Mixed pricing strategies in museums: Examining the potential of voluntary contributions for capturing consumer surplus

Abstract

Existing pricing strategies at museums rarely reflect actual market demand. This article compares willingness to pay with paid entrance fees to establish consumer surplus and investigates whether voluntary contributions can serve to capture part of this surplus. A donation box is tested under various design experiments to identify best-performing contribution mechanisms. The proposed method is tested at the New Zealand Maritime Museum, with data collected via an intercept visitor survey and through electronic donation box counters. While the results indicate the existence of a significant consumer surplus, practically none of this surplus was gathered via voluntary contributions, and under normal circumstances less than 1% of museum visitors offered a donation. An animated donation box increased donations to 2% but remained well below the percentage of visitors that claimed to have a positive consumer surplus. The findings conclude that pay-what-you-want principles seem to have limited potential in mixed pricing strategy environments.

Keywords

cultural heritage donation behaviour experimental economics revenue management stated preferences willingness to pay

Introduction

It is widely accepted that museums – as part of the diverse field of cultural heritage – offer significant value to society in general and to their immediate surroundings in particular, with the effects of their cultural capital moving beyond immediate economic contributions. As noted by the Institute of Museum and Library Services (2008: 3): ‘…they connect the whole society to the cultural, artistic, historical, natural, and scientific understandings that constitute our heritage; and they collect and conserve tangible objects – animate and inanimate – for the benefit of future generations’. Within a total value framework, Martin (1994) identifies two general classes of benefits: use value and non-use value. As the term suggests, the former relates to values that are directly attributable to museum visitors (i.e. the ‘users’). Non-use value, on the other hand, refers to the benefits assigned to a museum that are unrelated to (current) visitation, including option, bequest and existence value.

This study focuses specifically on use value of current museum visitors, in order to estimate the total utility received from a visit. Such use value – alternatively denoted ‘value for money’ by Ashworth and Johnson (1996) – can be quantified as the sum of the entrance fee and additional consumer surplus (Seaman, 2006). Since museums are generally not traded on a free market, and consequently lack an indicative pricing mechanism, developing knowledge of the underlying use value inclusive of consumer surplus is important for revenue management and public funding decisions, especially in the current climate of increased privatization (Ponzini, 2010; The Economist, 2002). Decreasing government levies would increasingly expose cultural institutions to free market economics and necessitate a shift towards user-pays principles and alternative sources of income.

The main purpose of this study is, therefore, to gain an understanding of the actual use value of a museum visit – over and above any existing entrance fee – via the use of a stated preference contingent valuation method (CVM) (see e.g. Mourato et al., 2004; Sanz et al., 2003; Thompson et al., 2002). Knowledge gained in this respect will provide a better understanding of the actual economic value of museum visits, which can then be used for price setting purposes and discussions on government levies. Secondly, the research will link hypothetical-stated willingness to pay (WTP) values with the revealed behaviour of both entrance fee payments and voluntary donations to the attraction and investigate whether part of the consumer surplus can be redistributed to the attraction via voluntary contributions. Thirdly, such voluntary contributions are of further interest since studies on charitable giving and the underlying psychological motivations suggest that innocuous design alterations might increase revenue (e.g. Agerström et al., 2016; Fathi et al., 2014; Perrine and Heather, 2000). Therefore, informed design principles might optimize charitable donations – in terms of revenue generation – and bring them closer to covert WTP values, leading to a redistribution of income from the customer to the provider.

In order to test these effects, a visitor survey was conducted at the New Zealand Maritime Museum in Auckland. The museum operates as a private trust while receiving approximately 55% of their total budget through a levy provided by the Auckland Council Regional Amenities Funding Act. As a result, Auckland residents are granted free entry. Wider New Zealand residents and international visitors need to pay an admission fee of NZ$20 (standard adult fee). This differentiated price structure can potentially offer interesting data on differences in consumer surplus and the subsequent likelihood of voluntary financial contributions to the museum via the donation boxes present. Furthermore, since admissions currently only amount to 9% of the yearly museum revenue, there is interest in increasing the revenue gained through visitation (Voyager New Zealand Maritime Museum, 2017).

The article is structured as follows. Staring from a broader perspective, the second section discusses value assessment of cultural assets, specifically in terms of use value, and potential bias present in the CVM. The third section provides a brief discussion on voluntary contribution systems and general behavioural theories on charitable giving. The fourth section discusses the methodological setup for the data collection and analysis and the fifth section presents the empirical results. Finally, the last section offers a conclusion and general recommendations.

Estimating the use value of museums

As a non-market, demand-based, approach, CVM has received significant attention in the valuation of cultural properties (for an overview, see Noonan, 2002, 2003). The CVM is a survey method that approaches value through a hypothetical WTP or willingness to accept question. Due to its hypothetical nature, the method can provide utility values for public resources and intangible goods, for both the users and the wider population (Venkatachalam, 2004). Seaman (2006) and Snowball (2013) mention how CVM is ultimately concerned with uncovering a demand curve for a specific asset (within the subject of this study: cultural heritage). While such demand curve could be established for use and non-use – thereby approaching total economic value (see e.g. Bedate et al., 2009; Queensland Government, 2009; Tohmo, 2004) – the method is equally applicable for a focused approach on use value. For example, Arbrecht (2014) applied a CVM to estimate use value for both the Vara Konserthus and the Nordic Watercolour Museum and found values for the latter museum of €9.20 for locals and €10.10 for non-locals, which, aggregated, led to a total value of €1,480,350. Lampi and Orth (2009) adopted a CVM survey with visitors to the Museum of World Culture in Sweden – prior to the introduction of an entrance fee of SEK40 in 2007 – and established a mean WTP of SEK56.6.

