exclusion’ from state benefits and the circuits of the modern economy, or its growth has been argued to be driven more by a voluntary decision to ‘exit’ the formal economy. Here, each is briefly reviewed in turn.

Since the late 1980s, a group of structuralist scholars have viewed the expansion of undeclared work to result from the emergence of a de-regulated open world economy and characterised it as unregulated, low paid and insecure work conducted under ‘sweatshop-like’ conditions by marginalised populations excluded from the formal labour market who conduct such work out of necessity due to no other options being open to them (Castells and Portes 1989; Sassen 1997; Gallin 2001; Amin et al. 2002; Hudson 2005; Davis 2006; Ahmad 2008). It is therefore deemed to be an inherent component of contemporary capitalism and the direct by-product of capital accumulation and restructuring within the global economy. As Fernandez-Kelly (2006: 18) puts it, ‘the informal economy is far from a vestige of earlier stages in economic development. Instead, informality is part and parcel of the processes of modernization’. Indeed, for Davis (2006: 186), such ‘primitive forms of exploitation … have been given new life by postmodern globalization’.

Undeclared work is therefore viewed to exist at the bottom of a hierarchy of types of employment and to be akin to ‘downgraded labour’ with undeclared workers receiving few benefits, low wages and with poor working conditions (e.g., Castells and Portes 1989; Sassen 1997; Gallin 2001). It is seen to be direct result of employers seeking to reduce costs, such as by sub-contracting to businesses employing undeclared labour under ‘sweatshop-like’ conditions (e.g., Bender 2004; Espenshade 2004; Hapke 2004; Ross 2004) and ‘precarious’ or ‘false’ self-employment. Undeclared work is also viewed to be a direct outcome of the demise of the intended full-employment/comprehensive formal welfare state regime characteristic of the Fordist and socialist era (Amin et al. 2002; Hudson 2005). In the new post-Fordist and post-socialist era, those of no use to capitalism are no longer maintained as a reserve army of labour and socially reproduced by the formal welfare state. Instead, they are off-loaded, resulting in their increasing reliance on undeclared work as a survival practice. Such work is thus extensive in marginalised populations where the declared realm is weak, since its role is to act as a substitute. It is undertaken by those involuntarily decanted into this realm and conducted out of necessity as a survival tactic (Castells and Portes 1989; Sassen 1997; Amin et al.2002). Undeclared workers, in sum, are portrayed as unwilling and unfortunate pawns within an exploitative global economic system which has made work more precarious and poorly paid.

The ascendancy of the neoliberal economic model during the 1980s moved the discussion on undeclared work away from the notion that it is a precarious form of waged work and last resort conducted out of economic necessity. For a group of neo-liberal commentators, the growth of undeclared work was an outcome of people deciding to voluntarily exit from the declared realm, rather than a product of involuntary exclusion (Cross 2000; Gerxhani 2004; Maloney 2004; Snyder 2004; Perry and Maloney 2007). As Gerxhani (2004: 274) puts it, workers ‘choose to participate in the informal economy because they find more autonomy, flexibility and freedom in this sector than in the formal one’. Such work is thus re-read as a form of dynamic entrepreneurship in which the self-employed voluntarily display an intrinsic preference due to the flexible hours, opportunity for economic independence, better wages, and avoidance of taxes and inefficient government regulation (Maloney 2004; Packard 2007).

For the neo-liberals, undeclared workers are cast as heroes throwing off the shackles of a burdensome state (e.g., Sauvy 1984; De Soto 1989) and undeclared work viewed as a direct result of over-regulation of the market (Minc 1982; Sauvy 1984; De Soto 1989). As De Soto (1989: 255) asserts, ‘the real problem is not so much informality as formality’. Undeclared workers are thus the last bastion of untrammelled enterprise culture in an over-regulated economic system, and the recent growth of undeclared work is evidence of a resurgence of the free market against state regulation. Undeclared work, in consequence, is the people's ‘spontaneous and creative response to the state's incapacity to satisfy the basic needs of the impoverished masses’ (De Soto 1989: xiv–xv).

