The importance of quality signals for the success of visual artists

The importance of quality signals for the success of visual artists Julie Hooft Graafland (10356398)

Jindi Zheng (supervisor) Bachelor’s Thesis; Economics

This document is written by Student Julie Hooft Graafland who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

Abstract

The primary art market consists of all artworks that are sold for the first time. Especially in the primary market, the value of an artwork and the quality of an artist are hard to determine, as there is usually a lack of information. Quality signals are used to credibly convey

information in the non-transparent market, reflecting upon the level of performance of artists and their work. Galleries establish art prices and influence the reputation of artists. As galleries play a dominant role in the industry, it is crucial for an artist’ career to get a gallery affiliation.

This study extends the knowledge about the role of quality signals in the primary art market by investigating how awards, subsidies, and reviews affect the probability of artists gaining a gallery affiliation after graduation. The main findings of this study suggest that both awards and subsidies positively relate to the probability of affiliating with a gallery. No significant results were found for the effect of reviews on gallery affiliation.

Quality Signals in the Visual Arts

In May 2015 Picasso’s masterpiece Les Femmes d’Alger was sold for $179 million in New York, setting a world auction record (www.cristies.com). In the secondary art market, which is mostly dominated by private dealers and auction houses and consists of artworks that have been sold at least once before, extremely high art prices are not unusual. However in the primary art market, where artworks are sold for the first time, art prices are usually much lower, as art purchases come along with more risk and uncertainty. How can a potential buyer know if a new artwork will be worth anything in a few years? According to Caplin (1989), the majority of artworks sold on the primary market will never appear on the secondary market. What makes these few artworks that do appear again on the market successful? This thesis will elaborate on the importance of quality signals for the success of artists in the primary art market. Do quality signals of artworks in the primary market predict

the success of the artist? Signals are defined as awards, reviews, and subsidies received by

the artists and I use gallery affiliations as proxy for being successful.

The demand for contemporary art has increased tremendously in the last decade; its market share in the global art market has gone up with almost 8 percentage points (Ehrmann, 2011). However, only a few artists earn a major part of total earnings in the primary sector (Rosen, 1983). This boosts the importance of the primary market, which consists only of new contemporary artworks (Velthuis, 2005). As there is relatively little known about the

functioning of the primary art market, this thesis hopes to create more clarity, which can be helpful for both art insiders as well as outsiders.

The art market is a very complex market, as it produces goods that carry greater symbolic than material value (Hirsch, 1972). This means that the economic worth of an artwork is primarily determined by its artistic value, constructed within the art world, rather than the materials used or the scarcity of supply (Beckert & Rössel, 2013).

The lack of information on the primary art market makes it difficult for potential buyers to assess the value and quality of an artwork. Intangible factors, such as status, determine the competitive performance of artists. But these intangible factors cannot be perceived directly by customers. Opinions of others, so called signals, are therefore used to determine quality (Washington & Zajac, 2005).Examples of common signals used in the art world are reviews, subsidies, and awards.

The concept of signals is commonly used in the art market and stems from the signalling theory. The signalling theory (Spence, 1973) states that conveying informative signals from one party to another can solve the problem of asymmetric information. In markets with asymmetric information, signals, which can be send by third parties as well as the artist him or herself, provide guidance in consumers’ buying behaviour. Quality signals are essential for the primary art market to function properly.

Relatively few studies have been done on the primary art market compared to the secondary market (Rengers & Velthuis, 2002). A reason for this might be the non-transparent character of the primary market and the limited availability of data. Moreover, most

economic studies focused on the buyer’s perspective rather than the artist’s perspective (Velthuis, 2011). Although there is some existing literature and findings on the primary art market, more research should be done to fully understand the complexity of this market.

In this project I use a database from the Rijksakademie, which consists of 428 artists and covers the time period from 1990 to 2009 and I show with a Probit model that quality signals are positively associated with the chance of gaining a gallery affiliation after

graduation. The associated influence is minor although significant and gives an insight into the complex mechanisms of the primary art market.