Such research on WTP – as reflecting use value – has typically uncovered a significant consumer surplus over and above existing entrance fees. Since ticket prices often do not reflect market demand, but are chosen through a combination of tradition, benchmarking, price of substitute goods (e.g. other recreational attractions) and political decisions (O’Hagan, 1998), understanding of the underlying demand curve can benefit management choices. Importantly, it can identify the optimal point between an increase in revenue and a decrease in tourism demand (Mourato et al., 2004), therefore, being a useful tool in situations where visitor numbers need to be limited for preservation purposes (Montemagno, 2002). Furthermore, knowledge of consumer surplus and WTP might assist revenue generation through discriminatory pricing strategies, over and above the common practices of offering reduced fees for specific socio-economic and/or geographic groups (Huntington, 1993).

While previous studies indicate the usefulness of WTP studies for heritage management and revenue generation, hypothetical bias is a concern, and one of the main critiques levied at WTP studies (Diamond and Hausman, 1994). Hypothetical bias originates from the fact that stated preference techniques value attitudes without having respondents face an actual spending situation, most often leading to an overstatement of WTP (Loomis, 2014; Zawojska and Czajkowski, 2017). A variety of hypotheses have been postulated to explain potential differences between actual and stated values, leading to different suggested solutions – either ex ante or ex post. As a general principle, respondents need to consider a question to be consequential – that is, have a relevant outcome that is potentially influenced by the response given to the question (Carson and Groves, 2007). Diamond and Hausman (1994) agree that bias is more likely to occur when there is a lack of actual consumer preferences, which would imply that the WTP for actual use (i.e. visitor value) can be considered more reliable than option and non-use values.¹ Within political science, Hamlin and Jennings (2011) note, however, that preferences might be expressive, rather than instrumental, specifically when individual weight on the final outcome is seen as marginal. For instance, within a heritage valuation context, one might overstate personal WTP to reflect identity or moral values, without considering a direct outcome effect in terms of an increase in price.

While hypothetical bias is a concern, various solutions have been proposed and, as noted by Snowball (2013), carefully designed and interpreted non-market valuation studies have been shown to be generally reliable and valid. A meta-analysis by Murphy et al. (2005) found a ratio of hypothetical to actual WTP median values of just 1.35, while Zawojska and Czajkowski (2017) note that studies satisfying the requirements of incentive compatibility suggest a close similarity between stated and actual preferences. However, neither of these analyses focuses on cultural assets. This article proposes a specific method to reveal hypothetical bias within WTP studies for museums, by adopting a revealed preference approach through a cash validity study (see e.g. Champ et al., 1997), related to the admission strategy.²

Voluntary donations and behavioural insights into charitable giving

O’Hagan (1995) notes that the majority of payment mechanisms at cultural heritage attractions are of a hybrid nature, combining voluntary contributions with either a free admission policy – making it a pay-what-you-want pricing strategy – or a compulsory admission fee – in which case the voluntary contribution is seen as a true additional donation.³ Even in cases where a compulsory admission fee is levied, non-personalized, fixed pricing invariably leads to the existence of consumer surplus which might partly be captured through additional donations – with real WTP potentially being the sum of the admission fee and the additional donation in this case. In order for donations to fulfil such potential, it is necessary to optimize additional voluntary contributions in order to help bring revenue closer to the real, individual WTP.

A large strand of research has devoted attention to the underlying motives for voluntary payments, largely distinguishing between altruistic, egoistic/strategic and normative motives (Natter and Kaufmann, 2015). Pure altruism is considered when all utility to the donor is derived from a selfless concern for others (Becker, 1974), while impure altruism implies underlying motives such as a warm glow effect or the achievement of moral satisfaction and favourable self-conception, specifically in cases where the individual donor identifies strongly with the organization (Andreoni, 1990; Bhattacharya et al., 1995; Kahneman and Knetsch, 1992; Klein and Goethals, 2002; Savas, 2016). Bertacchini et al. (2010) found such intrinsic – either pure or impure – motives to be more effective drivers for charitable giving than purely providing fiscal incentives. Consumer contribution via voluntary payments can be further explained by egoistic/strategic motives such as signalling social status, receiving recognition or acquiring social influence, thereby primarily maximizing the donor’s utility. In this sense, prosocial behaviour suggests a payoff for generosity through potential future interactions (Fathi et al., 2014; Natter and Kaufmann, 2015; Roberts, 1998). Such underlying motives are particularly evident in museum donations in the form of corporate social responsibility programmes, where philanthropy can become a marketing instrument in its own right (Porter and Kramer, 2002). Other researchers have adopted social norms to frame voluntary payments and incorporate the actions of others as affecting intrinsic motives. The normative account originates from norm psychology and states that people are motivated to behave in a similar fashion to others in their surrounding environment, leading to conformity with shared local and global norms (Agerström et al., 2016; Deb et al., 2014; Kallgren et al., 2000). Specifically, within tipping and pay-what-you-want environments, social exchange norms are sometimes invoked, falling within equity theory and emphasizing the concern of fairness and reciprocity in relationships as a driver (Azar, 2007; Kim et al., 2014).

Understanding various motives for the frequency and size of voluntary contributions, combined with studies on consumer-specific and product-specific success factors (see e.g. Natter and Kaufmann, 2015; Savas, 2016), has further informed a range of field experiments – specifically concerning charitable donations – designed to understand how design factors can significantly influence voluntary contributions in terms of frequency and/or average donation size. Focusing specifically on design factors that are related to the above-mentioned normative theory, based on earlier findings that people behave more generously when being gazed upon, Bateson et al. (2013), Ekström (2012) and Fathi et al. (2014), among others, showed how even artificial images of ‘watching eyes’ around the donation box significantly increased average donations. Agerström et al. (2016) focus on descriptive social norms and found a stronger influence of local norms by manipulating the information embedded in a charity brochure. Their findings show similarity with the results of Schultz’ (1998) experiment on curbside recycling which found that including feedback on the recycling behaviour of other residents led to the largest behavioural improvement. Perrine and Heather (2000) also studied the effectiveness of message type and found a significant relationship between the average donations and the vividness of information, specifically including a picture instead of a textual plea for help. This coincides with insights from the UK Behavioural Insights Team (2013) about the need to attract attention, as well as reward charitable behaviour by encouraging reciprocity, in order to activate social exchange norms.