Portraying undeclared work as undertaken largely by micro-entrepreneurs choosing to operate off-the-books in order to avoid the costs, time and effort of formal registration and in preference to declared employment (De Soto 1989, 2001; Small Business Council 2004; Cross and Morales 2007; Perry and Maloney 2007), this perspective which depicts participation in the undeclared realm as a universal voluntary choice (Biles 2008 ,2009) views such individuals as circumventing legal regulation and thereby establishing a ‘real’ free market. The main problem therefore is not undeclared work itself but rather the problems of operating in the declared realm that force individuals to escape to the undeclared sphere.

In recent years, an alternative explanation for the decision to ‘exit’ has emerged. This is inspired by a small stream of critical, post-colonial, post-structuralist, post-development and post-capitalist thought that has sought to transcend the conventional ‘thin’ portrayal of monetised exchange as universally market-like and profit-motivated and adopted ‘thicker’ portrayals that unravel the complex and messy characters and logics of monetised transactions (Davis 1992; Escobar 1995; Chakrabarty 2000; Bourdieu 2001; Leyshon et al. 2003; Zelizer 2005; Gibson-Graham 2006). Mirroring this in the field of undeclared work, a similar post-structuralist reading has been adopted which moves beyond a ‘thin’ depiction of undeclared work as market-like profit-motivated work, and undeclared workers as everywhere rational economic actors swayed by the cost/benefit ratios confronting them. Instead, it draws attention to how much undeclared work is conducted for and by kin, neighbours, friends and acquaintances and for reasons other than purely financial gain (Williams 2004; Williams and Windebank 2005; Persson and Malmer 2006; Smith and Stenning 2006; Round and Williams 2008; Jensen and Slack 2009; Nelson and Smith 2009). It also draws attention to undeclared work as a form of resistance in response to the exploitation of workers in the neo-liberal global economic system (Whitson 2007a ,b; Biles 2008 ,2009). For post-structuralists, undeclared work is thus again the result of exit rather than exclusion, but the decision to exit is more for social and redistributive rationales.

Until now, these competing explanations have been largely viewed as mutually exclusive. Most explanations see the development of undeclared work as taking place according to a single unique ‘logic’. Commentators have either advocated some single logic as universally valid and not even mentioned the others, or have paid lip-service to the others but contended that a particular logic predominates (De Soto 2001; Amin et al. 2002; Hudson 2005). Although a few commentators have begun to call for greater integration of these explanations, such as by arguing that ‘exclusion’ is more applicable to waged undeclared work and ‘exit’ to own-account informal workers (Chen 2004; Perry and Maloney 2007), or that exclusion is relevant to relatively deprived populations and exit to relatively affluent groups (Evans et al. 2006; Gurtoo and Williams 2009), few studies have sought to evaluate the wider validity of these more integrative explanations. In consequence, attention now turns to a survey conducted during 2005/6 in Moscow.

To evaluate critically the validity of these contrasting theorisations of undeclared work, a survey was conducted during late 2005 and early 2006 in Moscow, a global city of some 10.4 million inhabitants where previous surveys reveal a long legacy of the use of undeclared work practices as a livelihood practice (Clarke 2002; Caldwell 2004; Brooke 2006; Ledeneva 2006; Shevchenko 2009). As Clarke (2002) highlights, in the Soviet period, there were a wide range of services that no state enterprise provided and which were delivered privately by individuals to one another, such as small construction and decorating jobs, repair of radios, televisions and washing machines, clothing repairs, care for the elderly and sick, private transport and private tuition. Indeed, until 1987, doing such odd jobs for financial reward was illegal. This meant that undeclared work (the so-called ‘second economy’) was ‘illegitimate’ per se and reciprocity between kin, friends and neighbours became a necessary coping tactic (Ledeneva 1998). Great care should be taken, therefore, not to extrapolate the findings from this survey, such as on its size, to either Russia more generally, post-socialist societies or other regions of the world.

It is important at the outset, moreover, to provide a brief review of the context in which household employment decisions are made, in particular of Russian labour market legislation and the tax and benefits regimes. Superficially, the context in which household employment decisions are made appears quite simple. Taxation, for example, is on the surface extremely straightforward, with a uniform rate of 13 percent on personal income and a corporation tax level of 20 percent. Post-Soviet economies, however, are in practice rather more opaque. As Round et al. (2008) outline, many formal employees receive two wages from their formal employer, one formal and the other in cash at the end of the month. This allows employers to evade payroll taxes and state pension contributions, and makes wage payments very uncertain for employees as they have no guarantee that their unofficial wage will be paid. If it is not, they have no recompense because if they complain to the state, they will be open to accusations of tax evasion on previously paid informal payments. Labour legislation is also extremely problematic in Russia. Although there are a myriad of labour regulations aimed at protecting workers, few adhere to these regulations in practice (see Round et al. 2008). As Clarke (2000) displays, many people rely on contact networks to obtain employment, as opposed to applying for posts in open competition, and in some instances bribes are required to obtain employment (Round et al. 2008).

