14 Abbing (1989) 126.
15 Cf. Towse (1993).
16 Frank and Cook (1995) 200-9.
17 Cowen (1998) 131 challenges Frank and Cook"s thesis of cultural impoverish-ment. He defends the position that cultural diversity increases.
18 The fact that essential qualities cannot be measured does not exclude that cer-tain qualities, which matter to the consumer, can be measured, like singing, for instance. Cf. Hamlen (1991) and Towse (1993).
19 In this context Rosen (1981) 846 uses the word "charisma", which has a slightly different meaning.
20 Authenticity in the arts was treated in chapter 1 in relation to the a.s.sumed sa-credness of art.
21 Everybody is necessarily creative and authentic at times for instance, in .nd-ing solutions in a problematic relationship or in solving problems at work.
22 I do not want to suggest that people have no urge to conform. It could well be argued that both the urge to conform and the need for distinction, i.e., the wish to be "somebody who differs from everybody else" are powerful impulses in many people"s lives. In our culture, however, the latter has probably gained more momentum, and the artist serves as its model. Although it is true that by strongly identifying with artists or stars people apparently become less of an au-thentic individual themselves, this kind of identi.cation and degree of symbiosis are usually temporal. In the long run they help in the formation of ident.i.ties. These ident.i.ties naturally exist in relation to the "ident.i.ties" of the social groups or subcultures somebody takes part in. After all, distinction can only exist in a social environment.
23 Frank and Cook (1995) 29, emphasize that, although ent.i.ties other than per-sons may compete in winner-takes-all markets, identi.able individuals are al-most always involved, when the stakes are high.
24 Abbing (1989) 124-135, Frey and Pommerehne (1989) 137-164, Jeffri (1989), Towse (1993), Elstad (1997) and Rengers and Madden (2000). Many other studies, which con.rm low incomes and a skewed distribution of incomes, were reviewed by Throsby (1994), O"Brien (1997) and Wa.s.sall and Alper (1992). A different opinion comes from Filer (1986), who states that artists earn only a lit-tle less (not more than 10% less) than comparable professionals, but his use of census data has been criticized, for instance by Menger (1999) 553. See also note 213.
25 Baumol and Bowen (1966) 103.
26 Frey and Pommerehne (1989) 152-55.
27 "Surveys (and of.cial statistics) on artists show they are younger, better quali-.ed in the sense of having higher levels of educational achievement, earn less and work longer hours than other professionals." Towse (2001) 485.
28 Menger (1999) 545.
29 Throsby (1994) 18.
30 Alper, Wa.s.sal et al. (1996).
31 Throsby (1996).
32 Menger (1999) 545. Earlier Peac.o.c.k, Shoesmith et al. (1982) 39 noted that be-tween 1970 and 1980 in Britain the real value of performers" incomes working in established companies decreased.
33 On the basis of census data, Cowen and Grier (1996) 14-20 discovered that in-comes and the numbers of artists in the us rose slightly between 1970 and 1990. However, in the census data, people are categorized by their primary source of income. Many artists earn most of their incomes outside the arts. Measuring the income of a group already selected on a .nancial criterion can lead to incorrect or irrelevant results.
34 Throsby (1996) 228.
35 Meulenbeek, Brouwer et al. (2000) 36. Fifty percent earned less than 100 Euros a month. And more than 75% could not make a living, in the sense that they earned less than the equivalent of social security bene.ts in the Netherlands. Unlike in some other surveys, the respondents had no incentive to hide their ille-gal earnings.
36 Rengers and Madden (2000) 339 discovered that in Australia artists with low art incomes and artists with high art incomes work long weeks, while artists with intermediate incomes tend to work shorter weeks in the arts. The latter group probably consists of predominantly non-.ne arts artists.
37 Meulenbeek, Brouwer et al. (2000) 24.
38 Struyk and Rengers (2000) 24.
39 Median income implies that 50% of a group earns less and 50% more than this .gure. Because income distribution is always skewed, the median income is less than the average income. The top of the median income is not in.uenced by co-incidentally high incomes. Therefore, it is generally more informative than aver-age income. Nevertheless, average income is often used, because median income .gures are not always available.
