POVERTY AMONG SELF-EMPLOYED BUSINESSPEOPLE IN A RICH COUNTRY: A MISUNDERSTOOD AND DISTINCT...

1. Introduction

If the topics of entrepreneurship and poverty are linked, entrepreneurship is seen as a way to escape poverty. This is apparent, for example, in the title of a recent book, La culture entrepreneurial, un antidote la pauvret (The enterprise culture, an antidote to poverty) (Fortin, 2002). However, research on poverty among people who became poor as a consequence of running their own businesses and a fortiori in a rich country, is lacking. This can be a consequence of the popular stereotype, also embraced by the academic world, of high income self-employed businesspeople. Nonetheless, scientific research that compared incomes of self-employed people with those of employees showed that a large number of the self-employed have a low income (Bradbury, 1997; Hamilton, 2000; Meager and Bates, 2001). According to Meager and Bates (2001), an important avenue for future research is to attempt to identify more clearly the characteristics of individuals and activities that are associated with economic success in self-employment (in both the short and longer term), and those associated with less favorable outcomes. We intend this article to remedy these deficiencies by creating a quantitative and qualitative picture of poverty among self-employed businesspeople, with Belgium as case in point. The quantitative picture aims at quantifying the extent and characteristics of poverty among self-employed people. The nature of poverty among self-employed businesspeople, the causes and the problems that stem from poverty constitute the qualitative picture. They lead us to the formulation of policy recommendations.

The article is structured as follows. After the literature review, we explain the methodology and present our empirical findings. The study ends with conclusions and policy recommendations.

2. Literature Review

2.1. Quantitative picture

There exists a limited amount of published research on self-employment income, compared with income of employees, often in the domain of labor economics. Hamilton (2000) used data from the 1984 panel of the survey of income and program participation to construct self-employment/paid employment earnings differentials in the USA. He found that most entrepreneurs enter and persist in business despite the fact that they have both lower initial earnings and lower earnings growth than they would in paid employment. For example, after 10 years in business, median entrepreneurial earnings are 35 percent less than the predicted alternative wage on a paid job of the same duration, regardless of the self-employment earnings measure used (net profit, the amount withdrawn from the business in salary by the business owner [the draw], or the equity-adjusted draw [the sum of the draw in period t and the change in business equity between the beginning of period t and period t+1]). The present value to the median entrepreneur of a business lasting 25 years is over 25 percent less than the present value of a paid job of the same duration. According to Hamilton, the differential cannot be explained by the selection of low-ability employees into self-employment. The earnings of self-employed high school graduates are not significantly different from those of college dropouts, whereas the difference is significant in paid employment. Hamilton adds that the estimated earnings differentials may understate the differences in compensation across sectors since fringe benefits are not included in the measure of employee compensation. He concludes that the self-employment earnings differential reflects entrepreneurs' willingness to sacrifice substantial earnings in exchange for the nonpecuniary benefits of owning a business (such as "being your own boss").

Using UK data from the late 1980s and the 1990s, Meager and Bates (2001) showed that, relative to a current employee, a current self-employed person has more than three times the odds of being in the lowest decile group of the overall labor income distribution. There is a distinct self-employment effect, not reducible to personal and occupational characteristics. Looking at the likelihood of falling into the highest decile, the key finding was that, although the self-employed are overrepresented in this group, there is no significant impact of being self-employed, once the other variables (sex, age, sector, occupation, working time, and education) are controlled for. If prior employment status is taken into account, a self-employed person again has higher odds of being in the lowest decile, compared with someone with no self-employment experience. For those not currently self-employed, however, previous self-employment has less impact on the odds of falling into the lowest decile, and the effect is not statistically significant. This finding led Meager and Bates to the conclusion that, for those currently in the labor market, it is current self-employment that increases the risk of poverty. Once again, the impact of self-employment experience occurs at the lower end of the income distribution. It is other personal and employment characteristics that increase the likelihood of a higher income (especially being male, being qualified, working full time, being in managerial and professional occupations, and being in specific sectors). When looking at those aged 55-69, Meager and Bates found that those with self-employment experience have 1.8 times the odds of falling into the lowest income decile compared with those who have worked only as employees. Self-employment in work history has no significant effect on the odds of falling into the top decile among those aged 55-69. In addition, those aged 55-69 with self-employment experience (whether currently retired or working) are more likely to have both low incomes and low savings in later life than those with employee work histories. Meager and Bates conclude that the growth of "new self-employment" since the early 1980s in the UK may have led to a growing labor market segment whose self-employment is associated with low and unstable incomes, insecurity, and relative poverty in later life.