Those running small businesses face similar problems. Although tax is superficially simple, with a 20 percent corporate tax level, business owners can face significant problems when engaging with tax authorities. Often, they are asked to pay bribes to ensure a ‘light touch’ tax audit and a lower tax bill. This means that many decide to avoid contact with the state to reduce the number of informal payments requested. This extends not just to tax officials, but also the police and fire authorities, planners and bureaucrats involved in providing trade licences. Ironically, therefore, many operate on an undeclared basis to avoid the corruption, bribery and informality which pervades the declared realm. Reinforcing this, in a 2006 survey in response to the question, ‘what hinders economic growth in Russia?’, the most common response by 36 percent was ‘bureaucratic abuse, corruption and bribery in higher authorities’ (Yakovlev and Avraamova 2008: 264). Few small businesses are fully legitimate, therefore, as informality enables them to avoid the predatory nature of the state. This particularity of the Russian labour market and regulatory context, therefore, needs to be taken into account when considering whether the findings are more broadly applicable to other geographical contexts.

To gather data on the nature of the undeclared realm in Moscow, and given that previous studies in not only Moscow (Pavolvskaya 2004) but also Russia (Clarke 2002; Kim 2002; Pickup and White 2003), East-Central Europe (Wallace and Haerpfer 2002; Wallace and Latcheva 2006) and western nations (Renooy 1990; Leonard 1998) identify significant variations in undeclared work across affluent and deprived populations, maximum variation sampling was used to select three contrasting districts. Firstly, an affluent district in the west of Moscow okrug, namely Krylatskoe, was chosen, secondly, one of the most deprived districts located in the South-East okrug of Moscow, and third and finally, a mixed area to the west of the Moscow region, about 25 km out from the Moscow ring road (which is the Moscow city border), namely Leshkovo, where ‘new Russians’ live in large expensive houses alongside benefit-dependent pensioners and the ‘working poor’.

In each district, a spatially stratified sampling methodology was used to select households for interview. If there were some 1,000 households in the district and 100 interviews were sought, the researcher called at every 10th household. If there was no response and/or an interview was refused, then the 11th household was visited, then the 9th, 12th, 8th and so on. In practice, the response rate was 15 percent for the 10th household rising to 40 percent if the 9th and 11th households are included, 66 percent with the 8th and 12th household and 95 percent with the 7th and 13th household. This provided a spatially stratified sample of each district. In total, 313 face-to-face interviews were conducted in the three districts. Although these findings are not representative of Moscow as a whole, they do unravel the heterogeneous forms of undeclared work in this city and avoid the pitfall of analysing only one population segment which might engage in a narrow range of types of undeclared work.

Given how previous research indicates that respondents find it difficult to recall instances of undeclared work when relatively unstructured open-ended interviews are used, and that the resultant responses are often not comparable (e.g., Leonard 1998), the decision was taken to implement a structured interview. Besides gathering background data on gross household income, the employment status of household members, their employment histories, ages and gender, firstly, and drawing upon a question used in the New Democracies Barometer (NDB) surveys (Wallace and Haerpfer 2002; Wallace and Latcheva 2006), respondents were asked about the forms of work that they most rely on to maintain their living standard. Secondly, and using the now well-established ‘household work practices’ method (Wallace 2002) to analyse how households get-by and whether undeclared work is used, the sources of labour that the household last used to complete 25 common domestic services were analysed and why they had used that source of labour, thirdly, whether they had undertaken any of these 25 tasks for others (either on a paid or unpaid basis) during the past year and why they did so, and fourthly, a range of additional open-ended questions on undeclared work and its relative importance.

The interviews were conducted by a native Muscovite with a doctorate who had previously worked with the western researchers as a researcher on previous projects. Nevertheless, and given that Russian surveys are sometimes believed to be notorious for their level of fraud, checks were conducted on the interviewer. Whenever an interview was conducted, the interviewer had to note down the postal address and/or telephone number, not least so that we could contact them about follow-up in-depth interviews. The researchers did indeed later contact by telephone some 10 percent of the respondents (30 interviewees). No evidence of interviewer fraud was identified.

