Evaluating the Participation of the Unemployed in Undeclared Work: Evidence from a 27-nation European survey

How common is it for unemployed people to participate in undeclared work? Are unemployed people more likely to engage in undeclared work than those in declared jobs? Do they receive the same rewards as those in declared jobs from their undeclared work? And is the extent and nature of their participation in undeclared work everywhere the same? This paper seeks answers to these questions in the context of the 27 member states of the European Union (EU-27).

To do this, the first section will review the existing literature on the participation of the unemployed in undeclared work. This will display the prevalence of two dominant views: a ‘marginalisation’ perspective which holds that the unemployed disproportionately participate in and gain from undeclared work, and a ‘reinforcement’ perspective which argues that the unemployed benefit less from undeclared work than those in declared employment, meaning that undeclared work consolidates, rather than diminishes, the inequalities produced by the declared realm. To evaluate critically the validity of these rival perspectives in the EU-27, the second section will introduce the methodology used in a 2007 Eurobarometer survey of undeclared work comprising 26,659 face-to-face interviews in the EU-27 followed in the third section by the results. This will reveal that the marginalisation perspective is valid in Southern Europe and the reinforcement perspective in Nordic nations. In East-Central Europe and Continental Europe, as well as the EU-27 as a whole, however, the marginality and reinforcement perspectives are not mutually exclusive but co-exist with the participation of the unemployed in undeclared work representing a ‘reinforced marginalisation’ of their position. The final section then seeks tentative explanations for this situation.

At the outset, however, undeclared work needs to be defined, or what is sometimes called the ‘informal’, ‘shadow’, ‘underground’, ‘cash-in-hand’, ‘hidden’ or ‘off-the-books’ sector/economy (Williams 2004). Reflecting the widespread consensus, undeclared work is here 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 (OECD 2002; Williams 2004, 2006; Williams and Windebank 1995; European Commission 2007; Schneider 2008; Dekker et al. 2010). The only difference between declared and undeclared work, therefore, is that undeclared work is not declared for tax, social security or labour law purposes when it should be declared. If other differences exist, then it is not undeclared work. If the goods and/or services traded are illegal (e.g., illegal drugs) for example, then it is part of the wider ‘criminal’ economy, and if unpaid it is part of the separate unpaid economy. Of course, in practice, the boundaries between spheres, such as undeclared and criminal activities, sometimes overlap and blur, such as when in-kind favours or gifts are involved in exchanges. Consequently, only legal goods and services and only transactions involving strictly monetary exchanges are here defined as undeclared work. Finally, only wholly undeclared work is here analysed. Under-declared work where declared employees receive from their declared employer two wages, a declared wage and an undeclared (‘envelope’) wage (Williams 2009) is not explored, not least because the focus here is upon the participation of the unemployed in undeclared work.

Two contrasting views prevail regarding the participation of the unemployed in undeclared work, namely the marginalisation and reinforcement perspectives. Here, each is reviewed in turn.

The ‘marginalisation’ perspective holds that undeclared work is concentrated amongst marginalized populations such as the unemployed, who disproportionately participate in and gain from this realm (Gutmann 1978; Rosanvallon 1980; Parker 1982; Henry 1982; Castree et al. 2004; Katungi et al. 2006; Ahmad 2008).

In major part, this view results from a particular historical situation regarding the structure of employment in Europe. In France, for example, Rosanvallon (1980) asserted that high unemployment at the end of the 1970s had not resulted in the traditional French reaction of rioting in the streets because the unemployed, many of whom were not entitled to unemployment benefits at the time, were working undeclared. In the early 1980s recession in the UK, meanwhile, Parker (1982: 33) claimed that ‘with high unemployment more and more people are getting caught up in the web of the underground economy’, whilst Robson (1988: 55) asserted that ‘the informal economy is more feasible as an alternative prop to those who are out of work …’. Indeed, so dominant was this marginalisation perspective during the 1970s and 1980s that Pahl (1988: 249) claimed it was ‘in danger of becoming a social scientists’ folk myth’. Despite this warning, it has not abated. Throughout the 1990s, it continued to be expounded. As Stauffer (1995: 1) argued, ‘the informal sector can act as an important buffer against unemployment’, whilst Blair and Endres (1994: 288) asserted that ‘The role of the informal sector in providing a source of support for unemployed workers or individuals receiving public assistance is an important function of the unobserved sector’. Indeed, it has continued to be voiced in the twenty-first century (e.g., Office of the Deputy Prime Minister 2003; Castree et al. 2004; Davis 2006; Katungi et al. 2006; Ahmad 2008). The on-going belief, therefore, is that ‘a significant percentage of the officially unemployed are in reality working “off the books”’ (Gutmann 1978: 26).