In the household budget survey by the National Office for Statistics (1998), about 20 percent of self-employed businesspeople in Belgium found it difficult to very difficult to live on their monthly disposable income (compared with 16 percent of manual employees and 12 percent of white-collar employees). A study carried out in 2000 at the OCMW (OCMWs are public institutions for providing social services) in Ostend (a city in Belgium) revealed that 10 percent of people on the statutory minimum income had been self-employed once or several times (Dewaele, 2000). Van Hecke and Marx (1999) studied poverty among farmers and horticulturists in Flanders (Belgium has three regions: Flanders, Wallonia and Brussels). They came to the conclusion that one in four farming households had significant financial difficulties. According to the Nationaal Actieplan Sociale Insluiting 2003-2005 the poverty risk for self-employed businesspeople in Belgium is as high as 10 percent compared with 3 percent for employees (Working group Indicators NAPIncl, 2004). Poverty risk equals the percentage of people with an income below 60 percent of the median income, which is the Belgian and European standard poverty line. In the Netherlands, the Poverty Monitor of the Sociaal en Cultureel Planbureau (Social and Cultural Planning Office) and the Centraal Bureau voor de Statistiek (Central Statistical Office) (2001) showed that 16 percent of self-employed businesspeople were experiencing financial difficulties in 1998. Furthermore, the probability of receiving an income below the poverty line was four times as high for self-employed businesspeople as for those in salaried employment. The Dutch research also showed that the agricultural sector was vulnerable (20 percent of farmers had a low income, compared with 16 percent of the total group of self-employed businesspeople).

Previous research also shows the greater polarization of incomes among the selfemployed when compared with those in traditional wage employment. Hamilton (2000) showed for the USA that the distribution of self-employment earnings exhibited greater dispersion than that of the wage distribution. Approximately 13 percent of business owners earned more than $20 per hour, when the equity-adjusted draw was used (8.8 percent for draw and 7.6 percent for net profit), compared with only 4.2 percent of employees. Substantial returns for self-employment are clearly concentrated among a few entrepreneurs. This is also confirmed in the study of Meager and Bates (2001) in the UK. The ratio of ninetieth to tenth percentile equals 22.36 for the self-employed, compared with 6.29 for employees. Also, within the age group 55-69, the income distribution (including incomes from pensions and savings) of those with self-employment experience is more polarized than the distribution for those with only wage employment experience. The Gini coefficient (0 = perfect equal distribution and 1 = completely unequal distribution) equals 0.49 for men with a work history that includes self-employment, compared with 0.35 for men who have always been employees during their work history (for women, these Gini coefficients equal 0.52 and 0.42 respectively). Parker (1997) discovered, on the basis of a study of the pattern of UK selfemployment inequality between 1976 and 1991, that the self-employment income distribution became progressively more unequal. Parker's results suggest that this rise in inequality was principally caused by an increase in the heterogeneity of the self-employed themselves, which was particularly pronounced between 1981 and 1986. In contrast, the increased numbers engaged in self-employment seems to have had no discernible effect on inequality.

2.2. Causes of poverty and business failure

In the literature on poverty among self-employed businesspeople in Belgium, economic and personal causes are cited. The economic causes include interest on tax arrears and penalties that are higher than the actual debt (Efrem vz w, 2001). Personal causes refer to administrative inability, poor financial management, divorce, illness and accident. Due to the strong ties between family and business life, personal factors may have an influence on business, and vice versa (Dewaele, 2001). Finally, the weaker employment status of the self-employed businessperson (in terms of incapacity to work, pensions, etc.) plays an important role in the poverty issue (Dewaele, 2000; Steverlynck, 2001).