Research undertaken in the field of voluntary contributions – and specifically charitable donations – can thus inform the management of cultural heritage attractions. An interesting field of study lies in identifying whether the effects extend from the experimental laboratory settings and charity giving to voluntary contributions for museum visits, which are most often collected through impersonal donation boxes. This study will use a field experiment setup to establish visitors’ WTP, independent from the admission fee, investigate the relation between stated WTP and additional voluntary contributions and test the effects of various normative designs on these contributions.

Methodology

Survey procedure and instrument

Between 2 November 2016 and 27 November 2016, a visitor intercept survey was conducted in the New Zealand Maritime Museum in Auckland. Since only actual visitors were surveyed, the calculated economic value of the cultural site was limited to its actual use value. While the study, therefore, did not take full advantage of the possibilities of stated preference methods, it did avoid overestimation due to hypothetical bias (see e.g. Bedate et al., 2009; Diamond and Hausman, 1994). Visitors were approached by a research assistant at the end of their museum visit and invited to participate in a self-completed questionnaire. A self-completion method was preferred since questions about WTP and donation behaviour could potentially lead to social desirability bias. It should also be acknowledged that there might be a difference in WTP values depending on whether respondents are approached prior to or after a visit to a cultural attraction, although such effect was not found by Mourato et al. (2004). While, in reality, consumers have to decide on the acceptability of an admission fee ex ante, the requirement to collect actual data on donation behaviour in this study necessitated the use of an ex post questionnaire.

The survey instrument was presented in English and consisted of two versions: a questionnaire for Aucklanders and a questionnaire for non-Aucklanders. Although most of the questionnaire content was similar, a few changes were required, given the differentiated pricing structure of the New Zealand Maritime Museum whereby locals are allowed free admission. The questionnaires consisted of four main parts: a section on the current visit and visitor experience, a section on WTP for museum admission, a section on donation behaviour and a final section collecting personal details.

The WTP question was preceded by a ‘cheap talk’ script, as suggested by Carlsson et al. (2005) as a way to reduce hypothetical bias and was formulated as: ‘Imagine the museum increases the entrance fee. What would be the maximum entrance fee you would be willing to pay in order to still visit the museum?’ While contingent valuation studies typically describe hypothetical WTP scenarios in terms of a change, directly relating the WTP amount to an existing product/service has commonly been adopted in heritage and museum studies (see e.g. Arbrecht, 2014; Lampi and Orth, 2009; Mourato et al., 2004). The WTP values were presented via a payment card vehicle, as also used by Mourato et al. (2004), with eight levels. Recognizing the difference in baseline admission fee – with Aucklanders enjoying free entry and non-locals paying a standard fee of NZ$20 – as an important starting point effect, for local Aucklanders, the suggested price levels were not willing to pay anything – NZ$5 – NZ$10 – NZ$15 – NZ$17 – NZ$20 – NZ$25 – NZ$30, while for non-Aucklanders, the payment card contained the values: not willing to pay a higher amount – NZ$22 – NZ$25 – NZ$30 – NZ$35 – NZ$40 – NZ$45 – NZ$50. If respondents were unwilling to pay a higher admission fee, an additional question was asked in order to understand whether or not the lack of WTP was the result of a protest bid, that is, bids that are reflective of an objection to the CVM, rather than an actual measure of value (Grammatikopoulou and Olsen, 2013).

The next section asked respondents whether they had offered a voluntary contribution via the museum’s donation box via a binary question. Since it could be assumed that these questions are subject to social desirability bias, the question was pre-loaded via the following statement:

There is a donation box in the Maritime Museum in case visitors want to offer a voluntary donation. People can find donation boxes unnecessary since there is already a specific entrance fee. Did you put any money in the donation box?

For affirmative responses, respondents were asked to indicate the donated amount in NZ$ via an open-ended question. Respondents who had not contributed were asked why.

Experimental setups of donation box design

During the period of the visitor intercept survey, four different experimental setups were tested with regard to the donation box. The donation box was located near the only entrance/exit of the museum, prior to the visitor intercept survey and out of sight of the research assistant. At the start of each day, the donation box was loaded with NZ$20, keeping both the total amount and the currency denominations constant over observation days. At the end of each day, the final donation amount was determined to obtain a record of total daily donations. Furthermore, an infrared beam within the donation box allowed for a precise count of donation numbers.

The period from 2 to 6 November served to establish a baseline, presenting a non-modified design with the generic message: ‘With your help we can do more. Thank you for your generous support’. From 9 to 13 November, the effect of a ‘watching eyes’ poster (see e.g. Ekström, 2012; Fathi et al., 2014) was tested. The poster was located right next to the donation box and included the innocuous message: ‘Please keep an eye on your children around the waterfront’. During the third stage, from 16 to 20 November, the effect of social norms (see e.g. Agerström et al., 2016; Schultz, 1998) was empirically tested by adapting the generic message on the donation box to include the local norm as found during the first week of observation. However, due to the low amount of actual donations in this period, direct reference to voluntary contributions of other visitors could not be made. Instead, the message was based on an alternative survey question which asked respondents how important community support to cultural heritage is. Therefore, the final norm reference became: ‘74% of our visitors find community support for museums important. Help us achieve more’. Finally, the last experiment which ran from 23 November to 27 November was primarily based on the idea of reciprocity as a motivation for voluntary contributions (Azar, 2007; Kim et al., 2014; UK Behavioural Insights Team, 2013) while also acknowledging the findings of Perrine and Heather (2000) and the UK Behavioural Insights Team (2013) that vividness of information can increase donations. As described by Porter (2011: 133): ‘Receiving something gratis conjures a sense of indebtedness and incites deep-seated feelings of reciprocity’. To activate such motives, an interactive, animated Lego^®-display was built within the donation box, activated by movement around a sensor attached to the exterior of the donation box.

Sample descriptives

A total of 453 completed surveys were collected. While no visitors refused to participate upon approach by the research assistant, language barriers resulted in automatic non-participation. Information on such missing samples was not collected. Of the 453 respondents, 52.5% were male and 47.5% female, while 29.6% were local from within Auckland and 70.4% were non-local. This is important given the pricing structure of the museum. Table 1 gives an overview of the responses collected per experimental setup, as well as the total number of billable visitors passing through the museum gates during this same period and counted via the museum’s visitor management system.