In the EU-27, evidence to support this marginalisation perspective is found in surveys of Belfast (Leonard 1994), Brussels (Kesteloot and Meert 1999) and Romania (Stănculescu 2004). For example, Stănculescu (2004) finds that 27 percent of all Romanian households earn undeclared incomes, but 61 percent in households in which all members are non-employed. Whether this disproportionate participation of the jobless in undeclared work is more widely valid across the EU-27 has not been so far evaluated.

Over the past three decades, however, a reinforcement perspective has emerged that superficially appears to argue the inverse, namely that unemployed people benefit less from undeclared work than those in declared jobs and that the undeclared economy therefore reinforces, rather than reduces, the inequalities produced by the declared economy.

Across the EU-27, support is found in small-scale studies which reveal higher participation rates in undeclared work amongst the employed than unemployed, and how only a very small proportion of undeclared workers are regular ‘working claimants’. This has been shown in France (Foudi et al. 1982), Germany (Heinckel et al. 2008; Pfau-Effinger and Sakac Magdalenic 2010), Greece (Hadjimichalis and Vaiou 1989), Italy (Mingione 1991), Lithuania (Krumplyte and Samulevicus 2010), the Netherlands (Van Geuns et al. 1987), Poland (Surdej 2005; Surdej and Slezak 2009), Portugal (Lobo 1990b), Romania (Parlevliet and Xenogiani 2008), Spain (Benton 1990; Lobo 1990a), the UK (Pahl 1984; Morris 1994; Williams 2001, 2004, 2006; Williams and Windebank 2005) and Sweden (Persson and Malmer 2006).

For example, in Sweden, whilst 13 percent of the population engage in undeclared work, 9 percent of the registered unemployed do so (Persson and Malmer 2006). Pedersen (2003), meanwhile, identifies that the unemployed constitute only a small proportion of the undeclared workforce: 20.7 percent in Germany, 9.9 percent in Denmark and 9.2 percent in the UK. In East-Central Europe, similarly, empirical evidence widely supports this reinforcement perspective (Wallace and Haerpfer 2002; Wallace et al. 2004; Wallace and Latcheva 2006; Rodgers and Williams 2009; Williams and Round 2009). Wallace et al. (2004) identify that just 5 percent of undeclared workers in East-Central Europe are unemployed people receiving benefits, and Surdej and Slezak (2009) in Poland that the unemployed represent just 16 percent of undeclared workers. In Spain, meanwhile, Benton (1990) identifies that 5 percent of undeclared workers claimed social security benefit, although this was at a time in Spain when unemployment benefits were not widely available to the non-employed.

It has been also widely shown that the unemployed receive lower earnings from undeclared work than the employed (Mattera 1985; Renooy 1990; Fortin et al. 1996). A segmented undeclared labour market has been therefore identified in which those in declared jobs engage in relatively well-paid undeclared work and unemployed people in lower-paid undeclared work (Lobo 1990a,b; Lemieux et al. 1994; Fortin et al. 1996; Williams and Windebank 1998; Neef 2004). In a survey of English localities for example, Williams (2004) finds that the average hourly wage rate of the registered unemployed (£2.40) is 54 percent lower than that of the employed (£5.15).