In the USA, between 13 and 14 percent of personal bankruptcies are actually small business bankruptcies in disguise (Cleaver, 2004). It was observed that, when business owners go under, they go under harder. Therefore, it is also useful to look at the determinants of business failures. However, we stress that not every bankruptcy leads to poverty of the owner-manager and that poverty of the entrepreneur is not always caused by the bankruptcy of his/her business. An analysis of the determinants of business failure in 16 European countries indicated that failure can be rarely attributed to a single factor or cause (ENSR, 1997). The European study classified the causes into four different categories. The first category consists of external factors. These encompass changes in the market or industry structure, inadequate product or production capacity and regulatory changes (for example, changes in professional requirements and environmental standards). In the apparel and accessory retailing industry in the USA region of Iowa, the competitive environment (competition from discount stores, inability to compete in a trading area and failure to offer saleable merchandise assortments) was also identified as a cause of small business failure (Gaskill, Van Auken and Manning, 1993). The second category of causes in the European study consists of financial problems linked to an inability to absorb shocks due to an inadequate capital base. These shocks include a rise in interest rates or other fixed costs, changes in exchange rates, increases in the number of settlement days or in bad payment, or shifts in banks' credit policies. Bad or poor management is the third cause in the European study. This most frequently cited cause of business failure is somewhat simplistic. As Argenti (1976) puts it, while everyone agrees that bad management is the prime cause of failure, no one agrees what "bad management" means nor how it can be recognized except that after the company has collapsed - then everyone agrees how badly managed it was. Gaskill, Van Auken and Manning (1993) referred to weak managerial and planning functions (for example failure to generate a long-term business plan, a personnel plan, etc.) as cause of small business failure. However, Perry (2001) found a statistically significant but weak relationship between planning and firm failure. According to him, a possible explanation for the lack of a strong relationship lies in the fact that firms with fewer than 5 employees rarely undertake formal (written) planning, where it may have limited value and utility for such tiny businesses. The European study designated the fourth category of causes as "other possible factors". These include small legal and financial consequences of failure to the owner; bad luck, fraud and negligence on the part of agents involved with the enterprise or personal problems with key enterprise personnel. Gaskill, Van Auken and Manning (1993) added factors related to vendor relations (poor relations with vendors, difficulties in receiving merchandise and inadequate financial accounting or record keeping) and to premature business growth and overexpansion. Chen and Williams (1999) examined the role that some key state fiscal measures and federal transfer grants to states play in explaining business failures in low-technology and high-technology industries in the USA. Higher sales taxes tend to increase business failure rates. Development assistance programs (providing business advice, technical assistance and jobs training) tend to improve small business survival rates for high-technology industries, but they are ineffective for low-technology industries. Van Praag (2003) studied person-oriented drivers of business survival and success of young white small business owners in the USA. She discovered that compulsory or forced exits were more probable for business owners that started at a younger age or because they are unemployed, and who had little experience in the same industry as the business venture. Dummy factors for industry turned out to be insignificant.

3. Methodology

Our empirical study comprises a quantitative and a qualitative section. In order to depict poverty in statistical terms, we analyzed the net professional income of the total population of self-employed people with self-employment as their main occupation in Belgium for the period 1995-2000, according to the Rijksinstituut voor de Sociale Verzekeringen der Zelfstandigen (RSVZ: "National Social Insurance Institution for the Self-Employed", which is a government institution responsible for optimal application of the social security system for the self-employed). Net professional income is equal to gross professional income minus business expenses and operating losses. The term "self-employed" indicates the employment status of an individual. From the RSVZ data, it cannot be determined which self-employed people are "self-employed businesspeople". We use the term "self-employed businessperson" to denote someone who runs his/her own business. The incomes from those with self-employed status but who do not run their own business (for example, the directors of a financial institution) are also shown in the RSVZ database. We "restrict" our quantitative analysis to those who are self-employed as their main professional activity (this dataset does not allow restriction to solely self-employed businesspeople). Those who are self-employed in a second occupation or those who are self-employed in a main occupation that is treated as equivalent to a second occupation are not included, since they could distort the results due to their lower income from self-employment. The RSVZ database contains the incomes of self-employed people, but not of their households.

In addition to income, the following information is available from the RSVZ database: gender, sector, nationality, age categories, regions and number of years' income from the main occupations of self-employed people. We investigate whether the income varies according to those characteristics. When our data were from a survey, we would have estimated the impact of the characteristics by means of the appropriate statistical techniques. However, we stress that this is not necessary as we work with the data from the whole population of the self-employed with self-employment as their main activity. We do not need to use equations that enable us to extrapolate the results to the whole population. We can present the real data.

We are aware of potential critical comments about the use of declared net professional income as a measure of poverty. There is a suggestion of suspicion about the income of the self-employed. Some people believe that the actual income of self-employed people is higher than the sum that they declare to the tax authorities (i.e., as they work for cash in hand). As Meager and Bates (2001) state, there is no reliable way to adjust income data for underreporting. Anyway, the situation of the self-employed person can be even more distressing than is apparent from his/her income. If his/her income is too low, it is necessary to draw on personal assets in order to survive. This decrease of the personal assets is not included in the income statistics. According to Holm (2000), the income of a self-employed person needs to be 50 percent higher than the income of a salaried employee, in order to cover the risk of bankruptcy and income taxes. Finally, there are practical and intrinsic reasons to use income as a measure of poverty. From the practical viewpoint, income enables poverty thresholds to be delineated. A family is considered as poor or not having enough to live decently if their income falls below that threshold. Intrinsically, adequate economic resources (income) are an important condition to reach a certain level of welfare and social participation. Statistics send out the first signals that a self-employed person's business is not doing well.