Table 1.

Number of responses (n) per experimental setup.

Sample	Experimental design
Sample	Baseline	Watching eyes	Social norms	Reciprocity
Locals	27	42	40	25
Non-locals	65	89	77	88
Total sample	92	131	117	113
Total billable visitors	988	1617	1938	980

Of the local respondents, the large majority (42.1%) lived in Central Auckland, in close proximity to the New Zealand Maritime Museum; 17.4% originated from Eastern Auckland and 13.2% from Southern Auckland. North Shore and Western Auckland were equally represented by 17.4% each, while the far North of the metropolitan region was least represented by 5.8%. The 319 non-local respondents represented 32 different nationalities; the United States (22.6%) and Australia (22.3%) were best represented, followed by New Zealand (11.0%), the United Kingdom (10.0%), Canada (6.6%), Germany (4.4%) and Switzerland (3.4%). All remaining nationalities accounted for less than 3.0% individually and a combined total of 19.4%. It seems clear that, due to language barriers, Asian nations are underrepresented in the sample.

Table 2 offers a summary of descriptive variables that might serve as explanatory variables to understand differences in WTP. In the total sample of 453 respondents, the ages of 60 and above (24.8%) and 20–29 (22.6%) were best represented, followed by the categories 30–39 (18.4%), 50–59 (13.9%), 40–49 (10.6%) and finally the youngest age group of 16–19 (9.7%). The age distribution differed between locals and non-locals, with the latter accounting for a higher number of relatively older visitors, while the youngest age groups were predominantly local. Information was also collected on yearly household income. While this question had a large non-response of 42.2%, the collected responses indicated an average income level of NZ$96,519.23 (standard error = 9483.386), comparable to the 2016 New Zealand mean of NZ$95,898 (StatsNZ, 2017), although non-local, specifically international, visitors were more likely to have above-average incomes. As could be expected, non-local were predominantly first-time visitors (92.1% as compared to 58.2% of locals) and locals had a higher tendency to be accidental visitors, being motivated by the lack of admission (31.6%), social motives such as spending time with friends and family (31.6%) or even the weather outside (9.0%). In comparison, non-local visitors predominantly came specifically out of an interest in museums (55.3%), although more random motivations such as being in the area (28.1%) cannot be discounted. Finally, when asked about the importance of financial support by the community – via public financing – for the protection of cultural heritage, a large majority found this very important (69.4% of locals and 67.9% of non-locals) to important (29.9% of locals and 31.8% of non-locals).

Table 2.

Summary of explanatory variables.

Variable	Categories	Local (n = 134; %)	Non-local (n = 319; %)
First visit	Yes	58.2	92.1
	No	41.8	7.9
Motivation*	A general interest in museums	37.6	55.3
	To see a specific exhibition	9.0	11.5
	In the area or passing by	33.1	28.1
	To learn about the local area/destination	15.0	24.6
	It was recommended (word of mouth, guide book, etc.)	16.5	22.0
	Curiosity	28.6	25.9
	As part of tour, a ‘must see’ visitor attraction	3.8	5.1
	Saw signage or advertising	0.8	6.4
	Spending time with friends and family	31.6	6.1
	It’s free/it’s good value for money	31.6	5.8
	Because of the weather outside	9.0	4.2
Support	Very important	69.4	67.9
	Important	29.9	31.8
	Not important	0.7	0.3
	Very unimportant	0.0	0.0
Gender	Female	53.7	44.8
	Male	46.3	55.2
Age	16–19	16.4	6.9
	20–29	26.1	21.1
	30–39	23.9	16.0
	40–49	19.4	6.9
	50–59	5.2	17.6
	>59	9.0	31.4
Income	Mean yearly household income in NZ$	$96,519.23	$107,577.60

*Multiple response questions.

Results

WTP and consumer surplus

A large portion of respondents indicated that they were prepared to pay a premium over and above the current admission fee. For local residents, 26 respondents (19.7%) refused to pay any entrance fee. Asked about the reasons for objection via a multiple response question, a majority of 72% stated that they had already paid for entrance through taxation, while 44% also agreed with the statement that the museum should be free as part of local heritage and identity. Only one person (4%) stated a lack of funds as the main reason for a non-positive WTP. From these motives, it can be deduced that the zero bid only accurately reflects utility for one person, while the other respondents clearly attach a positive value to the museum but object to the suggested payment mechanism. Since it is a concern of the payment card elicitation method that respondents limit their WTP to the listed upper bound, it is also important to identify the number of respondents willing to pay the maximum suggested entrance fee. Only three local respondents (2.3%) chose the upper limit of NZ$30.

Prior to establishing WTP, non-locals were explicitly asked about the museum entrance fee that had been paid. While standard adult admission is NZ$20, 11.6% of non-locals entered the museum free of charge, with another 0.6% paying NZ$15, 33.1% paying NZ$17 – the standard seniors/students entry – and 1% reported paying NZ$18. The remaining 53.7% paid the full admission of NZ$20. Taking these starting values into account, 81 non-local respondents (25.6%) were unwilling to accept a higher admission fee. The main reason given for rejection adhered to the principle that heritage should be offered at a low price (41.6%). For 24.7% of non-local respondents, the museum was not worth a higher entrance fee, while 22.1% stated a lack of money. Eleven non-local respondents (14.3%) furthermore stated reasons of fairness, with the difference between the charges for residents and non-residents perceived as too large. Finally, four people (4.7%) stated it was the government’s responsibility to fund heritage, 2.6% felt the educational scope of the museum necessitated free access and one person (1.2%) felt the museum was already currently overpriced. At the other end of the payment card listed bids, four people (1.3%) chose the maximum amount of NZ$50.