To analyse the rival explanations, firstly, the prevalence and importance of undeclared work will be evaluated; secondly, whether it is driven more by exclusion or exit along with how this differs across populations; and third and finally, the relationship between different types of undeclared work and different explanations. The outcome will be a more context-bound understanding of undeclared work that transcends the current universal hues.

Asking respondents what work is most important for securing their livelihood, along with what is second most important, Table 1 reveals that one in three (32 percent) primarily rely on undeclared work for their livelihood and a further one in nine (11 percent) assert that undeclared work is the second most important contributor. Some 43 percent thus cite this sphere as either the principal or secondary means of securing a livelihood. As such, undeclared work is not a residue or leftover. It is a core livelihood practice, as previously identified elsewhere in East-Central Europe (Wallace and Haerpfer 2002; Wallace and Latcheva 2006).

Who, therefore, relies on undeclared work? In both the modernisation and structuralist theories, it is marginalised populations excluded from declared work. However, Table 2 reveals that just 25 percent of working-age households with nobody in declared work cited undeclared work as their primary strategy, but 40 percent of households with one declared wage earner and 35 percent of households with multiple declared wage earners did so. Indeed, of all households chiefly relying on undeclared work, just 20 percent had nobody in declared work (22 percent of all households surveyed), 25 percent were multiple-earner households (35 percent of households surveyed) and 55 percent single-earner households (43 percent). To portray undeclared work as used by marginalised Muscovites excluded from the declared labour market is therefore erroneous.

Is the structuralist theory therefore not relevant in Moscow and undeclared work more a chosen alternative? To evaluate this, the rationales for participation will be analysed firstly by how they vary across populations and secondly, by how the rationales differ according to the type of undeclared work.

Do people engage in undeclared work due to their exclusion from declared work or because they are voluntarily exiting the declared realm either because of the problems associated with working legitimately (as neo-liberals assert) or for rationales grounded in alternative social and redistributive motivations (as post-structuralists assert)? To evaluate this, firstly, consumers' rationales are evaluated and secondly, suppliers' rationales.

Examining consumers' rationales, and as Table 3 displays, some 40 percent of undeclared work is used due to their lack of choice and/or exclusion from the declared realm, and a further 21 percent included this issue in their explanation. Other consumers, however, use the undeclared realm in preference to declared work. Neo-liberals assert that this is because of the problems associated with sourcing declared goods and services, including the availability, quality and speed of the provision of legitimate goods and services. In this Moscow survey, some 26 percent of undeclared work was sourced purely for this reason and a further 24 percent included this issue in their rationale. Third and finally, post-structuralists similarly assert that consumers use undeclared work in preference to declared work but explain this in terms of social or redistributive rationales. Some 8 percent of undeclared labour was primarily to help someone and/or as a favour amongst friends, colleagues and acquaintances. A further 11 percent included this issue in their explanation.

consumers' rationales, nevertheless, vary significantly across populations (see Table 3). Whilst ‘exclusion’ is more commonly cited amongst low-income households and those living in deprived neighbourhoods, ‘exit’ factors, particularly regarding poor formal provision, are more frequently cited amongst high-earners and those in affluent and rural areas. Social/redistributive rationales, meanwhile, are more frequently the chief rationale amongst higher-income households and affluent urban areas. Participation in undeclared work for low-income groups and those in deprived areas is therefore more a result of exclusion, whilst it is more due to voluntary exit amongst high-income groups and those living in affluent areas. In consequence, different explanations are relevant to different groups.

Turning to suppliers' rationales, Table 4 again differentiates those engaging in undeclared work according to whether it is primarily due to their exclusion from the declared realm (e.g., because they could not find a regular job, the employer insisted on them being paid off-the-books or because there is no alternative) or a decision to exit the declared realm and if so, whether this is due to problems faced working legitimately or for social/redistributive rationales. The finding is that 27 percent of undeclared work is explained purely in terms of exclusion from the declared realm, 50 percent solely due to a desire to voluntarily exit the declared sphere, as argued by neo-liberal and post-structuralist commentators, and 23 percent for a mix of both exit and exclusion rationales. Breaking down the decision to voluntarily exit the declared realm, 38 percent of all undeclared work was to avoid the costs, time and effort of formal registration and/or in preference to declared work, and 12 percent for social and/or redistributive reasons, such as a desire to help someone who is in need of help who would otherwise be unable to afford to get the work completed.