To explain this relative lack of participation of unemployed people in undeclared work compared with those in declared jobs, and their lower gains from the undeclared realm, adherents to this reinforcement perspective commonly cite four reasons:

• the unemployed lack the resources (e.g., car, tools) necessary to engage in a wide range of undeclared work (Pahl 1984; Williams 2004);

• they receive fewer opportunities to engage in undeclared work (e.g., Morris 1995; Komter 1996; Williams 2006);

• the unemployed lack the skills and competencies to conduct better paid undeclared work (Renooy 1990; Mingione 1991; Fortin et al. 1996). If their skills and competencies are inappropriate for finding declared work, there is no reason to believe that they are appropriate for finding undeclared work; and

• the unemployed fear being reported to the authorities, not least because claiming welfare benefits is popularly considered a more serious offence than tax evasion (Cook 1997; Williams 2004).

In sum, two apparently mutually exclusive perspectives exist, one of which argues that the unemployed disproportionately conduct undeclared work and the other that those in declared jobs disproportionately benefit from undeclared work. Until now, the only evidence supporting each of them is from small-scale studies of particular localities, regions or nations. No evaluation of these contrasting perspectives has been conducted on a pan-European level. Below, therefore, we start to fill this gap.

In May and June 2007, survey was conducted of undeclared work in the EU-27. Special Eurobarometer No. 284 (‘Undeclared work in the European Union’), which one of the author's of this paper helped design, involved 26,659 face-to-face interviews in the 27 member states of the European Union (EU), ranging from 500 in smaller member states to 1,500+ interviews in larger countries.

In all nations, a multi-stage random (probability) sampling method was applied. A number of sampling points were drawn with probability proportional to population size (for total coverage of the country) and to population density according to the Eurostats NUTS II (or equivalent) and the distribution of the resident population in terms of metropolitan, urban and rural areas. In each of the selected sampling units, a starting address was then drawn at random. Further addresses (every nth address) were subsequently selected by standard ‘random route’ procedures from the initial address. In each household, meanwhile, the respondent was drawn at random (following the ‘closest birthday rule’). All interviews were conducted face-to-face in people's homes and in the appropriate national language with adults aged 15 years and over. Data was collated using CAPI (computer assisted personal interview) where this was available and then loaded onto a statistical package (namely SPSS) for analysis purposes.

The face-to-face interview schedule adopted a gradual approach to the more sensitive questions. This firstly asked questions about the respondents' attitudes towards undeclared work and having established some rapport, the second section then asked questions regarding their purchase of goods and services on an undeclared basis in the last 12 months along with their reasons for doing so and thirdly, questions regarding their supply of undeclared work including the type of work conducted, hours spent, the average remuneration per hour, who they did the work for and their reasons for doing so. Socio-demographic data was collected on their current occupational status, gross formal income, age, gender, education, hours worked in formal employment and so forth. Here, the focus is upon the 1,507 respondents (6 percent of the surveyed population) who reported their employment status as unemployed.

Before discussing the findings, the validity and reliability of the results needs to be addressed. The mean time that the interviews lasted with the 1,507 unemployed respondents was 45 minutes and these respondents were not generally viewed by the interviewers as hesitant about discussing undeclared work. Respondent cooperation was deemed excellent in 57 percent of cases, fair in 36 percent and average in 7 percent. In only 7 cases (0.4 percent) was cooperation deemed bad by the interviewer. Below, therefore, the results are analysed.

The 26,665 respondents were asked to name the population group in their country most likely to undertake undeclared work. Table 1 reveals that popular opinion is strongly supportive of the marginalisation perspective, with 41 percent citing the unemployed as the group most likely to conduct undeclared work followed by 23 percent citing illegal immigrants.

However, popular opinion on who does undeclared work is not everywhere the same. For analytical purposes, the 27 member states surveyed are here grouped into four geographical regions, namely: Continental Europe, UK and Ireland (Belgium, Germany, France, Ireland, Luxembourg, the Netherlands, Austria and the UK); Eastern and Central Europe (Bulgaria, Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Romania, Slovenia and Slovakia); Southern Europe (Cyprus, Greece, Spain, Italy, Malta and Portugal), and the Nordic countries (Denmark, Finland and Sweden). Popular support for the marginalisation perspective was more extensive in East-Central Europe (where 56 percent cited the unemployed as the group most likely to engage in undeclared work) and Continental Europe (44 percent) than in Southern Europe (26 percent) and the Nordic nations (28 percent).