Hackl and Pruckner (1999) present various methods for estimating WTP under payment card elicitation. In this article, we adopt the simple minimal legal WTP model, interpreting the chosen payment card values as a ‘legal’ commitment to pay. The mean WTP can then simply be obtained as ∑(A_i × P_i) with A_i being the different values of the payment card and P_i the respective frequencies in the sample. Since true WTP might actually lie between the chosen value and the next value on the payment card, this minimal legal WTP can be considered a lower bound value (Tian et al., 2011). Average local WTP after elimination of protest bids equalled NZ$14.55 (standard error = 0.605), with a median of NZ$15. In the non-local sample, the average WTP after elimination of protest bids was NZ$26.55 (standard error = 0.450), with a median of NZ$25. It is worth noting that the other main museum in Auckland, the Auckland War Memorial Museum, charges an entrance fee of NZ$25 to international tourists, hence, possibly serving as an anchor for the choice in WTP.

Using these results, aggregate demand curves could be constructed, using the respective local and non-local visitor totals as index in order to improve comparability of the two groups. It can be seen from Figure 1 that the differentiated price regime for non-locals creates a drop-off prior to the standard NZ$20 entrance fee due to visitors who entered the museum at reduced cost and were unwilling to pay a suggested higher amount. Comparing current entrance fees with the demand curves of Figure 1 suggests the existence of a positive consumer surplus (calculated as the difference between ticket price paid and stated WTP and exclusive of protest bids) of, on average, NZ$14.55 (standard error = 0.605) for locals – similar to average WTP due to the zero admission fee) and NZ$9.51 (standard error = 0.480) for non-locals.

Figure 1.

Demand functions for locals and non-locals – excluding protest bids.

A binary logistic regression⁴ tested the effect of the explanatory variables of Table 2 on the willingness of respondents to pay a higher amount than the current admission fee (i.e. whether or not respondents exhibited a positive WTP). In order to simplify the analysis, age was recoded into dummy variables for 16–29 years old and 30–49 years old. Furthermore, to reduce dimensionality in the 11 motivational categories, a multiple correspondence analysis (MCA) was performed (see also Brida et al., 2016). MCA extracts latent factors from a contingency table and is, therefore, an appropriate alternative to factor analysis in the case of categorical variables. Results of the MCA lead us to retain five dimensions with a cumulative variance of 54.7%. Table 3 reports the coordinates of the variables within the latent dimensions. The first motivation (situational) is primarily constructed through a positive answer on ‘Spending time with friends and family’, ‘Because of the weather outside’ and ‘It’s free/It’s good value for money’. This motivation stands out from the other dimensions as being less focused on the actual product offer of the museum and can be linked to a more situational interest. The second dimension (strong interest) covers positive answers on ‘A general interest in museums’, ‘To learn about the local area/destination’, ‘Curiosity’ and ‘Saw signage or advertising’. Upon reading, this dimension seems to cover the strongest intrinsic motivations for visiting the museum by respondents with an active interest and curiosity towards the destination. An interpretation for dimension 3 (recommended) is provided by positive answers on ‘It was recommended’ and ‘As part of a tour’, with the former being linked quite exclusively to this factor. The fourth motivation (tour) is defined strongly by the category ‘As part of a tour’, which might further be linked with weather events ‘Because of the weather outside’, potentially suggesting that people are more likely to enter an organized tour to an indoor museum when they are confronted with adverse weather conditions. Finally, dimension 5 (walk-in) combines ‘As part of a tour’, ‘Saw signage or advertising’ and ‘In the area or passing by’ might mostly be related to tourists who happened to pass by the museum and enter without any real prior expectations. A score for each respondent on these five MCA dimensions was calculated to further use as variables within the regression analysis.

Table 3.

Factorial coordinates of MCA.

Categories	Dim 1	Dim 2	Dim 3	Dim 4	Dim 5
General interest (No)	0.312	−0.528	0.009	−0.204	0.470
General interest (Yes)	−0.322	0.544	−0.009	0.210	−0.485
Specific exhibition (No)	−0.017	0.156	−0.011	0.051	0.190
Specific exhibition (Yes)	0.146	−1.316	0.091	−0.427	−1.600
In the area (No)	0.005	0.074	0.405	−0.194	−0.249
In the area (Yes)	−0.012	−0.181	−0.985	0.473	0.606
Learn about local area (No)	0.093	−0.228	−0.158	0.100	−0.036
Learn about local area (Yes)	−0.341	0.838	0.581	−0.365	0.133
Recommended (No)	−0.051	−0.053	−0.349	0.015	−0.134
Recommended (Yes)	0.203	0.211	1.389	−0.058	0.532
Curiosity (No)	−0.068	−0.356	0.186	0.089	0.048
Curiosity (Yes)	0.192	1.000	−0.521	−0.250	−0.134
Part of tour, ‘must see’ (No)	−0.028	−0.003	−0.048	−0.164	0.025
Part of tour, ‘must see’ (Yes)	0.570	0.052	0.979	3.377	−0.518
Signage/advertising (No)	0.058	−0.074	0.009	0.005	−0.086
Signage/advertising (Yes)	−1.194	1.525	−0.192	−0.103	1.763
Spend time with F&F (No)	−0.282	0.010	−0.012	0.123	−0.057
Spend time with F&F (Yes)	1.815	−0.067	0.080	−0.789	0.367
Free/value for money (No)	−0.236	−0.147	0.084	0.031	0.097
Free/value for money (Yes)	1.543	0.961	−0.551	−0.203	−0.637
Weather (No)	−0.133	−0.031	−0.035	−0.102	−0.016
Weather (Yes)	2.269	0.530	0.592	1.754	0.275

Note: MCA: multiple correspondence analysis. Explained variance Dim 1 = 12.9, Dim 2 = 12.1, Dim 3 = 10.8, Dim 4 = 9.6, Dim 5 = 9.2.