40 Heilbrun and Gray (2001, ed princ 1993) 15, Menger (1999) 542. See also sec-tion 7.7.
41 Cf. Menger (1999) 607-8.
42 Frank and Cook are not primarily interested in the height of low incomes in winner-takes-all markets. They are interested in social ef.ciency. They demon-strate that from a social point of view, there tends to be too many partic.i.p.ants in these markets. Their a.n.a.lysis implies that even when the other causes of low in-comes in the table do not apply, there would still be a surplus of partic.i.p.ants and therefore, social waste. Frank and Cook (1995) 107-110. Frank and Cook (1995) 102-103 do not agree with the popular notion that society needs a very big pond in order to catch a few very talented artists. They evidently disagree with Adler (1985) who has tried to offer a scienti.c basis for the notion of the big point. In Chapter 9, I will return to this point.
43 Menger (1999) 554 presents related explanations.
44 A .gure can be drawn (.gure 1 presented at: ) in which the y-axis stands for monetary income and the x-axis for non-monetary income. Ua,b, etc. are the indifference curves of an average artist with respect to both forms of income and Up,q, etc. are similar indifference curves of a non-arts pro-fessional. Where the indifference curves of the artist cross the curves of the oth-er professional, the slope of the .rst is greater.45 Frey (1997) and Towse (2001) a.s.sume that artists who seek status are intrinsi-cally instead of extrinsically motivated. See note 3 of chapter 4. In my opinion, the way they distinguish intrinsic and extrinsic motivation is confusing. There-fore, I use non-monetary versus monetary rewards and motivation.
46 Economists have been reluctant to use the concept of psychic income because they think it has too often served as an easy way out for the explanation of phe-nomena, which are otherwise dif.cult to explain. Cf. Towse (1996b) 310.
47 This shows in the changing slope of the indifference curve in .gure 1 mentioned in note 44.
48 In a .gure (.gure 2, presented at: ) which is a more de-tailed version of .gure 1 mentioned in note 44, the kink in the indifference curves of artists, where the survival constraint for artists is approached, ex-presses this phenomenon.49 Not to complicate matters, in this chapter, the term "risk" also covers the notion of uncertainty.
50 Cf. Singer (1981b), Frey and Pommerehne (1989) 157 and Frank and Cook (1995) 116-118. Singer was probably the .rst to acknowledge the importance of risk taking in explaining artist behavior. He applied the Friedman-Savage gambling model to the visual arts.
51 Throsby (1994) 19.
52 Cf. Frank and Cook (1995) 103-105.
53 Frank and Cook (1995) 104.
54 Menger (1999) 553-4. See also note 63 of Chapter 6.
55 Menger (1999) 601.
56 Another outcome can be a so-called pig cycle, which exists in a mild form in most professions: periods of too much entrance alternate with periods with too little.
57 In the Netherlands, for instance, the percentage of self-taught artists with no of-.cial art education is approximately 12%. This percentage for visual artists was found by Meulenbeek, Brouwer et al. (2000) 21 and the same percentage was found by Struyk and Rengers (2000) 22 for performing artists. There are indi-cations that among younger artists this percentage is even lower.
58 The percentage of students at Dutch art schools with more highly educated par-ents, according to Allen and Ramaekers (1999) 18, is 40% higher than in other higher education inst.i.tutions. If .gures were available for the .ne arts depart-ments at arts colleges this .gure could be even higher.
Structural Poverty
1 Cf. Frey and Pommerehne (1989).
2 Cf. Montias (1987).
3 Cf. Hoogenboom (1993) and Stolwijk (1998).
4 Different types of bohemian artists can be distinguished cf. Kreuzer (1971). The social backgrounds of various bohemian artists varies considerably. Never-theless, a majority of bohemian artists probably come from families with above-average incomes.
5 The effect of more stringent restrictions of numbers can be shown in a .gure (.g-ure 3 presented on ). With average hourly income on the y-axis and the total number of hours that all artists spend making art on the x-axis, the relevant part of the supply curve of artistic hours S moves to S", which appears above and to the left of S. S" is also steeper than S. With more control applied, the intersection with the demand curve also moves up and income rises as well. Cf. Frey and Pommerehne (1989) 160.6 In the us there was support for artists in the 1930s through the Work"s Progress a.s.sociation, which was part of Roosevelt"s New Deal.
7 Chapter 9 discusses other arguments for subsidization.
8 For increased spending, see note 28 of Chapter 7 and for low incomes note 24 of Chapter 5.
9 Presently I prefer to ignore the sixth explanation, the winner-takes-all explana-tion, because artists are not exceptional in that respect.