To what extent, therefore, does popular opinion reflect the lived practices? As Table 2 displays, whilst 4 percent of respondents reported conducting undeclared work in the last 12 months, this figure was 9 percent of those defining their employment status as unemployed and this is a statistically significant difference. The popular belief that the unemployed disproportionately participate in undeclared work therefore appears validated.

However, this does not refute the reinforcement perspective. Although the unemployed have higher participation rates than the employed, they only undertake a small proportion (11 percent) of all undeclared work across the EU-27 and earn 57 percent less per hour from undeclared work (€8.04) than those in formal jobs who work undeclared (€14.08). The unemployed therefore earn only 9 percent of the gross undeclared income in the EU-27, and this is due to the longer hours they spend working undeclared (98 hours on average per annum compared with 71 hours for the wider population).

Those in declared jobs, meanwhile, although 50 percent of the surveyed population, conduct 57 percent of all undeclared work, receive 59 percent of total undeclared income and earn both more per hour and have higher annual undeclared incomes. This supports at the EU-27 level the reinforcement perspective that the undeclared economy reinforces, rather than reduces, the inequalities produced by the declared economy. Consequently, the marginality thesis and reinforcement thesis are not mutually exclusive. Although the unemployed disproportionately engage in undeclared work, the employed benefit more from the undeclared economy, meaning that undeclared work reinforces, rather than reduces, the inequalities produced by the declared economy.

Do all groups of the unemployed, however, disproportionately participate in undeclared work but benefit less than the employed from working undeclared? Table 3 reveals that all the unemployed sub-groups groups surveyed disproportionately conduct undeclared work, meaning that the marginalisation perspective applies to all groups of the unemployed. Nevertheless, some unemployed sub-groups have higher participation rates than others, namely men, those aged 40–54 years old, those whose education ended between 16 and 19 years old and those living in rural areas.

Examining whether all groups of the unemployed benefit less from undeclared work than those in declared employment, this is also supported, although some benefit less than others. Unemployed women, for example, participate less in undeclared work than unemployed men and earn just 74 percent the hourly wage rate of men. Similarly, the younger unemployed, the unemployed with lower levels of education and those living in larger urban areas both participate less and earn lower hourly earnings. Moreover, no sub-groups of the unemployed match the hourly undeclared wage rates of those in declared jobs (€14.08) and only one unemployed sub-group exceeds the hourly undeclared wage rate of manual workers (€11.36), namely those unemployed aged over 55 years old. One can therefore identify a segmented undeclared labour market amongst the unemployed in terms of hourly wage rates and total undeclared incomes which has little overlap with the segmented undeclared labour market that exists amongst the employed. Whichever unemployed group is analysed across the EU-27, not only is the marginalisation perspective valid but so too is the reinforcement perspective that undeclared work benefits the employed more than the unemployed.

Table 4 evaluates whether the participation of the unemployed in undeclared work varies across different EU regions. Before doing so, however, the apparently counter-intuitive finding that a smaller proportion of the population participate in undeclared work in Southern Europe than in the EU-27 in general and Nordic nations more particularly, requires explanation. Just because participation rates are lower in Southern Europe does not mean that the undeclared economy is smaller. In Southern Europe, although a smaller portion of the population engage in undeclared work, those doing so are more likely to undertake undeclared waged employment, work longer hours, and the total amount of undeclared work conducted and undeclared income earned per capita is higher than in the EU-27 as a whole and Nordic nations in particular (Williams and Windebank 2011). Meanwhile, participation rates are higher in Nordic nations but more undeclared work is composed of small odd-jobs conducted for and by kin, friends, neighbours and acquaintances on an own-account basis. The outcome is that despite a lower participation rate in undeclared work in Southern Europe, the size of the undeclared economy relative to the declared economy is larger than in Nordic nations.