Again, however, the rationales vary across population groups and areas and these variations are statistically significant. Voluntarism, and therefore the neo-liberal and post-structuralist explanations, is more commonly cited by undeclared workers living in higher-income households and affluent localities. Exclusion from the legitimate realm, meanwhile, is more common amongst low-income households, and those living in deprived areas. It is not just across different socio-economic and spatial populations, however, that variations exist in the reasons for participating in undeclared work.

To evaluate how the rationales for participation varies by the type of undeclared work, the most usual approach would be to compare undeclared waged work and undeclared self-employment (ILO 2002a ,b; Neef 2002; Hussmanns 2005). However, this Moscow survey displays the need for a richer, more textured understanding that recognises a continuum of kinds ranging from, at one end, undeclared employees engaged in work close to formal employment to, at the other end, kinds more akin to unpaid community exchange. To start to unpack the varieties along this spectrum, three broad types are here distinguished that in practice blur into each other, namely undeclared salaried employment, undeclared self-employment and socially embedded undeclared work conducted for kin, friends and neighbours.

Of all undeclared work identified in this survey, some 25 percent is informal waged employment, 18 percent is informal self-employment, and 57 percent is own-account work conducted for closer social relations. Examining the reasons for participating in each of these different types of undeclared work, Table 5 reveals that those studying undeclared work from the perspective of exclusion are more likely to find what they seek in waged employment; some 51 percent is conducted for factors related to exclusion and 26 percent for rationales associated with both exit and exclusion. Similarly, the desire to exit the legitimate realm is greater in own-account work. Indeed, neo-liberals depicting participation more as a voluntary decision to exit the formal economy so as to avoid the costs, time and effort of formal registration, will find what they seek in undeclared self-employment (52 percent of such endeavour is conducted for this reason), whilst post-structuralists depicting engagement to be a choice conducted for social/redistributive rationales find this reason in own-account work embedded within closer social networks of familial and community solidarity (52 percent is conducted for such reasons).

It is important to be aware, however, that even if there is a statistically significant relationship between these different varieties of undeclared work and the contrasting explanations, it is far from the case that all work in each category can be explained exclusively in terms of either exclusion or exit. To see this, each variety of undeclared work is briefly reviewed in turn.

Over half (51 percent) of undeclared waged employees do so due to their exclusion from the declared realm, 23 percent for reasons related to a desire to exit the formal economy, and 26 percent for a mixture of both exclusion and exit factors. Undeclared work, therefore, is largely undertaken out of economic necessity and/or because they had no other choice, but not exclusively so.

Indeed, some forms of waged undeclared work are more likely to be associated with exclusion than others. To see this, we here analyse firstly, wholly undeclared waged work where the undeclared employee works without a formal contract and/or wholly off-the-books (5 percent of all reported undeclared work) and secondly, quasi-declared employment where declared employees receive two wages from their declared employer, an official declared wage and an additional unofficial non-declared (‘envelope’) wage (20 percent of all reported undeclared work).

Those engaged in wholly undeclared waged work are more likely to do so for reasons of exclusion (63 percent) or for reasons of both exit and exclusion (23 percent). Just 14 percent do so for exit reasons. This is reflected in the fact that 70 percent do not have declared jobs, and such undeclared waged work is what Pfau-Effinger (2009) terms a ‘poverty escape’ type in that it is their main source of income and largely done out of necessity. When conducted as what she terms the ‘moonlighting’ type, that is, on a temporary or part-time basis as a way of earning some additional income by those in declared jobs, then it is more often conducted primarily out of choice.

Those receiving envelope wages, meanwhile, are less likely to cite exclusion: 47 percent cite exclusion, 29 percent exit and 24 percent both exit and exclusion. Nevertheless, even if this wage arrangement is sometimes voluntarily entered into by employees, the majority felt that it had been imposed by employers seeking to avoid paying their full social insurance and tax liabilities, as well as potentially redundancy pay in that withdrawing it encourages employees to voluntarily leave. Most employees were against this arrangement since it affected their social security and pension entitlements as well as their ability to get credit and loans. This Moscow study thus reinforces earlier findings elsewhere that this wage practice is largely imposed on employees by employers (Sedlenieks 2003; Karpuskiene 2007; Woolfson 2007; Zabko and Rajevska 2007; Williams and Round 2009).