10 In .gure 3 described in note 5 (and presented on: ) the relevant part of the curve that represents total number of hours worked in the arts at different income levels runs almost horizontal. The three inclinations in combination with the survival constraint for artists cause a high income-elastic-ity in the relevant part of the curve. According to Towse (2001) 485 this is con-.rmed by empirical evidence. The effect of misinformation, on the other hand, is that a persistent overestimation of expected returns shifts the "supply" curve to the right and thus increases the total number of hours worked in the arts and therefore, also the number of artists.11 See also note 13 of Chapter 4.
12 The notion of an oversupply of artists can be found, among others, in Menger (1999) and Towse (2001). The use of terms like "oversupply" and "overproduc-tion" is not unproblematic. I return to the subject in sections 9.6 and 9.8.
13 Even when artists on average do not become poorer, poverty can be said to "in-crease" because extra available money leads to more poor artists per hundred thousand inhabitants.
14 The model I present in this section leads to largely the same result as Throsby (1994a), "A work-preference model of artist behavior". In Throsby"s model, artists tend to forsake consumer goods in order to make art. Pa.s.sionate artists want to maximize their time doing art at the expense of monetary gain and the leisure time that comes with working less. My approach is more general and more in line with neo-cla.s.sical thinking in the sense that artwork is not a goal in itself but a means for gaining non-monetary income like private satisfaction and recognition. Therefore, artists are inclined to forsake earned income for non-monetary income. In both approaches there is a survival constraint for artists. Although the approaches differ, the results can be the same. Moreover, it should be noted that my model refers to .ne artists and not necessarily to all artists, i.e., including applied artists. For other comments on Throsby"s model see Rengers and Madden (2000).
15 In theory it"s possible that more subsidization does not by itself lead to more artists, but that, instead, more artists and more subsidies have a common cause. For instance, the Netherlands is a relatively prosperous country. On the one hand, because the Dutch are prosperous and have enough money to do some-thing about the economic plight of artists, they are perhaps more likely to be concerned about the low incomes in the arts and may decide to increase subsi-dies for artists. On the other hand, because of the prosperity in the Netherlands, people may be more tempted to seek out a romantic alternative in the arts. Therefore, it is possible that higher subsidies do not necessarily create more artists, but that both have a common cause. Although it is not impossible that there are also general causes that contribute to both subsidization and high numbers, I nevertheless argue that subsidization primarily causes the large numbers of artists and overcrowding.
16 Throsby (1994a) in his work preference model for artists, presents indirect evi-dence for this thesis. Cf. Towse (2001) 485.
17 An international comparison of income .gures is tricky, as will be pointed out below. Nevertheless, in most areas, evidence more likely points in the direction of lower than higher incomes. (As noted before, visual artists in the Netherlands earn close to zero.) 18 Higher subsidies in mainland Western Europe are only partly offset by lower donation and spending levels. Therefore, I claim that subsidies result in a net-in-crease in numbers.
19 Cf. Ernst (1999). Measuring methods are almost always indirect; they often de-pend on a form of registration that only exists in a speci.c country, like registra-tions based on unique tax measure, for instance. The relativity of measures is also evident because of the fact that in one country there can be two ways of counting artists whose outcomes deviate by more than 100%. And, even though the registration of subsidies appears to be less complicated, the way in which subsidy levels are measured in different countries, especially subsidies by local governments, also differs too much to enable a comparison that makes sense. Cf. ibid. Ernst.
20 Even then, some reservation in the interpretation of results would be required because of the necessary limitation inherent in the de.nition of who is counted as an artist and who is not.
21 Dun, Muskens et al. (1997) 9.
22 Dun, Muskens et al. (1997).
23 Haanstra (1987) 22-5.
24 Dun, Muskens et al. (1997) 51 also point to the increase in the number of stu-dents but I see this as an independent variable that contributed to the attraction of the program instead of the other way around.
25 Although 40% of the money spent on the program continued to pour into the visual arts through other subsidies, overall spending by the ministries of social welfare and culture on visual art went down by more than 40%. Because spend-ing on the visual arts by local governments remained largely the same, the total reduction was approximately 20%.
26 Dun, Muskens et al. (1997).
27 Dun, Muskens et al. (1997) 46-8.
28 IJdens and Vet (2001) 31.
29 Cf. section 5.7.
30 In their direct effects, increased spending, donations, and subsidies tend to in-crease the number of artists, and they sometimes in.uence the income distribu-tion and average income .gures as well, while the willingness to work for a low income and therefore the inclination to become an artist remains the same. In the case of signals that are interpreted positively by artists, numbers increase and average income can decrease because artists become more willing to work for low incomes. The inclination of people to become artists increases. (The supply curve shifts to the right and starts to run more horizontally.) 31 This is also the opinion of Grampp (1989a) 114.