Table 4 gives an overview of the binary logistic regression results. Due to a large number of missing cases on yearly household income, regressions were tested inclusive and exclusive of this indicator. Collinearity diagnostics showed no reason for concern, with all Variance Inflation Factor (VIF) values between 1 and 3.1, mean VIFs of 1.588 and 1.125 and tolerance statistics well above 0.2. The models that include yearly household income had a higher predictive power and will, therefore, be the ones analysed in more detail. Broadly, though, all of the models lack sufficient predictive power, with a best-case Nagelkerke R² of just 0.184 for the local sample and 0.167 for non-local visitors. Furthermore, the log-likelihood (LL) difference between the constant-only model and the model inclusive of all predictors is non-significant in the local sample (−2LL = 76.250, p = 0.393) and the non-local sample (−2LL = 153.014, p = 0.36). While, understandably, the difference between locals and non-locals had a significant effect on whether or not a positive WTP and associated consumer surplus were found, additional effects are barely noticeable. While we might have expected motivated museum visitors – as indicated through specific reasons to visit such as a general interest in culture or an interest in a particular exhibition – to have a higher WTP (e.g. Apostolakis and Jaffry, 2013), this was not reflected in the data. The hypothesis that visitors who primarily enter the museum because of situational factors – specifically free access for locals – or to spend time with friends and family are likely to exhibit a lower WTP was not supported. The only relevant motivational factor was found in the non-local case for people who visited the museum based on recommendations or by following a tour, exhibiting a lower probability of having a positive WTP (β = −0.521, p = 0.001). On the other hand, non-local visitors between the age of 30 and 49 were found to be significantly more likely to be willing to pay an entrance fee above the current level (β = 1.524, p = 0.022). There also seems to be an indication that local repeat visitors are less likely to be willing to pay an entrance fee, although this relationship was only found on an α < 0.1 level (β = −1.214, p = 0.079). This result might suggest that loyal return visitors see their admission free entry as an acquired right. Or, alternatively, it might be felt that an admission fee would limit future visitation possibilities and is, therefore, resisted. While the effect of income on WTP is quite well-established (e.g. Horowitz and McConnell, 2003; Lee, 2016), such a relationship was not present in the case of the New Zealand Maritime Museum. Our findings are comparable to a number of other studies that failed to establish socio-economic and motivational determinants to WTP behaviour, with only a few study design variables able to explain the variation in WTP (see e.g. Christie, 2007; Jacobsen and Hanley, 2009; Kim et al., 2007; Murphy et al., 2018; Togridou et al., 2006). Christie (2007) and Foster et al. (1997) provide a possible explanation through their findings that models on real WTP have larger explanatory power than models on hypothetical WTP. This might further explain the lack of model fit in our research, since WTP questions were purely hypothetical and, as can be seen later, the link to actual behaviour was very weak.

Table 4.

Results of binary logistic regressions.

	Locals				Non-locals
	Including income		Excluding income		Including income		Excluding income
	B (standard error)	Exp(B)	B (standard error)	Exp(B)	B (standard error)	Exp(B)	B (standard error)	Exp(B)
Constant	1.483 (2.045)	4.408	2.072 (1.326)	7.941	0.997 (1.380)	2.711	1.374 (1.365)	3.950
Repeat visitors	−1.214* (0.692)	0.297	−0.701 (0.493)	0.496	−0.516 (0.773)	0.597	−0.228 (0.686)	0.796
Yearly household income (NZ$1000)	0.000 (0.000)	1.000			0.000 (0.000)	1.000
Age 16–29 years	−0.219 (1.103)	0.804	−0.244 (0.783)	0.784	0.190 (0.494)	1.209	−0.224 (0.416)	0.799
Age 30–49 years	−0.031 (0.986)	0.969	−0.621 (0.735)	0.537	1.524** (0.665)	4.590	0.737 (0.543)	2.089
Motivation: Situational	0.261 (0.218)	1.298	0.084 (0.138)	1.087	0.234 (0.202)	1.263	0.088 (0.175)	1.092
Motivation: Strong interest	−0.327 (0.268)	0.721	−0.159 (0.203)	0.853	−0.065 (0.159)	0.937	0.143 (0.132)	1.154
Motivation: Recommended	0.176 (0.286)	1.193	0.067 (0.212)	1.069	−0.521*** (0.187)	0.594	−0.280* (0.163)	0.756
Motivation: Tour	0.167 (0.243)	1.182	0.108 (0.187)	1.114	−0.089 (0.211)	0.915	−0.103 (0.205)	0.902
Motivation: Walk-in	0.264 (0.351)	1.302	0.176 (0.247)	1.192	−0.134 (0.200)	0.874	−0.058 (0.186)	0.943
Importance of community support	−0.735 (0.793)	0.479	−0.463 (0.542)	0.629	0.291 (0.457)	1.338	0.601 (0.389)	1.824
Model summary
N	83		132		153		258
Nagelkerke R²	0.184		0.089		0.167		0.080
−2LL	76.250		123.349		153.014		196.842

Note. LL: log-likelihood.

***p < 0.01.

**p < 0.05.

*p < 0.1.

Capturing consumer surplus via voluntary contributions

Given that a majority of respondents experienced a positive consumer surplus, and such consumer surplus is indicative of a suboptimal price setting, a next step in the research tried to identify whether voluntary contributions served to capture at least part of the surplus. Such additional pricing systems might be deemed more favourable than an actual increase in entrance fees since they do not negatively affect demand or political willingness to support. More specifically, currently the New Zealand Maritime Museum receives a levy of NZ$2,025,000 via the Regional Amenities Funding Act of the Auckland Council. Given that the recorded numbers of Auckland visitors for 2016 were 22,069 adults and 13,991 children, it means that for every adult visitor the museum receives NZ$91.76 of support from the Council, a sum that is much higher than could be achieved via entrance fees (as stated in the previously discussed WTP results).

The number of daily voluntary contributions was identified via electronic hardware within the donation box, while the total donation amount was counted at the end of each observation day. Over the total 20 observation days, only 49 museum visitors offered a monetary contribution with the donations totalling just NZ$122.50 or an average donation of NZ$2.50. These numbers were further compared to the questionnaire responses on donation behaviour, in order to identify the possibility of social desirability bias. Even though the question was preloaded and the questionnaire was respondent-completed and anonymous, in 7 of the 20 observation days (35.0%), there was a positive difference between stated donations and actual donations, both in number of donations and in donation amounts. While the social desirability bias did not seemingly affect a large number of respondents, we could infer that, at minimum, 2.4% of respondents had been dishonest about donating to the museum.