This research hopes to provide some new insights into the role of quality signals in the primary art market. It could fill the gap in literature about quality signals and the success of artists.

This paper is organised as follows: Firstly, the existing literature is investigated. Then the theoretical framework is explained followed by an elaboration of the research method. Next the analysis and results are presented, after which a discussion follows. Finally a conclusion is given.

Literary Review Signalling theory

Michael Spence (1973) originally proposed the signalling theory to solve the problem of asymmetric information. Connelly, Certo, Ireland and Reutzel (2011) generally described the signalling theory; the sender of a signal, the insider, has access to information that is not reachable for outsiders. The signal contains information about the quality or value of a specific good that can support a potential buyer with his or her purchase decision.

As art is considered an experience good, of which the value and quality are difficult to determine, information about the quality of art can be transferred through signals (Kirmani & Akshay, 2000). There are different classifications of signals but they all have in common that they credibly convey information from one party to another party, reflecting upon the level of performance (Graffin & Ward, 2010).

Literature on value signals stated that in order to provide useful information on estimating value a signal should be as specific and clear as possible (Dawar & Parker, 1994; Hertzendorf, 1993). Moreover, the attribution theory emphasised that credibility of the source of the quality signal determines whether galleries and potential buyers rely on value signals (Eagly & Chaiken, 1975). Signals from third parties are perceived as far more valuable when

compared to quality signals that stem from producers themselves, as third parties are able to independently rank, review and rate products (Dean & Biswas, 2001).

According to Wijnberg and Gemser (2000) the dominant value system in the visual arts is the expert-based value system, in which experts determine value by giving their opinion about a product. Experts gain from giving their opinion, either by financial means or an increase in status.

In short, a value signal has to be clear, explicit, and precise and send by an expert that is perceived to be credible (Gemser, Leenders, & Wijnberg, 2008; Leenders, Gemser, & Wijnberg, 2004).

Price determination on the primary market

The primary art market consists of artworks that are sold for the first time. This includes new artworks from unknown artists as well as artworks from well-known artists. However, the latter occurs very rarely. There is uncertainty on the primary market because information about the artist and the artwork are often missing, resulting in a limited number of buyers and a more unstable market (Schönfeld & Reinstaller, 2007).

There are two agents on the primary art market: The sellers, mainly art galleries, and the buyers. These two determine the nature of competition in the primary art market.

Auctions are exclusively used on the secondary art market. On the primary market, auctions would only add uncertainty to the economic value of contemporary art (Beckert, 1996). Here galleries usually make sales based on set prices, which, however, are subject to negotiation. These set prices keep artwork values to some extent predictable, as they are based on a set of tacit pricing rules (Velthuis, 2011).

According to Rengers and Velthuis (2002) galleries are price maximizers rather than profit maximizers. Tacit pricing rules help them to make pricing decisions along the different stages of an artist’ career. New works of unknown artists, for example, are priced based on

the price of similar works of art. This price usually starts low, as new artists have little price history. Prices increase due to an increase in an artist’ reputation, sales and time. Besides, medium and the size of an artwork also play a role in the pricing decision, as well as the competition between art galleries themselves (Schönfeld & Reinstaller, 2007). Galleries try to forecast the reaction of competitors and set prices in such a way that there is no incentive for other galleries to undercut their price (Schönfeld & Reinstaller, 2007). Price decreases are avoided in the art market, as this might be perceived as a negative quality signal.

Gallery affiliation

According to Marshall and Forrest (2011) art galleries have not only great influence on the monetary value of an artwork but also on the status or reputation of an artist. Because of the dominant role galleries play in the primary art market, gallery affiliation is crucial for an artist’ career development and success.

In the art world galleries select artists and signal them to the public (Velthuis, 2003). They promote their artists and help them with their artistic developments. Promoting artists can be done in various ways: exposing works in the gallery as well as promoting works among other galleries, encouraging purchases to potential buyers and getting an artists’ work into museum expositions and public shows. Moreover, galleries provide context for an artist’s work within the gallery. Through a gallery an artist is able to reach a wider and larger audience (Caves, 2000). The gallery tries to find buyers whose purchase supports the quality judgment of the gallery, as this boosts the artist’s reputation, the prices, and indirectly the galleries reputation as well (Schönfeld & Reinstaller, 2007). Gallery affiliation increases the chance for artists to get noticed by experts and thus the chance for getting more reviews and awards and ultimately a successful career (Bystryn, 1978).