Table 1 compiles a number of descriptive statistics of the annual income of the selfemployed for the period 1995-2000. The mean annual income of the self-employed increased every year in the period 1995-2000. The median is always below the mean, which indicates an unequal income distribution. A limited number of high-income self-employed people cause an increase in the mean.

The statistics do not allow us to understand the nature of poverty among self-employed businesspeople, the causes and problems. Therefore, a qualitative approach is absolutely necessary. In addition, it enables us to study self-employed businesspeople and not the selfemployed, which is a shortcoming in quantitative income analysis. We held individual interviews with 15 public social welfare centers (OCMWs), four expert witnesses from supporting organizations and 20 self-employed or previously self-employed businesspeople living in poverty. Two researchers independently completed a total of almost 60 hours of interviews. In order to detect self-employed businesspeople in poverty, we appealed to the OCMWs and supporting organizations. They enquired whether self-employed businesspeople were prepared to cooperate with the study. Whenever the businessperson agreed, we immediately made an appointment for an individual interview. In the selection of the self-employed businesspeople, we were guided not by considerations of statistical representativeness, but by diversity (e.g., by sector and number of years in charge of their own business). During the personal interview, the self-employed businesspeople first told their story. We ensured that at least the nature of the poverty, the causes and the problems were always addressed. To illustrate our qualitative findings, the interviewees are quoted; the source of the quotation (the self-employed businessperson, the spouse of the self-employed businessperson, an expert witness from a supporting organization or a social worker) is then mentioned.

Table 2 gives the breakdown of the 20 self-employed businesspeople according to a number of personal and business characteristics. Interviews were held with 14 men and six women of various ages. The average number of years of running their own business among those interviewed was 14. Only three entrepreneurs had been in business for less than five years. Seven interviewees had been running their own business for over 20 years. Eight entrepreneurs had been running their own business for between 10 and 20 years. For two interviewees, the number of years of self-employment was unknown. These figures on duration make it already clear that experienced self-employed businesspeople can also fall into poverty. The self-employed businesspeople interviewed came from various sectors. In terms of form of business entity, they were mainly one-person businesses. In nine cases, it was a business that they had started themselves. Nine other self-employed businesspeople took the business over from a third party (five) or a relative (four). Among self-employed businesspeople who had had several self-employed businesses, we restricted ourselves in the table to their most recent business.

4. Empirical Findings for Belgium

4.1. Quantitative picture

According to our own calculations, based on the net professional income of those who are self-employed in their main occupation, approximately one-quarter of those who are selfemployed in their main occupation are living below the poverty line (defined as 60 percent of the median income, or a monthly income of 523 euros after tax) within the group of selfemployed people. By working with 60 percent of the median income, we are adopting the relative or statistical method for determining the poverty line, which is used in Belgium and Europe (Cantillon, Deleeck and Van den Bosch, 2001). When we use the poverty line for the whole Belgian population (also 60 percent of the median income, or a monthly income after tax of 594 euros in 1998), one third of the self-employed in their main occupation are living in poverty. These results confirm other empirical findings in the UK, the USA, Belgium and the Netherlands that a substantial number of self-employed have a low income.

Approximately 37 percent of the self-employed with at least six years' income (N = 319,511) had experienced at least one year below the poverty line during the period 19952000, within the group of those who are self-employed in their main occupation. Figure 1 shows that nearly 70 percent of them have been living below the poverty line for at least two years (100-31.98).

Self-employed people with at least six years' income from their self-employed main occupation who remained in the lowest income category throughout the six-year period (from 1995-2000 inclusive: for each year, we calculated four income categories based on quartiles), recorded average annual real income growth of - 1.59 percent. Self-employed people who remained in the higher income category throughout that period experienced annual income growth (+2.42 percent).