Given that the museum received 5523 billable (i.e. excluding education trips) visitors during the observation period, the number of voluntary contributions (49) was less than 1% (0.89%). The hypothesis that voluntary contributions via donation boxes can be an indication of actual WTP and are capable of extracting consumer surplus from visitors is, therefore, rejected. This result does not confirm the findings of Chao et al. (2015) who suggest that pay-what-you-want pricing can be more profitable than fixed pricing and capture part of the deadweight loss. However, as also noted by Gravert (2017), experiments with pay-what-you-want pricing have generated mixed results, and reference prices, social norms and gain or loss framing can significantly affect the outcomes.

In order to investigate reasons for non-contribution more closely, Table 5 presents the results of the associated multiple response question. One hypothesis to explain the lack of voluntary contributions in the New Zealand Maritime Museum is the pricing structure. For non-local visitors, there is essentially no pay-what-you-want system in place, since a standard entrance fee of NZ$20 is levied. Even for visitors with a higher intrinsic WTP, a set entrance fee could give the impression that the cost of visit has been covered and there is no imbalance in reciprocity. As can be seen from Table 5, almost half of non-local visitors (47.3%) no longer felt a need to donate money to the museum due to the established entrance fee. While this is not the case for Aucklanders, who are entitled to free entry and, therefore, could be considered as a pay-what-you-want case, the fact that others are charged upon entrance implicitly links to their own entitlement as Auckland rate payers. This can strengthen the feeling that the zero admission fee is warranted due to existing public support for the museum, as mentioned by 15.6%. Other important stated reasons were a lack of visibility of the donation box (41.3% for locals and 33.9% for non-locals) and a lack of small change to donate (46.8% for locals and 20.1% for non-locals). The former might be resolved by opting for a true pay-what-you-want pricing regime and asking the voluntary contribution at the entrance which could have further beneficial effects since Fielding and Knowles (2015) found a significant increase in charitable donations under verbal cues. However, such verbal cues might be deemed inappropriate in a setting where people have strong opinions about the non-commercial educational function of museums. The latter – a lack of small change – indicates an inherent weakness in the traditional donation box approach, that is, requiring instantaneous cash donations. Specifically given the technological development of contactless payment systems, already well-established in New Zealand, new donation systems need to be devised to account for trends in payment behaviour that increasingly render carrying cash obsolete. Apart from contactless donation boxes and apps for on-the-spot micropayments, developments could include adding online donation requests to the museum’s website, thereby removing the need for instantaneous payment.

Table 5.

Stated reasons for not donating.

Categories	Local (n = 109; %)	Non-local (n = 283; %)
Already paid for my entrance through local tax/entrance fee	15.6	47.3
Did not see a donation box	41.3	33.9
No small change at hand	46.8	20.1
Not interested in donating	5.5	5.7
Already involved in charity	0.0	1.1
No money	2.8	4.2
Other purchases made in gift shop	1.8	1.1

Finally, Table 6 looks at the effect of the four experimental designs on donation box contributions. However, due to the small number of donations, statistical results need to be interpreted with care. In order to exclude potential inference due to differences in subsample characteristics during the different experimental stages, a series of preliminary tests were conducted. The results of χ² tests for local to non-local ratio differences (p = 0.204), repeat visitation (p = 0.701), gender (p = 0.575), age (p = 0.343), the analysis of variance tests for yearly household income (p = 0.061) and the five motivational categories (p’s = 0.688, 0.272, 0.215, 0.735, 0.706) all indicated that the subsample characteristics were comparable.

Table 6.

Effect of donation box designs on voluntary contributions.

Experiment	Donation numbers (% billable visitors)	χ² test (df = 1)		Donation amount
Experiment	Donation numbers (% billable visitors)	χ² value	p Value	Total	Average
Baseline	9 (0.91%)			15.70	1.74
Watching eyes	11 (0.68%)	0.425	0.513	48.85	4.17
Social norms	9 (0.46%)	2.121	0.144	29.80	3.31
Reciprocity	20 (2.04%)	4.262	0.038	31.15	1.56

Note: Values in italics indicate the percentage of donations on total number of visitors and should not be left out. Their statistical significance in turn is calculated via the chi-square test and associated p-value given here.

Because total and average donation amounts were influenced by outlier values, we focussed solely on donation numbers. A χ² test on expected versus observed donations, comparing each experiment with the original baseline, only found a significant effect in the final experiment that was designed to induce feelings of reciprocity (χ² = 4.262, p = 0.038), doubling the number of donations. Neither the watching eyes poster nor reference to social norms had any discernible effect on the number of voluntary contributions. Both the watching eyes and social norms experiments were based on the existence of strong behavioural norms shared by local communities and guiding individual behaviour. While this has been found valid in terms of charitable donations (e.g. Agerström et al., 2016; Fathi et al., 2014) and recycling (Schultz, 1998), the non-significance of our experiments might imply a lack of shared norms for museum donations. This could partially be affected by respondent cultures where a number of nationalities are used to free admission systems and have come to regard it as both a right and a situation whereby funding responsibility is diverted to states and local councils.

The lack of effect of the social norm representation was further exacerbated by the suboptimal message that referred to general opinions on financial support for cultural heritage. While this was born out of a necessity that reflected the lack of voluntary contributions in the first place, it was mentioned by Agerström et al. (2016) that the more specific normative cues had most effect. Therefore, much of the potential influence might have been lost due to a lack of detailed, descriptive and localized reference to donation numbers and average donation amounts. Still, such a reference would require stable and high donation numbers at the outset, in order not to revert to untrue and manipulative statements.