The importance of status

Besides galleries, experts, institutions and media determine the artistic status of an artwork or artist. This status or reputation of an artist, which he or she has accumulated over the course of his or her career, is of central importance in the primary art market (Beckert & Rössel, 2013). It functions as a quality signal, enabling buyers to assess the economic value of artworks (Beckert & Rössel, 2013). The reputation of art galleries is determined by summing up all the individual artists they represented in the present and past (Schönfeld & Reinstaller, 2007). The reputation of the gallery and the artist are tangled and reinforce each other. Galleries with high reputation are known for selecting high potential artists and thus reducing the individual buyer’s risk of art purchase. Reputation enhances the functioning of the market (Schönfeld & Reinstaller, 2007).

At the same time, the status or reputation of the sender of a quality signal is

important. As quality signals are based on the judgements made by viewers of an artwork, the uncertainty of the artistic value of a work is actually the uncertainty regarding the correctness of these judgements on its artistic quality (Beckert & Rössel, 2013). Therefore signals from well-known and high status experts and galleries are more reliable than those with a low reputation.

Theoretical Framework

In this section the role of the independent variables awards, reviews, and subsidies will be discussed and three hypotheses are presented.

Awards

Institutions such as companies, art schools and newspapers can distribute awards. Awards provide the winner not only with a monetary reward but also with a certain amount of fame (Anand & Watson, 2004). Artworks from an artist that received an award are

perceived as of higher quality than that of his or her competitors. By constructing specific categories and orders awards contribute to the clarification of the value of an artwork or artist. Moreover, awards give a buyer examples and knowledge of what to look for in an artwork. Awards help to indicate the main players in a certain field (Anand & Watson, 2004).

Hypothesis 1: The number of awards positively correlates with an artist’ probability of

getting a gallery affiliation.

Reviews

Reviews from experts can signal the quality of a specific artist or artwork to the public. Good reviews can be useful in determining the value of art. However, a comment should be made with this type of signal. When an artist becomes popular or well known, he or she might attract more media attention of popular non-art magazines. These magazines contain reviews written by reviewers who do not have the reputation of being an expert in the primary art market and are more interested in the personality of the artist rather than the work he or she makes. The role of reviews in the primary art market is thus ambiguous (The

Economist , 2009).

Moreover, Eliashberg and Shugan (1997) argue that critics that write review papers are not opinion leaders but rather reflectors of public taste. Reviews acts as predictions

whether an audience will finally like the work but they cannot affect the success or failure. At the same time reviews provide visibility, which is important for artists, regardless if a review is good or bad (Shrum, 1991). The visibility of a review seems to be more important than the content it conveys.

Hypothesis 2: The number of reviews does not correlate with an artist’ probability of getting

Subsidies

Government and corporate subsidies are common in the visual arts as it increases social and cultural benefits. Subsidies are essential for artists as it gives them the opportunity to focus on their art practice (Rengers & Plug, 2001). Subsidies can finance artists’ events and activities or individual artists. In the Netherlands, for example, government subsidies are an essential source of income for visual artists1 (Rengers & Plug, 2001). Subsidies to artists, especially merit-based, are given based on several selection criteria, and thus contain a certain value signal, as not all artists receive subsidies. Subsidies therefore might indirectly influence the art market. Corporations’ motive to subsidise visual artists is often to gain a distinct reputation among other corporations, ultimately attracting more customers and thus generating profit (Creyer & Ross, 1997). As with governmental subsidies, corporations influence the visual arts environment by selecting artists for a subsidy (Witte, 2008).

Hypothesis 3: The number of subsidies, especially merit-based, positively correlates with an

artist’ probability of getting a gallery affiliation.