As in the UK and the USA, the income distribution of self-employed people in Belgium is very unequal and much less equally distributed than that of the Belgian population as a whole. The Lorenz curve shows that 70 percent of self-employed people (those on the lowest incomes) only earn one-third of the total income of those who are self-employed in their main occupation (see Fig. 2). That means that 30 percent of those who are self-employed in their main occupation (those who are earning most) are receiving two-thirds of the income pie. The Gini coefficient within the group who are self-employed in their main occupation is 0.5, compared with 0.3 for the Belgian population overall (Vranken, Geldof, Van Menxel and Van Ouytsel, 2001). The great income inequality can also be seen from the fact that the annual income of the top 10 percent of earners is more than 12 times higher than that of the lowest 10 percent of earners (45,356 euros compared with 3,656 euros).

Finally, we examine the influence of the following characteristics on belonging to an income category (we distinguish between four income categories based on quartiles) and on the number of years that a self-employed person (with at least six years' income) spent in poverty (based on the poverty line within the group of self-employed in their main occupation), as this is a gap in the literature. The following characteristics were available: gender (male, female), age (<35, 35-50, 51-65, >65), nationality (Belgian, non-Belgian), sector (retail, liberal and intellectual professions, services, manufacturing and primary sector) and region (Belgium has three regions: Flanders, Wallonia and Brussels). It is mainly selfemployed people over 65 years of age working in the primary sector who belong to the lowest income category, which confirms existing research in Belgium and the Netherlands. Among non-Belgians in Brussels, we are more likely to find the youngest self-employed (both men and women) in the lowest income category. On the other hand, males over 65 in Wallonia and Flanders with a liberal or intellectual professional and with Belgian nationality are mainly found in the highest income category. Among male non-Belgians in Brussels, the over-65s working in retail or the services sector are more likely to be in the highest income category (40 percent and 50 percent respectively).

The highest percentages of self-employed with over two years of poverty in 1995-2000, over 50 percent were found among self-employed women. In Wallonia and Flanders, this is the case with self-employed people with Belgian nationality in the services sector (60.59 percent and 55.91 percent) and the primary sector (63.41 percent and 57.81 percent). For self-employed females with a foreign nationality, this only applies in the same sectors in Wallonia. Among self-employed males with Belgian nationality, the highest percentages are found in the primary sector (27.80 percent in Wallonia, 29.88 percent in Flanders and 20.95 percent in Brussels). The percentages of male self-employed with a foreign nationality with over two years below the poverty line are higher in all sectors, apart from the primary sector and manufacturing in Brussels, than among male self-employed with Belgian nationality.

4.2. Nature of poverty among self-employed businesspeople

The interviews made clear that self-employed businesspeople living in poverty are isolated.This isolation is the consequence of concealing the problems. They perceive the failure of their business as a personal failure: "7f is not only a failure of the business, but also a failure of the self-employed businesspersori" (Social worker). This can be explained by the specific nature of self-employed businesses, which is expressed in the entanglement between working and private life. Setbacks with the business are personal blows to the selfemployed businessperson and his/her family. Conversely, family problems can cause harm to the self-employed business. Divorce seems to be both a cause and a result of poverty among self-employed businesspeople.

Because of the feeling of personal failure, self-employed businesspeople living in poverty are afraid of the reaction of their friends and associates. That means that they cut themselves off from the outside world, and sometimes even leave the village or town where their business was located:

"When the business failed, people accosted us in the street. Then my wife always started to weep. We didn't want to talk about it and didn't want to have any contact with former business partners. We decided to withdraw and moved to our native village" (Self-employed businessperson).

Those providing assistance observe that poverty does not bring self-employed businesspeople together. Among poor people who are not self-employed, poverty does bring them closer. They often form organizations and live ghetto style. A second difference with poor non-self-employed is that poverty among self-employed businesspeople does not fit the "conventional" image of poverty and deprivation from one generation to the next:

"Social policy of the past few years wanted to combat 'conventional' poverty. However, poverty among self-employed businesspeople is a combination of social and economic factors" (Expert witness from a supporting organization).

A third difference is that there has been no institutionalized support for self-employed businesspeople suffering from poverty. They feel that they have been abandoned to their fate. Furthermore, there are few interfaces between social work and self-employed businesspeople. OCMW social workers, who admittedly get a positive evaluation from self-employed businesspeople in poverty, have to wrestle with complex legislation regarding the selfemployed (for example, concerning their social security status) and with the complexity of a self-employed business (e.g., screening of the financial situation). OCMWs do not always have well-grounded (legal) knowledge, the necessary time, sufficient staff and finances to understand and help self-employed businesspeople in difficulties. Therefore, some OCMVKs do not immediately pay them a living wage or they do not grant any financial support to self-employed businesspeople who are still running their businesses. Nevertheless, there is no legal prohibition of OCMWs giving financial support to active self-employed businesspeople (Van Noyen, 2003). For self-employed businesspeople themselves, social assistance is unknown territory. Concepts such as "welfare" and "support" hold negative connotations and induce shame:

"To ask for help at the public social welfare centers is extremely humiliating. You have to ask for support, while you have fought your own battles for over 20 years" (Self-employed businessperson).