32 When it"s mainly highly successful artists who are pro.ting, average incomes are likely to show a slight increase.
33 See also O"Hagan (1998) 148-151.
34 Through the "Intermittent de Spectacle", income is transferred from the average employee to the average performing artist. Because of.cially there is no govern-ment involvement, it is seen as more of a collective donation than a subsidy. The effect is the same, however.
35 Although the amount of money the nea gives in the form of grants is relatively small, the grants are well-known and .ne arts graduates usually .rst turn to the nea.
36 Because they are scattered, the signaling effect of the Regional Boards is weak-er. The same does not apply to the Arts Council, but at the moment it does not give grants to artists. This situation is likely to change in the near future.
37 IJdens (1999) gives other examples. Nooy (1996) also points to the unintended effects of subsidization. He argues that unwanted effects can be reduced by a pe-riodic change in the system of subsidization. Another interesting example is pre-sented by Benhamou (2000) who discusses the effects of different public poli-cies on the temporary and self-employment situations among performing artists in France and England.
38 It is only recently (in 2000) that the Dutch government changed its policy. Using .nancial means, it installed an indirect numerus clausus by discouraging schools from accepting too many students. In France, England, and Germany meanwhile, the population of .ne arts students continues to grow unabated.
39 It seems possible that the pressure of the art lobby encourages colleges to in-crease their numbers even more in order to offer teaching jobs to the growing number of artists who cannot .nd proper employment elsewhere. In this re-spect, Menger (1999) 607 suggests that the training method both adapts and contributes to the "oversupply" of artists. Towse (1996b) 318-19 mentions the existence of a "training lobby" in the uk, which is engaged in "rent seeking", the seeking of pro.ts for the own group. It has a .nancial incentive to supply more training positions to students.
40 Among others, Towse (2001) and Linden and Rengers (1999) point to the fact that the parents of art students are better educated than those of other students in higher education. In the Netherlands, there are at least 40% more students at arts colleges with higher educated parents than in other schools for higher edu-cation. See Allen and Ramaekers (1999) 18. These .gure apply to all depart-ments in art colleges, including the relatively large applied apartments. If .gures were known for the students in only the .ne arts departments in comparison to other students the difference is likely to be even larger. This is con.rmed by re-search done by Lannoye. She questioned a group of art students, a majority of whom were .ne arts students. According to her data, almost 70% more art stu-dents than other students had parents with a higher than average education. Lannoye (2000) 37.
41 Authorities are often aware of the fact that professional artists use allowances to continue working but administrators often turn a blind eye to it. In the Netherlands, this became quite clear when it took several years to agree on the wik-program, a new subsidy program whose aim was to stop artists from de-pending on social bene.ts. In that period of time authorities were instructed to postpone a strict application of the of.cial rules for social bene.ts in the case of artists who made illegal use of social bene.ts. When illegal was no longer "ille-gal", it revealed just how widespread the illegal but tolerated utilization of bene-.ts had always been.
42 Benhamou (2000).
43 Presently in the Netherlands the so-called Melkert-baan plan is a good example.
44 Meulenbeek, Brouwer et al. (2000) 34. (The median yearly income in 1998 was 1200 Euro as one of the authors of the report, Natasha Brouwer informed me.) 45 Meulenbeek, Brouwer et al. (2000) 34. These artists earn less than the level of social security bene.ts in the Netherlands.
46 According to Janssen (2001) 342, who presents other references, in the nine-teenth century, most artists came from the middle cla.s.s (in most cases from a professional family) and most had also gone to public school or a university. Therefore, family .nancial a.s.sistance was possible and likely, but as far as I know, no empirical .ndings exist regarding the extent of family patronage. The same applies to the present family patronage. See also note 40.
47 See note 40.
48 Another explanation for the different social background of art students is that these students have more cultural capital and therefore more chances to succeed in the arts.
49 Throsby (1996) 232-4, Menger (1999) 602.
50 Wa.s.sall and Alper (1992).
51 Throsby (1996) 235.
52 These .gures are derived from Meulenbeek, Brouwer et al. (2000) 25-6.
53 Linden and Rengers (1999).
54 Wa.s.sall and Alper (1992) 191.
55 This is a trend not only in Britain as Benhamou (2000) demonstrates, but in sev-eral European countries. In France, however, due to the incentives of a speci.c system of social bene.ts, the trend has been most strongly towards shorter con-tracts rather than to situations of self-employment. See Benhamou (2000). In countries like the Netherlands and Germany, both developments occurred.