Research method Research setting

The main purpose of this thesis is to define and examine the extent to which quality signals (reviews, awards and subsidies) relate to the probability of getting a gallery

affiliation. As this research tries to explain relationships between variables, it falls under so-called explanatory research (Saunders, Lewis, & Thornhill, 2009).

The dataset used for this research contains specific data of artists, collected through files from the Rijksakademie and the Internet. This falls under archival research, containing

1 _{In 1995 government funding in the Netherlands received by visual artists made up 43% of their total income} earned in that year (Rengers & Plug, 2001).

secondary data analyses (Saunders, Lewis, & Thornhill, 2009). Although the dataset is longitudinal I will do a cross-sectional analyses.

Dataset

The dataset covers 428 alumni of the Rijksakademie van Beeldende Kunsten in Amsterdam, in the time period from 1990 to 2009. The Rijksakademie van Beeldende Kunsten focuses on developing talent in the fine arts. The two-year residency program offers visual artists a platform for further development of their work, podia for artistic presentations and connections to an international network. The Rijksakademie has been internationally regarded as a high quality education program (www.rijksakademie.nl).

Over 2000 beginning artists worldwide apply every year and only 25 get accepted for the program. The artists in the dataset belong to the more prominent category of artists within the primary art market, as they already received a quality signal by being accepted to the academy. Most of the artists at the Rijksakademie have already completed an art academy and have 3-5 years of autonomous art practice experience. About half of the participants come from abroad. The average age of participants is around 35 years. Before 1990, the Rijksakademie was a traditional art academy, significantly differently from the

internationally recognised academy it is today. Therefore the dataset covers only the time period from 1990 to 2009.

For each artist resident, data regarding sales, awards, reviews, exhibitions, subsidies, gallery affiliation and social demographics were collected. The database is originally based on the Rijksakademie artists’ documentation, regularly updated by staff members. This database was supplemented with information gathered over the Internet, using online databases such as artfacts.net and artnet.com. The database was recently updated for all the previous criteria until December 2015. It contains 428 artists and approximately 16.000 value signals.

Econometric model

The dependent variable in the Probit model is whether one gets a gallery affiliation after graduation. The regression model used is as follows: Pr (aff_after_grad=1) = Φ(β0 + β1

aff_before_grad + β2 reviews_total + β3 awards_total + β4 subsidies_total + β5 disc_vis + β6 disc_vav + β7 age + β8 age2 + β9 female)

In the formula, the dependent variable aff_after_grad is binary, Φ is the cumulative standard normal distribution, reviews_total, awards_total and subsidies_total are independent variables, and aff_before_grad, disc_vis, disc_vav, age, age2, and female are control

variables. (See appendix A for a description of all variables.)

All signals used as explanatory variables were gathered before the artists got a gallery affiliation (after graduation) and were therefore valid to use to predict the effect on gallery affiliation. Reviews from all types of different media sources were recorded, national as well as international, in art and non-art journals. Awards recorded came from institutions such as companies, art schools and newspapers. Need-based subsidies as well as merit-based

subsidies were taken into account.

The different art disciplines, fine arts, photography and audio-visual, are used as control variables, as some disciplines might be more popular to the public than others.

Gender is also included as a control variable, as according to Rengers and Velthuis (2002)

male artists are more successful than female artists as they generally sell more artworks. Lastly, age and age2 are added as control variables.

This research applies mostly to the main players in the primary art market. According to Joy (1996) artists, galleries as well as other intermediary institutions are considered as main players.

Analysis and results Descriptive statistics

Table 1 displays the descriptive statistics for the variables used in the regression. For the dummy variables the percentage of artists that had a value one for a particular dummy is given. There are more females in the sample (63%) compared to males (33%) and the average age of the artists is 36. Most artists have fine arts as their main artistic discipline (50%), followed by audio-visual (29%), photography (14%) and miscellaneous (7%). All three explanatory variables are positively skewed.

Table 1

Descriptive statistics (428 observations)

Variable Mean and std. dev.