Some 18 percent of all undeclared work is conducted as self-employment. In stark contrast to undeclared waged work, just 20 percent is undertaken chiefly due to exclusion, 56 percent for exit reasons (particularly problems associated with operating in the formal realm) and 24 percent for both exit and exclusion reasons. However, this varies significantly across groups.

The undeclared self-employed are clustered at either end of the income spectrum with some 35 percent in the lowest quartile of households surveyed in terms of gross income and a further 40 percent in the highest quartile. Those undeclared self-employed in the lowest-income quartile largely have no declared job (90 percent of them) and conduct this work as a survival tactic. These necessity-oriented undeclared self-employed thus reflect the exclusion perspective that depicts undeclared workers as marginalised people off-loaded by capitalism and forced to eke out a living in this sphere as a last resort.

The undeclared self-employed in the highest-income quartile, however, usually engage in relatively well-paid undeclared self-employment which in three-quarters (70 percent) of cases directly arises out of opportunities from their declared work. This includes not only tradespersons, such as plumbers, electricians and builders, but also professionals such as lawyers who draw up contracts for non-declared payments. This work is largely voluntary and seen as a means of ‘topping up’ declared earnings. It is also a practice widely employed to avoid the ‘bribe taxes’, administrative corruption and ‘informal taxation’ in public life that pushes such enterprise and entrepreneurs into undeclared practices (Round and Kosterina 2005; Wallace and Latcheva 2006; Round et al. 2008). Indeed, for these self-employed, it is not so much undeclared enterprise that is in need of eradication, but rather, the bribes, corruption and informal taxation that are an inherent component of state institutions (Ledeneva 2006). Most of the higher-income self-employed running wholly illegitimate businesses, therefore, saw nothing wrong in working off-the-books. For them, it was the state and its officials who were corrupt and they who were in the right. They saw themselves as entrepreneurs and part of the burgeoning ‘capitalist’ culture but operating undeclared due to the corruption, bribery and bureaucracy inherent in the ‘old’ system they were replacing. They were willing to pay taxes but refused to do so whilst the income went into what they believe to be the pockets of state officials rather than used for the public good. For them, undeclared work is a chosen alternative to a harmful declared realm.

Previously, undeclared work conducted for close social relations such as kin, friends, neighbours and acquaintances was thought to be conducted purely for redistributive and/or social rationales (Williams 2004). This Moscow study, however, reveals that just 52 percent of such work was conducted for this reason. Instead, such work exists on a continuum ranging from own account work conducted for closer social relations mostly for exclusion rationales at one end, to work conducted mostly to help people out at the other end. The general rule is that the closer the social relations involved, the more the rationales are social and redistributive, whilst the more distant the social relations, the greater is the dominance of other rationales such as exclusion or the failure of the declared realm.

It should be noted, however, that kinship exchange has not been monetised. Instead, kinship exchange is either unpaid or, alternatively, gifts (but not money) are given and accepted, as Ledeneva (1998) has shown in her in-depth discussions of blat, networking and informal exchange. This socially embedded own account work, therefore, only applies to non-kinship exchange such as friends and neighbours. Indeed, some two-thirds of favours for friends and neighbours involve monetary payments, displaying that such paid favours are a key mechanism for the delivery of community self-help.

This Moscow survey reveals that different theories are valid for different types of undeclared work. When evaluating undeclared waged work, it is the structuralist reading of undeclared work as an inherent aspect of a new emergent form of capitalism that is most appropriate. Analysing undeclared self-employment, although the structuralist perspective is again relevant when examining necessity-oriented own-account workers in lower-income households, it is the neo-liberal thesis which is more appropriate when representing many engaged in undeclared self-employment in higher-income populations. In the realm of socially embedded undeclared work, meanwhile, it is the post-structuralist theorisation that is more valid.