Self-employed businesspeople find it hard to apply for social assistance, and they only do so at a late juncture. This requires them to take the unnatural step from independence (the main reason for deciding to run their own business (Bouckaert and Sels, 2001)) dependence.

Finally, self-employed businesspeople in poverty encounter substantial resistance and lack of understanding among some tax collectors, insolvency administrators and bailiffs whose work brings them into contact with self-employed businesspeople. Their arbitrariness, lack of transparency concerning fees to be charged and operating methods, and misuse of their powers are criticized:

"Bailiffs are bullies. They are vultures who only enlarge your debts. Their fees are unbearable. One bailiff asks 175 euros, while the other charges 620 euros. And you have to pay, otherwise they take everything" (Selfemployed businessperson).

4.3. Causes of poverty among self-employed businesspeople

The interviewees refer to similar causes, several of which may occur simultaneously. We can classify them under five headings: economic trends, government action, setbacks, personal causes and abuses by third parties.

First of all, self-employed businesspeople may become poor because of economic trends (economic recession, increased competition and unfavorable conditions in the sector), which confirms the literature on bankruptcy in Europe and the USA:

"The rise of the computer and digital cameras are death for photography. In addition, the big chains sell their products at a price at which a selfemployed businessperson cannot even purchase them"(Self-employed businessperson).

Government action (high tax burden, lengthy public works, the weak social security status of the self-employed, interest on VAT and tax arrears which exceed the actual debt, etc.) is a secondary cause (also stated in the literature on poverty and bankruptcy):

"The problems started during the public works in our village. The public works lasted more than one year and caused serious financial problems. I lost lots of customers and my debts were piling up" (Self-employed businessperson).

Setbacks (accidents, non-payment by clients, etc.) are a third cause and are also found in the literature on poverty and bankruptcy:

"My husband had an accident, for which he was not insured. For six months, he had to stay in a wheelchair. In order to continue the business, we had to employ two additional people. It would have been better had we stopped, because the wage cost was too high. Moreover, there were the hospital bills (...)" (Wife of a self-employed businessperson).

Personal causes (inadequate management skills, poor relations with business partner, doing business rashly or naively, divorce and health problems) also underlie a substantial proportion of poverty (personal causes were also mentioned in the literature on poverty and bankruptcy):

"The divorce caused business problems. My wife was responsible for the administration and bookkeeping. She also went to the bank, the customers.. .after the divorce I was obliged to hire someone to do all of this. But the labor cost was too high. Moreover my wife blackened my good reputation, which led to the loss of several customers" (Self-employed businessperson).

Finally, abuses by third parties (negligence by the accountant, harassment by (former) employees, misleading information from the vendor of a business to the new owner, etc.) can land self-employed businesspeople in serious difficulties. This fifth cause was mentioned neither in the literature on poverty nor in the literature on bankruptcy:

"The bookkeeper cooperated with the previous owners of the business who had lied about the number of clients. The bookkeeper knew it and he did not say anything. Now he refuses to give back the books, which is causing problems with my tax retum"(Self-employed businessperson).

Our interviews with self-employed businesspeople in poverty made it clear that these causes can also create poverty even among experienced self-employed businesspeople who had previously been very successful:

"I can't even figure out myself how a wealthy self-employed businessperson can become this poor. The surety and blind trust in my son-in-law ruined me. Afterwards it became clear that my son-in-law had gambled away all the money that I had given him" (Self-employed businessperson).

4.4. Problems accompanying poverty due to self-employment

The interviews with poor self-employed businesspeople indicated that they are faced with a diversity of problems: financial, psychological, physical and relational. Their financial difficulties vary from excessive expenditure and income that is too low, to a high level of debt. The largest creditors are often the income tax and VAT authorities and the social insurance funds that collect the social security contributions of the self-employed. Since selfemployed businesspeople run their own businesses, poverty for them goes hand in hand with mental and physical problems such as stress, fatigue, depression, loss of self-confidence, loneliness, continuing anxiety and even suicidal thoughts:

"The only thing you could be proud of, your business, is gone. You even start to doubt that which you excel in, your profession" (Self-employed businessperson).