Aff_after_grad 27.80% (0.45) Reviews_total 1.46 (2.72) Awards_total 0.63 (1.38) Subsidies_total 0.59 (1.32) Aff_before_grad 16.59% (0.37) Disc_vis 50.47% (0.50) Disc_vav 29.44% (0.46) Disc_pho 13.55% (0.34) Age 36.09 (7.92) Female 62.85% (0.48)

Table 2 shows the pairwise correlation coefficients of all variables and their significance level. Both the explanatory variables awards and subsidies display a positive correlation with gallery affiliation after graduation, at a significance level of 1%.

The variable reviews shows a minor positive correlation with gallery affiliation after graduation, but only significant at a 10% level. All explanatory variables positively correlate with each other, however, these correlations are relatively small so there is no threat of

multicollinearity. The only variables that are highly correlated with each other are age and age2, but this not a problem as the one is not a linear

transformation of the other by definition. Age and age2 also display a significant negative correlation with gallery affiliation after graduation and awards.

Table 2

Pairwise correlation coefficients

Aff_after _grad Reviews _total Awards_ total Subsidies _total Aff_befo re_grad

Disc_vis Disc_vav Disc_pho Age Age2 Female Aff_after_gr ad 1.00 Reviews_tot al 0.094* 1.00 Awards_tota l 0.22*** 0.18*** 1.00 Subsidies_to tal 0.14*** 0.30*** 0.15*** 1.00 Aff_before_ grad 0.14*** 0.086* 0.073 0.014 1.00 Disc_vis 0.020 -0.036 -0.098** -0.045 0.0021 1.00 Disc_vav -0.046 0.015 0.011** 0.0098 0.029 -0.65*** 1.00 Disc_pho 0.059 0.031 -0.019 0.071 -0.067 -0.40*** -0.26*** 1.00 Age -0.28*** -0.091* -0.18*** -0.074 -0.036 0.073 -0.025 -0.024 1.00 Age2 -0.25*** -0.084* -0.17*** -0.073 -0.020 0.073 -0.020 -0.027 0.99*** 1.00 Female -0.0085 -0.035 -0.065 -0.077 0.083* -0.065 0.062 0.022 0.063 0.056 1.00

Regression analysis

Table 3 displays the results of the Probit regression on the variable gallery affiliation

after graduation. The coefficients are the marginal effects on the binary dependent variable. I

carried out two regressions: One on the total sample, containing 427 artists, and one on a subsample, where I excluded all artists that had a gallery affiliation before graduation. In the total sample the control variable gallery affiliation before graduation is positively associated with the probability of gaining a gallery affiliation after graduation (see the extended table in appendix B). However, this result might be biased as some of the artists might be affiliated with the same gallery before and after graduation. Therefore, these artists are excluded in the subsample.

In both samples, reviews seem to have no correlation with the probability of gaining a gallery affiliation after graduation. However, awards are positively correlated with the

dependent variable at a significance level of 1%. This means that an increase in the number of awards by one is associated with an increase in probability of gaining a gallery affiliation of 4.5% in the total sample and 4.9% in the subsample. Subsidies show a smaller positive correlation, significant at a 5% level. In the total sample one extra subsidy is associated with a 3.2% increase in probability of affiliating with a gallery and this is 3.3% in the subsample.

Table 3

Probit regression on dependent variable Aff_after_grad

Total sample Subsample

Reviews_total -0.00028 (0.0072) -0.0012 (0.0072) Awards_total 0.045*** (0.015) 0.049*** (0.016) Subsidies_total 0.032** (0.015) 0.033** (0.015)

Control variables Yes Yes

Number of observations 427 356

Delta-method standard errors in brackets Subsample: all artists with aff_before_grad=0

Discussion Supported hypotheses

The regression results show that awards are, at a significance level of 1%, positively correlated with gallery affiliation, supporting the hypothesis that the number of awards positively correlate with an artist’ probability of getting a gallery affiliation. As, according to Rengers & Velthuis (2002), it is the main focus of galleries to select artists and affiliate with them, awards could be very helpful to get an indication of the new, talented artists in the market. Awards communicate value to galleries and are therefore important value signals. In this study reviews are not significant for gallery affiliation which supports the hypothesis that the number of reviews does not correlate with an artist’ probability of getting a gallery affiliation. However, the role of reviews in the art market should not be

underestimated. Reviews provide visibility for artists that might indirectly affect their reputation (Shrum, 1991). Moreover, reviews offer critical assistance for artists themselves (Wijnberg, 1995) and function as predictors for whether an audience will like an artist and its work (Eliashberg & Shugan, 1997).