Indeed, participation rates are lower in Southern Europe largely because of the non-monetization of community and kinship exchange. In Nordic nations, meanwhile, there appears to have been a monetization of work conducted for closer social relations. Money changes hands either to avoid the need for reciprocity at a later date, or to redistribute money in a manner where there is no connotation of ‘charity’ since the money is being given for work undertaken. This has previously been identified in smaller-scale studies in northern Europe (Boren 2003; Williams, 2004; Persson and Malmer, 2006). In Southern Europe, however, this familial and community solidarity work remains largely unpaid. The lower participation rates in undeclared work in Southern Europe compared with the EU-27 and Nordic nations, therefore, are not perhaps as counter-intuitive as they first appear.

How, therefore, does the participation of the unemployed in undeclared work vary across EU regions? As Table 4 displays, the marginalisation perspective is valid in Continental Europe, East-Central Europe and Southern Europe. However, this is not the case in Nordic nations; only 9 percent of the unemployed participate in undeclared work compared with 12 percent of the employed and 20 percent of the other non-employed. Indeed, although the unemployed represent 4 percent of the surveyed population, they conduct just 3 percent of all undeclared work and receive just 2 percent of total undeclared income.

In Nordic nations, therefore, solely the reinforcement perspective is valid. Although only 57 percent of the surveyed population, the employed conduct 63 percent of all undeclared work, earn 65 percent of total undeclared income, receive 28 percent more per hour from their undeclared work than the unemployed and their annual earnings are 13 percent higher. In Southern Europe, in contrast, the reinforcement perspective is refuted. Despite being 50 percent of the surveyed population, the employed conduct just 44 percent of undeclared work and earn just 43 percent of the undeclared income.

If only the marginalisation perspective is valid in Southern Europe and only the reinforcement perspective in Nordic nations, in Continental Europe and East-Central Europe, and the EU-27 as a whole, both the marginality and reinforcement perspectives are applicable. Although the unemployed have higher participation rates in undeclared work, they benefit less from such work than the employed. For example, in Continental Europe, despite being only 51 percent of the surveyed population, the employed conduct a disproportionate share of the undeclared work (57 percent), earn 63 percent of all undeclared income, 50 percent more per hour and 70 percent more in total undeclared earnings than the unemployed.

In sum, the marginalisation perspective is not valid in Nordic nations but it is in Continental Europe, East-Central Europe and Southern Europe, whilst the reinforcement thesis is valid in Continental Europe, East-Central Europe and the Nordic nations, but not in Southern Europe. In other words, in Nordic nations solely the reinforcement perspective applies whilst in Southern Europe only the marginalisation perspective is applicable. In Continental Europe and East-Central Europe, meanwhile, both the reinforcement and marginalisation perspectives are applicable. In these EU regions, similar to the EU-27 as a whole, a new perspective regarding the participation of the unemployed in undeclared work is required that depicts how undeclared work leads to a ‘reinforced marginalisation’ of their position. The unemployed disproportionately engage in undeclared work but their participation in this realm reinforces their marginalised position, earning lower wage rates and total annual undeclared incomes than the employed. This reinforced marginalisation becomes even more accentuated amongst certain sub-groups of unemployed; those disadvantaged in the declared economy (e.g., women) are in general most disadvantaged in the undeclared economy.

To explain the participation of the unemployed in undeclared work, the two issues that need to be analysed are firstly, why the unemployed disproportionately participate in undeclared work in the EU-27 as a whole but not in some regions and secondly, why when the unemployed participate in undeclared work they generally, but not always, benefit less than those in declared jobs.

One tentative explanation for the higher participation rates of the unemployed in undeclared work is that unlike many employed, they are forced into undeclared work in order to secure a livelihood, especially in EU regions where access to unemployment compensation benefits is not universal and/or the level of benefit is comparatively low relative to the average wage. Indeed, this survey provides tentative support for this explanation. The most common reason for the unemployed conducting undeclared work in all EU regions was that they could not find a regular job, except in Nordic nations where it was more a matter of choice, such as because they believed that both parties benefited from it. This intimates that the unemployed only disproportionately engage in undeclared work in welfare regimes with lower benefit recipient quotas and/or where the level of benefit payments is relative low compared with the average wage (Reissert 1994). In regimes where unemployment compensation benefits are more widespread and/or the level of benefit payments is relatively closer to the average wage, and therefore alternative survival mechanisms exist, unemployed people do not disproportionately undertake undeclared work.