Serious problems with self-employment can also result in divorce and in serious tensions within the family.

For most self-employed businesspeople, the problems drag on for several years. They attempt to overcome the problems in various ways. Some try to work even harder, whereas others incur more debt, which sometimes puts them into a debt spiral. Professional help is hardly ever called in. The business may be closed down, which sometimes gives a feeling of release. The question that arises is why self-employed businesspeople in difficulties do not cease their self-employment activities earlier. Wanting to stop and being able to stop are not things that they wish to contemplate or think themselves capable of doing. If being selfemployed is the realization of a dream for them, they want to continue as long as possible. Self-employed businesspeople are convinced that they can improve their situation, which is often not the case:

"The problems have lasted one year, because my husband wanted to continue. He thought each time that he could solve the situation, but the problems became only bigger. After a while, he started to see that. It would have been better if we had stopped earlier (Self-employed businessperson).

Self-employed businesspeople in difficulties are often very unsure about what is in store for them after they cease trading, which means they decide to carry on. If they still have loans outstanding and the self-employed business is their main source of income, it is difficult to give up:

"You are misinformed and afraid of the future. You imagine the most dreadful scenarios based upon what you hear and see on television about people who went bankrupt and have lost everything. There is a market place in our city where possessed goods are being sold. I couldn't imagine that our stuff would be sold there soon" (Self-employed businessperson).

Self-employed businesspeople cutting themselves off from the outside world or not being able to stop, combined with no institutionalized support when they are suffering from poverty and substantial resistance and lack of understanding, are characteristic of the culture in Belgium. "Failure" is stigmatized, and is neglected in the life cycle of a business and an entrepreneur. Policy ignores self-employed businesspeople in difficulty and in this way pushes self-employed businesspeople into poverty. This culture of neglect and condemnation of self-employed businesspeople in difficulties has four consequences. First, it is very difficult to overcome these difficulties and to remain a self-employed businessperson:

"A legal settlement can protect self-employed businesspeople for six months against creditors. Social security votes mostly against the extension of the legal settlement. The only way out is bankruptcy. One then reaches the poverty line and has to try to survive" (Expert witness from supporting organization).

Second, self-employed businesspeople in poverty who stop their independent activity are not caught by the social safety net. In Belgium, those self-employed for over nine years have, for example, no entitlement to a replacement income. The unemployment benefit is only paid when the self-employed person has started his/her business having previously been employed or unemployed, and gives up his/her self-employment within nine years. A third consequence is that self-employed businesspeople in difficulties are doomed to continue, which can exacerbate the situation. Finally, it is very difficult for a "failed" self-employed person to re-enter self-employment, certainly when there was a bankruptcy:

"Because of the bankruptcy, I cannot restart an independent activity in my name" (Ex-self-employed businessperson).

5. Conclusions and Policy Recommendations

In Belgium, the Netherlands, the UK and the USA, existing research compared selfemployment income with income of employees. In all these countries, it was found that self-employed businesspeople have a higher probability of falling into the lowest income groups and that there is a distinct self-employment effect. The National Offices for Statistics in Belgium and the Netherlands calculated that 16-20 percent of self-employed businesspeople have difficulties living on their monthly disposable income. Our own empirical findings for Belgium - on the basis of the analysis of the net professional income of the total population of about 450,000 self-employed people for the period 1995-2000 - indicated that approximately one-quarter of those who are self-employed in their main occupation are living below the poverty line. The poverty line is defined according to European standards as 60 percent of the median income.

Existing research in the UK and the USA also shows greater polarization of incomes among the self-employed when compared with those in traditional wage employment. Substantial returns for self-employment are concentrated among a few entrepreneurs. This is confirmed by our own empirical findings for Belgium. The Gini coefficient (0 = perfect equal distribution and 1 = completely unequal distribution) within the group who are selfemployed in their main occupation is 0.5, compared with 0.3 for the Belgian population overall.

Concerning the profile of self-employed businesspeople in poverty, existing research in Belgium and the Netherlands referred to the agricultural sector as being vulnerable. It was estimated that 20-25 percent of farmers in Belgium and the Netherlands face significant financial difficulties. Our study of the influence of entrepreneur- and enterprise-related characteristics on belonging to an income category showed that self-employed businesspeople over 65 years of age working in the primary sector mainly belong to the lowest income category.