Subsidies are positively associated with gallery affiliation at a significance level of 5%, which is in line with the hypothesis that the number of subsidies, especially merit-based, positively correlate with an artist’ probability of getting a gallery affiliation. Subsidies are an important source of income for artists and artists usually compete with one another to receive them (Rengers & Plug, 2001). By subsidising only a selected group of artists, institutions that offer subsidies influence the primary market selection environment (Wijnberg, 1994).

However, the dataset does not distinguish between merit- and need-based subsidies and therefore this part of the hypothesis can only be supported with literature and not the statistical results of this study. As the selection of merit-based subsidies is based on the

artistic skills of an artist rather than his or her financial needs, merit-based subsidies convey a greater qualitative judgement to potential buyers and galleries (Rengers & Plug, 2001). Although all three hypotheses are supported, the positive associations that the results showed are relatively small. This implies that there might be other determinants that influence the probability of getting a gallery affiliation after graduation. It should be noted that there might be omitted variable bias, as the three independent variables might be indirectly correlated with quality characteristics that increase the chance of getting a gallery affiliation. For example, consider the following scenario: Suppose galleries value innovation and therefore rather affiliate with avant-garde artists. Suppose likewise that art institutions that award prices, give awards to avant-garde artists2 because they praise their innovative ideas and expressions. In this scenario my results will show a positive association between awards and gallery affiliation, although the galleries might not have taken awards into account when deciding upon which artists to affiliate with.

Limitations

This research has several limitations. First, all artists in the sample are derived from the same source; the Rijksakademie van Beeldende Kunsten. The sample is not a random selection from the population and thus selection bias might be present. This means that significant results regarding the importance of quality signals cannot be generalised and apply to Rijksakademie alumni rather than to visual artists in other settings. As the dataset is based on secondary sources, there might be measurement biases. The data was double-checked with artfacts.net and artnet.com, but not all artists were present on these websites. Second, not all quality signals were taken into account. According to Velthuis (2003) the price of art, for example, is an important value signal about the quality of artists and their artwork. Moreover, art experience and years of education are not taken into account. Both

2 _{Avant-garde art is especially desired and highly valued by experts (Wijnberg & Gemser,} 2000).

could have a significant effect on the probability of affiliating with a gallery, as Throsby (2007) showed that artists with many years of education and experience are more successful in the art market than artists with little experience and education.

Third, success is measured in terms of gallery affiliation. Although gallery affiliation is very important for artists, career success could also be measured in terms of the number of sales or monetary value of sales. However, financial rewards seem to be less important in the art market. Thorsby (2006) argued that, in contrast with other sectors, artists are intrinsic rather than financially motivated.

Fourth, all signals are counted equally in the total number of value signals and not rated according to status and importance. An international award is valued as important as a regional award, which might not be correct. Also within a region some value signals might be more prestigious and more difficult to get than others. Moreover, some galleries might be harder to get an affiliation with than others, which is not taken into account.

The limitations mentioned above could help further research. It would be interesting to do a similar study with artists from other academies, and to include more value signals such as the price of an artwork.

Practical implications

The results of this study are applicable for artists, galleries and other intermediaries within the primary art market. Although the limitations mentioned before indicate that further research is needed, the results of this study do help to clarify the market. Moreover, this study gives outsiders a better understanding what determines and influences the success of artists and artworks.