Future research will need to evaluate this tentative explanation. Is it the case that although the unemployed generally engage in undeclared work out of necessity in order to secure a livelihood, in member states where benefit recipient quotas and/or benefit levels are higher, they do so more out of choice? This will require detailed analysis of firstly, whether benefit recipient and payment levels are insufficient to secure a livelihood in southern Europe (where the marginalisation thesis prevails) but sufficient in Nordic nations and secondly, in-depth qualitative research with the unemployed on the degree to which they conduct undeclared work out of necessity in southern Europe and out of choice in Nordic nations, as well as what they mean when they perceive themselves as ‘choosing’ to work undeclared (see Williams and Round 2010).

Why is it the case, however, that the unemployed benefit less from undeclared work than those in declared jobs? Evidence collected in this survey supports the earlier discussed cocktail of reasons, namely that those in declared jobs have fewer fears of being detected and fined and greater access to the resources, opportunities and skills to engage in undeclared work.

Firstly, to evaluate whether the unemployed feel more inhibited about engaging in undeclared work for fear of being detected and having their benefits curtailed, respondents were asked ‘people who work without declaring their income risk that tax or social security institutions find out and issue supplementary tax bills and perhaps fines. How would you rate the risk of being detected in [your country]?’. Across the EU-27, 44 percent of the unemployed rate the risk very high or fairly high compared with 37 percent of all respondents. Moreover, in EU regions where the unemployed more widely rated the perceived risk of detection as high or fairly high (55 percent in Nordic nations compared with 24 percent in Southern Europe), participation by the unemployed in undeclared work was lower.

Secondly, analysing whether the unemployed lack access to various resources which are a prerequisite for conducting some undeclared work, across the EU-27, just 51 percent of the unemployed have a car (compared with 68 percent of all respondents), 53 percent a computer (60 percent of all respondents) and 36 percent an internet connection at home (48 percent). However, the unemployed engaged in undeclared work are more likely to possess these resources (68 percent have a car, 71 percent have a computer and 59 percent an internet connection), intimating that those with such resources do perhaps find it easier to conduct undeclared work.

Thirdly, and examining whether the unemployed have fewer opportunities to engage in undeclared work, not only do they not hear about opportunities for undeclared work through their jobs since they are jobless, but this survey also finds that they engage in less undeclared work for closer social relations, such as kin, friends, neighbours and acquaintances, which constitutes 58 percent of all undeclared work in the EU, but just 48 percent amongst the unemployed. This suggests that joblessness leads to a narrowing of the possibilities for conducting undeclared work for even closer social relations.

Fourth and finally, and analysing whether the unemployed have fewer skills and competencies that potential employers might seek on an undeclared basis, this survey provides evidence to support this explanation. In the EU-27, although 19 percent of the unemployed stopped their education before 15 years old (compared with 28 percent of all respondents), 60 percent ended their education between 16 and 19 years old (46 percent of the whole sample) and just 21 percent at 20 years old or older (26 percent). Hence, they are less qualified than the wider population. The unemployed working undeclared, however, have spent more years in education with some 29 percent of unemployed undeclared workers ending their education at 20 years old or older, but 39 percent of all those in declared jobs working undeclared. The implication is that unemployed people with higher levels of education participate more in undeclared work but the unemployed working undeclared still have fewer years in education than those in formal jobs working undeclared.

This paper has evaluated critically two competing perspectives regarding the participation of the unemployed in undeclared work, namely the ‘marginalisation’ perspective which holds that the unemployed disproportionately participate in and gain from undeclared work, and the ‘reinforcement’ perspective which holds that the unemployed benefit less from undeclared work than those in declared employment. Reporting the results of a 2007 Eurobarometer survey on undeclared work comprising 26,659 face-to-face interviews conducted across the EU-27, including 1,507 unemployed people, the finding is that the marginalisation perspective is valid in Southern Europe and the reinforcement perspective in Nordic nations. In East-Central Europe and Continental Europe, as well as the EU-27 as a whole, however, the unemployed are more likely to participate in undeclared work, but they receive significantly lower earnings and gain less from undeclared work than those working undeclared who are in declared jobs. The outcome is a new ‘reinforced marginalisation’ perspective which holds that undeclared work reinforces the marginalised position of the unemployed relative to the employed.