In order to understand the nature of poverty among self-employed businesspeople, the causes and problems, we also held interviews in Belgium with public social welfare centers, expert witnesses from supporting organizations and self-employed or previously self-employed businesspeople living in poverty. Our qualitative findings made clear that self-employed businesspeople living in poverty are isolated, which is different from poor people who are not self-employed. This isolation is a consequence of cutting themselves off from the outside world (they feel the loss of their business is a personal failure), not fitting the "conventional" image of poverty and deprivation from one generation to the next, and the absence of institutionalized support for them. In addition, self-employed businesspeople in poverty in Belgium encounter substantial resistance and lack of understanding among some tax collectors, insolvency administrators and bailiffs with whom they are brought into contact. These officials sometimes behave arbitrarily, are not transparent regarding their fees to be charged and in their operating methods, and misuse their powers. This absence of institutionalized support for poor self-employed businesspeople explains why approximately 37 percent of the self-employed with at least six years' income in Belgium have experienced at least one year below the poverty line and why nearly 70 percent of them have been living below the poverty line for at least two years. Several factors - economic difficulties, government action, setbacks, personal problems and abuses by third parties can cause self-employed businesspeople in Belgium to end up in poverty. All but the last of these causes can be found in the scarce literature on poverty and bankruptcy. The diversity of causes makes poverty a multifaceted problem: financial, psychological, physical and relational.

The isolation of self-employed businesspeople living in poverty and not being able to stop is characteristic of the culture in Belgium. 'Failure' is stigmatized and is neglected in the life cycle of a business and an entrepreneur. Policy ignores self-employed businesspeople in difficulty, and in this way, pushes self-employed businesspeople in poverty.

From our quantitative and qualitative findings on poverty among self-employed businesspeople, we derive five policy recommendations that are in the first instance valid in the Belgian context. However, should research in other rich countries on poverty among people who became poor as a consequence of running their own businesses, which is now lacking, yield similar results in a comparable legal context, the policy recommendations can be generalized. Our policy recommendations are intended to lead to recognition and acknowledgement of a real but (still) misunderstood problem, and to prevent and combat poverty among self-employed businesspeople. Firstly, since there is no institutionalized support for self-employed businesspeople in poverty, we advocate integrated (having a single point of contact) and holistic (providing a total solution) support for them by a government body (a government and not a private body to avoid conflicts of interest). That government body would work according to the hub and spoke system (see Fig. 3).

The hub is formed by the (ex-)self-employed businessperson in difficulties and his/her person of confidence in the government body, the customer specialist. That person is a sounding board and detects and develops the needs of the self-employed businessperson. Based on the actual needs of the (ex-)self-employed businessperson, the customer specialist appeals to the internal (work in the government body) and/or external (work outside the government body) solution specialists. The solution specialists have know-how in a specific domain. Given that poor self-employed businesspeople face especially financial, fiscal, legal and psychological problems, we suggest that specialists in these areas work inside the government body (internal solution specialists). The customer and solution specialists are very familiar with the world of self-employed businesspeople and have enormous empathy. Customer specialists could, for example, consist of (ex-)self-employed businesspeople.

Secondly, in order to stop the arbitrariness of tax officials, administrators, bailiffs, etc., such authorities must publish their procedures, negotiating periods, scale of charges, rights and obligations of the (ex-)self-employed businessperson in difficulties, in comprehensible language and hand it over to the (ex-)self-employed businessperson in time. Thirdly, we are also advocating a fund financed by contributions from self-employed businesspeople, large companies and employees that can be used in the event of setbacks (one of the causes of poverty). Fourthly, in order to avoid tax arrears exceeding the actual debt and "strangling" the self-employed businessperson (another of the causes of poverty), the repayment of tax debt should first be deducted from the principal and not from the interest to be paid.

Finally, given that even experienced entrepreneurs who had been successful in the past can fall into poverty and to encourage self-employed businesspeople in severe difficulties to stop their independent activities in time, self-employed businesspeople that cease trading need a decent replacement income to fall back on. It is a notable paradox that the social safety net is the lowest for the independent entrepreneurs - the only risk takers in our society (as they work with their own money).

We hope this article can lead to the study of poverty among (ex-)self-employed businesspeople in other countries to see whether the findings for Belgium can be confirmed and supplemented. We wish that these studies could overcome the shortcoming in this research by undertaking a quantitative study of the household income of self-employed businesspeople (not only the income of the self-employed businesspeople and excluding the income of people who have self-employed employment status but who do not run their own businesses).

Acknowledgments

The authors wish to thank CERA Foundation for its financial support for this study and three anonymous reviewers for their constructive suggestions and remarks.