Conclusion

In the primary art market there is little information known about artists and their artworks, making it difficult for potential buyers and galleries to determine the value of an artwork or artist. Information about the quality of art can be transferred through so called quality signals that credibly convey information from one party to another party, reflecting upon the level of performance. Quality signals are valuable for potential buyers as they reduce the purchasing risk and uncertainty. As it is the responsibility of galleries to select, educate, support and promote artists, quality signals also help galleries with the selection of talented artists. Because galleries play such a dominant role in the non-transparent primary art market, it is crucial for artists to affiliate with a gallery in order to become successful. Galleries can boosts artists’ reputation, which is important for artists in the primary art market. Credible quality signals, such as awards and subsidies, should have a positive effect on the probability of gaining a gallery affiliation, indirectly leading to a more successful career for artists.

This study confirms that both awards and subsidies have a significant and positive relation with a visual artist’s career, measured in terms of the probability of gaining a gallery affiliation after graduation. However, this relation is relatively small, indicating that there might be other important determinants influencing the probability of affiliating with a gallery. Reviews have no direct effect on gallery affiliation, which can be explained by the fact that reviewers are often seen as reflectors of public taste rather than opinion leaders. However, reviews provide visibility of artists, which should not be undervalued.

References

Anand, N., & Watson, M. (2004). Tounament rituals in the evolution of fields: The case of the Grammy awards. Academy of Management Journal , 47 (1), 59-80.

Beckert, J. (1996). What is sociological about economic sociology? Uncertainty and the embeddedness of economic action. Theory and Society , 25 (6), 803-840.

Beckert, J., & Rössel, J. (2013). The Price of Art. European Societies , 15 (2), 178-195. Bystryn, M. (1978). Art galleries as gatekeepers: the case of the abstract expressionist. Social

Research , 45, 390-408.

Caplin, L. (1989). The Business of Art (2 ed.). Englewood Cliffs : Prentiec Hall Trade. Caves, R. E. (2000). Creative Industries: Contracts between Art and Commerce. Cambridge:

Harvard University Press.

Connelly, B., Certo, S., Ireland, R., & Reutzel, C. (2011). Signaling theory: A review and Assessment. Journal of Management , 37 (1), 39-67.

Creyer, E., & Ross, W. (1997). The influence of firm behaviour on purchase intention: Do consumers really care about business ehtics? Journal of Consumer Marketing , 14, 421-432.

Dawar, N., & Parker, P. (1994). Marketing universals: consumers' use of brand name, price, physical appearance, and retailer reptation as signals of product quality. Journal of

Marketing , 58, 81-95.

Dean, D., & Biswas, A. (2001). Third-Party Organization Endorsement of Products: An Advertising Cue Affecting Consumer Prepurchase Evaluation of Goods and Services. Journal of Advertising , 30 (4), 41-57.

Eagly, A., & Chaiken, S. (1975). An attributional analysis on the effect of communicator charcteristics on opinion change: the case of communicator attractiveness. Journal of

Ehrmann, T. (2011). Artprice: the 2010 art market annual report – China winner of the past

decade. Retrieved from Artprice: Artprice.com

Eliashberg, J., & Shugan, S. (1997). Film Critics: Inflluencers or Predictors? . Journal of

Marketing , 61, 68-78.

Gemser, G., Leenders, M., & Wijnberg, N. (2008). Why some awards are more effective signals of quality than others: a study of movie awards. Journal of Management ,

34 (1), 25-54.

Graffin, S., & Ward, A. (2010). Certifications and Reputation: Determining the Standard of Desirability Amidst Uncertainty. Organization Science , 21 (2), 331-346.

Hertzendorf, M. (193). I'm not a high-quality firm-but I play on TV. Rand Journal of

Economics , 24 (2), 236-247.

Hirsch, P. (1972). Processing Fads and Fashions: An Organization-set Analysis of Cultural Industry Systems. American journal of sociology , 77 (4), 639-659.

Joy, A. (1996, January). Framing art: the role of galleries in the circulation of art. 1-63. Hong Kong.

Kirmani, A., & Rao, A. R. (2000). No Pain No Gain: A Critical Review of the Literature on Signaling Unobservable Product Quality. Journal of Marketing ,64, 66-79.

Leenders, M., Gemser, G., & Wijnberg, N. (2004). Effects of award competition on market competition in the motion picture industry. 6th World Media Economics Conference. Montreal, Canada.