The paper then tentatively explains both the greater participation of the unemployed in undeclared work in the EU-27 and how this varies across EU regions, as well as why when they conduct undeclared work they generally benefit less than those in formal jobs. The higher participation rates of the unemployed in undeclared work is tentatively explained as being because in welfare regimes where benefit compensation payments are less widespread and/or benefit payments relatively low compared with the average wage, the unemployed are forced into undeclared work to secure a livelihood. Meanwhile, in regimes where unemployment compensation benefits are more widespread and/or the level of benefit payments is relatively closer to the average wage, unemployed people do not disproportionately engage in undeclared work and it is more a matter of choice.

To evaluate this, future research will need to analyse firstly, whether benefit recipient and welfare payment levels are less sufficient to secure a livelihood in southern Europe (where the marginalisation thesis prevails) than in Nordic nations and secondly, conduct in-depth qualitative research with unemployed participants on whether they undertake undeclared work more out of necessity in southern Europe and choice in Nordic nations as well as what workers mean when they perceive themselves as ‘choosing’ to work undeclared. To explain why the unemployed benefit less from undeclared work than those in declared jobs, meanwhile, this survey has provided tentative support for the view that those in declared jobs have fewer fears about the authorities and greater resources, opportunities and skills to conduct better paid forms of such work.

What, however, are the implications for policy? Until now, national tax, labour and social security authorities responsible for tackling undeclared work have generally targeted the unemployed, partly because they are seen as more likely to conduct undeclared work than other population groups and partly because they are seen to take somebody else's money rather than keep their own (Cook 1997). Although this survey reveals that it is generally valid to focus upon the unemployed since undeclared work is more prevalent amongst this group, this does not tackle the vast bulk of undeclared work. Neither is it appropriate for national governments in all EU regions to place the same emphasis on targeting the unemployed since the proportion conducted by the unemployed is higher in some EU regions than others. Nor is it appropriate for those authorities responsible for tackling undeclared work to adopt a blanket-approach towards all the unemployed. A more variegated approach is required since significant variations exist across different sub-groups. What is now required, however, is more detailed research on why the unemployed work on an undeclared basis in different contexts. Unless their motives are understood, policies cannot be tailored to tackle the reasons for their participation. In some places, policies might focus upon creating more declared job opportunities or improving social support for those excluded from the declared realm. In other places where it is a more chosen activity, it might be more a case of tackling the barriers (e.g., the time, efforts and costs) to participating in declared work that lead them to work undeclared.

In sum, this survey has revealed the variable extent of the participation of the unemployed in undeclared work across the EU-27 and in doing so, proposed a new ‘reinforced marginalisation’ thesis. What is now required are more nuanced context-bound understandings of why they do so in order to identify the policies required to tackle participation. If this paper stimulates the development of such richer textured explanations, as well as more nuanced approaches towards how it might be tackled, then it will have achieved its objectives.

Colin C. Williams is Professor of Public Policy in the Management School at the University of Sheffield in the United Kingdom. His research interests include the informal economy, work organisation and the future of work, subjects on which he has published some 20 monographs and 270 journal articles over the past 25 years. His recent books include Informal Work in Developed Nations (2010, Routledge), Rethinking the Future of Work: directions and visions (2007, Palgrave Macmillan), The Hidden Enterprise Culture (2006, Edward Elgar), A Commodified World? mapping the limits of capitalism (2005, Zed) and Cash-in-Hand Work (2004, Palgrave Macmillan).

Sara Nadin is a Lecturer in Organisational Behaviour and Human Resource Management in the Management School at the University of Sheffield in the United Kingdom. She was previously an ESRC post-doctoral research fellow at the University of Sheffield and has held lectureships at the University of Leeds and the University of Bradford. Her research interests are in the informal economy, entrepreneurship, small business and the psychological contract. She has recently been coordinating an Institute for Small Business and Entrepreneurship (ISBE) research and knowledge exchange grant on developing policies to help businesses make the transition from informal to formal enterprise.