Consequently, varying theories are appropriate for different types of undeclared work. No one theory fully captures the multifarious character of this sphere. By incorporating all of them, therefore, a finer-grained and more comprehensive understanding of the complex and diverse nature of undeclared work might be achieved. How, therefore, can these theories be coupled together to achieve this nuanced and multi-layered understanding? After all, superficially, these theories seem mutually exclusive. As this Moscow survey displays, however, they can be integrated by conceptualising firstly, a spectrum of types of undeclared work, ranging from undeclared work driven mostly by exit to undeclared work driven mostly by exclusion, with different combinations with varying emphases as one moves along the spectrum, and secondly, a spectrum ranging from undeclared work largely conducted as waged work through to undeclared work which is primarily own-account work (see Figure 1). Given that where each form of undeclared work sits in this grid might well vary from one socio-spatial context to another and from one time to another, Figure 1 depicts only the position of different types of undeclared work in contemporary Moscow.

Adopting this more integrative comprehension of the multifarious varieties of undeclared work helps stop the currently futile debates between those portraying these representations as competing theories despite each theory talking about very different forms of undeclared work (e.g., neo-liberals largely discuss undeclared self-employment while structuralists discuss undeclared waged work). It also enables a more nuanced approach towards tackling such work, helping identify those types that might be eradicated (e.g., low-paid sweatshop-like undeclared waged work), those that might be helped to make the transition into the formal economy (voluntarily chosen self-employment) and those that might be tacitly condoned (e.g., own-account socially embedded undeclared work), rather than adopting a ‘one size fits all’ approach which treats all forms the same. Perhaps most importantly, it might also enable the long-standing but simplistic debate about whether formalisation or informalisation is the route to progress and advancement to be replaced by finer-grained nuanced discussions about whether some forms of both undeclared work as well as declared work might be nurtured rather than all undeclared work or all declared work.

Evaluating the competing theories that variously read participation in undeclared work as driven by either exclusion from declared work (structuralism) or voluntary exit due to either the problems associated with declared work (neo-liberals) or more social/redistributive rationales (post-structuralist accounts), this paper reveals that in the Moscow districts surveyed, no one theory is universally valid. Although evidence can be found to support each and every explanation by looking at specific types of undeclared work and/or particular populations, no one theory accurately portrays the rationales for participation in all undeclared work and/or all populations. To more fully capture the diverse rationales for undeclared work, this paper therefore argues that only by incorporating all these theories will a finer-grained and more comprehensive context-bound understanding be achieved.

This Moscow study in consequence highlights the need to transcend a reliance on one theory over others. Now required are richer, more qualitative accounts of people's explanations for participation in order to tease out some of the more nuanced context-bound understandings of what leads different population groups to engage in undeclared work. Not least, the theoretical and methodological issue that ‘exit’ and ‘exclusion’ are not neat dichotomous terms needs to be unravelled along with how the meanings of ‘exit’ and ‘exclusion’ vary across population groups. For example, the opportunity structures within which an individual is operating his/her ‘choice’ to exit declared work will vary considerably across groups and this will be important to explore. So too will it be important to understand how the rationales for participation vary in other localities, regions and nations and how this influences the forms of undeclared work that are nurtured, eradicated and/or tacitly condoned in different places. At present, whether similar findings will prevail elsewhere is not known. The only comparable study is one conducted in Ukraine which displays similar findings to this Moscow study in terms of how rationales vary across population groups and forms of undeclared work (Williams and Round 2008). Whether this also applies elsewhere beyond these post-Soviet economies is as yet unknown.

In sum, this survey of undeclared work in Moscow has revealed that due to its heterogeneous character, the previous common tendency to privilege one particular theory and reject others needs to be transcended. To explain undeclared work, both exit and exclusion rationales need to be utilised, and it is suggested that only by appreciating how they all apply and become entwined to greater or lesser extents in different populations and types of undeclared work, that a more comprehensive understanding of this large sphere will be achieved. If this paper encourages such a re-theorising of undeclared work and encourages similar richer and more textured studies of the reasons for participation in undeclared work in different settings and populations, as well as more nuanced approaches towards how it might be dealt with, then it will have achieved its objective.

Colin C. Williams is Professor of Public Policy at the University of Sheffield. His research interests are in theorising the informal economy and devising public policy responses. Recent books include Informal Work in Developed Nations (2009, Routledge) and Rethinking the Future of Work (2007, Palgrave-Macmillan).

Dr John Round is Lecturer in Human Geography in the School of Geographical Sciences at the University of Birmingham, His interest are in postsocialism and transition processes.