Marshall, K., & Forrest, P. (2011). A framework for identifying factors that influence fine art valuations from artist to consumers. Marketing Management Journal , 21 (1),

Rengers, M., & Plug, E. (2001). Private or Public? How Dutch Visual Artists Choose Between Working for the Market and the Government. Journal of Cultural

Economics , 25 (1), 1-20.

Rengers, M., & Velthuis, O. (2002). Determinants of Prices for Contemporary Art in Dutch Galleries, 1992-1998. Journal of Cultural Economics , 26 (1), 1-28.

Rosen, S. (1983). The Economics of Superstars. The American Scholar , 52 (4), 449-460. Saunders, M., Lewis, P., & Thornhill, A. (2009). Research methods for business students

(5th ed.). Pearson Education Limited.

Schönfeld, S., & Reinstaller, A. (2007). The effects of gallery and artist reputation on prices in the primary market for art: a note. Journal of Cultural Economics , 31 (2), 143-153. Shrum, W. (1991). Critics and Publics: Cultural Mediation in Highbrow and Popular

Performing Arts. American Journal of Sociology , 97 (2), 347-475.

Spence, M. (1973). Job Market Signaling. Quarterly Journal of Economics , 87 (3), 355-374.

Stock, J. H., & Watson, M. M. (2012). Introduction into Econometrics (Third edition ed.). Pearson.

The Economist . (2009). How to Make Art History. Special report, The Art market.

Throsby, D. (2006). An artistic production function: theory and an application to Australian visual artists. Journal of Cultural Economics , 30, 1-14.

Throsby, D. (2007). Preferred work patterns of creative artists. Journal of Economics and

Finance , 31 (3), 395-402.

Velthuis, O. (2011). Damien's Dangerous Idea: Selling Contemporary Art at Auction. (P. Aspers, & J. Beckert, Eds.) Oxford: Oxford University Press.

Velthuis, O. (2003). Symbolic meaning of prices: constructing the value of contemporary art in Amsterdam and New York galleries. Theory and Society ,32 (2), 181-215.

Velthuis, O. (2005). Talking Prices: Symbolic Meanings of Prices on the Market for

Contemporary Art. Princeton University Press.

Washington, M., & Zajac, E. (2005). Status Evolution and Competition: Theory and Evidence. Academy of Management Journal , 48 (2), 282-296.

Wijnberg, N. (1994). National systems of innovation: Selection environments and selection processes. Technology in Science , 16 (3), 313-320.

Wijnberg, N. (1995). Selection processes and appropriability in art, science and technology.

Journal of Cultural Economics , 19, 221-235.

Wijnberg, N., & Gemser, G. (2000). Adding value to innovation: Impressionism and the transformation of the selection system in visual arts. Organizational Science , 2 (3), 323-329.

Witte, A. (2008). Professionalisering als paradoxale trend: Bedrijfscolleties in Nederland.

Appendix A

Table A

Description of variables

Aff_after_grad Gallery affiliation after graduation

Reviews_total Total number of reviews

Awards_total Total number of awards

Subsidies_total Total number of subsidies Aff_before_grad Gallery affiliation before

graduation

Disc_vis Discipline fine arts

Disc_vav Discipline audio-visual

Disc_pho Discipline photography

Age Age

Age2 Age squared

Appendix B

Table B

Probit regression on dependent variable Aff_after_grad

Total sample Subsample

Reviews_total -0.00028 (0.0072) -0.0012 (0.0072) Awards_total 0.045*** (0.015) 0.049*** (0.016) Subsidies_total 0.032** (0.015) 0.033** (0.015) Aff_before_grad 0.12** (0.051) Disc_vis 0.015 (0.050) 0.032 (0.054) Disc_vav -0.069 (0.057) -0.033 (0.060) Age -0.053*** (0.014) -0.066*** (0.015) Age2 0.00046*** (0.00016) 0.00062*** (0.00018) Female 0.026 (0.040) 0.063 (0.043) Number of observations 427 356

Delta-method standard errors in brackets Subsample: all artists with aff_before_grad=0