In confirmation of the findings of the UK Behavioural Insights Team (2013), the final experiment (reciprocity) proved to significantly increase donation numbers. When comparing the sample characteristics of this experimental setup with the baseline, we do note slight variations in the number of international visitors (70.7% during the baseline and 77.9% during the reciprocal setup), repeat visitation (81.5–85.7%), gender (46.7–53.1% female) and the age groups 16–19 (4.3–10.6%) and 40–49 (13–5.3%) which could influence the results at small observation numbers. However, it also has to be noted that neither of these variables were found particularly influential in terms of predicting WTP. If we do expect that the results are related to the specific experimental design, then three effects might have been at play here: (a) increasing visibility of the donation box, (b) establishing reciprocity or (c) capturing general curiosity. In order to test whether the increase in donations was solely due to increased visibility, we compared the ‘did not see a donation box’ stated reason for non-donation over the four different experiments. Since no significant difference was found, the hypothesis that the increase in contributions was merely due to increased attention could be rejected. The other effects could only be established through conjecture. Field observations could conclude that the interactivity of the box attracted more visitors. However, it is important to note that the display becoming active was contingent upon movement around the donation box and not on actual donations. It might, therefore, be hypothesized that donations were primarily a measure of additional gratitude and reciprocity, and not of curiosity.

Discussion and conclusion

Cultural and heritage resources are important contributors to the attraction potential of cities and regions. Because of regional and intersectoral spillover effects, the public sector often contributes significantly to the financing and management of such attractions. However, this also increases the vulnerability of cultural resources in times of decreasing government levies and shifts towards privatization. The present article uses a CVM to first establish use value and WTP for a museum, and then empirically tests whether additional voluntary contributions, in the form of donation boxes, can significantly supplement a museum’s income and serve to internalize part of the existing consumer surplus via pay-what-you-want principles. While the use of CVM, despite its critics, is well-established in offering a full scope on economic value, the link between WTP and actual behaviour through entrance fees and additional voluntary contributions offers a novel contribution to the literature. The results serve to increase the understanding of the potential for and viability of moving towards user-pays principles at museums via a mix of compulsory and non-compulsory contribution systems. This is especially significant since museums are not always in a position to raise entrance fees according to market demand due to a risk of losing their public character and associated governmental support.

Results collected from a visitor survey at the New Zealand Maritime Museum in Auckland confirm a WTP over and above current pricing strategies, leading to significant consumer surpluses. These are highest for local visitors who, under current regulation, can enter the museum free of charge. The consumer surplus for these visitors was on average NZ$14.55. Because of a standard admission fee of NZ$20 for non-locals, their average consumer surplus was lower at NZ$9.51. While theoretically these results show an opportunity for price increases, additional entrance fees are strategically unwarranted due to the adverse effect on visitor numbers. Donation boxes could then be seen as an innocuous and voluntary mechanism for collecting at least part of the existing consumer surplus. Particularly in the case of local visitors, the donation box of the New Zealand Maritime Museum serves as a sort of pay-what-you-want scheme, since these visitors pay nothing upon entry. Field observations of collected donations, however, showed that a standard donation box (the baseline experiment) does not manage to extract consumer surplus, with less than 1% of billable visitors offering a donation, which was NZ$1.74 on average. Alterations to the donation box design based on normative theory showed little promise; only by invoking social exchange norms via interactivity in the donation box did we notice a significant increase in donations to just over 2% of billable visitors. It is clear, though, that the percentage of voluntary contributions via the donation box is nowhere near the percentage of visitors who indicated a positive WTP above the current entrance fee. It has to be concluded from the experimental setup that simple donation boxes are ineffective tools for capturing consumer surplus and cannot be seen as viable alternatives to existing revenue streams. To achieve significant increase in such voluntary payments, it might be necessary to move away from purely impersonal demand systems. At the same time, we cannot ignore the possibility that the hypothetical WTP is overstated and might reflect expressive behaviour to show general support of cultural institutions, rather than instrumental behaviour (Hamlin and Jennings, 2011).

The results of the study indicate limitations to pricing flexibility in cultural attractions with modest visitor numbers, such as the New Zealand Maritime Museum. Care has to be taken in investigating price elasticity of demand and the ultimate effect of price on profitability. Specifically in such cases as the one investigated, where the current public funding – if calculated on a per head basis – is well above the WTP, entrance fees might only serve to lower public funding accordingly. While voluntary additional contributions, therefore, seem like an appropriate strategy for extracting additional WTP, even state-of-the-art donation boxes involving proactive, animated displays have limited effect on the revenue stream. It seems to suggest that adding pay-what-you-want principles to an already existing traditional entrance fee does not instigate additional payments but rather transmits the message that a necessary fee has already been paid. Therefore, a fully committed pay-what-you-want scheme might result in different findings. However, as previous literature has suggested, findings on the benefits of such pricing strategies are not unilaterally positive (Gravert, 2017).

Main study limitations arise from the small sample that was left after removing visitors who had not donated to the museum. The low propensity to donate yielded only a small sample size available to test the three additional experimental designs. Future research should take into account the donation patterns and extend the study sample size and study duration accordingly. Additional explanatory variables for WTP values should also be taken into account, since very limited causal variables were found in this study. Finally, our findings on the existence of a social acceptable bias lead to a general concern in research of this nature: how trustworthy are stated responses? Developing mechanisms to compare stated behaviour with actual behaviour where possible should be prioritized.

Footnotes

Author’s note

Bart Neuts is now affiliated with KU Leuven, Belgium.

Acknowledgements

The author would like to thank Eden Potter, Ivana Garcia, Chen Chen and Darlene Koo for designing the ‘watching eyes’ poster, Sam Gray for his help in designing the donation box and animated setup, Gurudatta Vishwasrao for survey assistance and the people of the New Zealand Maritime Museum for their friendly cooperation.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Auckland University of Technology, Faculty of Culture and Society Contestable Research under grant number AX16/12.

ORCID iD

Bart Neuts

Notes

References

Agerström

Carlsson

Nicklasson

, et al. (2016) Using descriptive social norms to increase charitable giving: the power of local norms. Journal of Economic Psychology 52: 147–153.

Andreoni

(1990) Impure altruism and donations to public goods: a theory of warm-glow giving. The Economic Journal 100(401): 464–477.

Apostolakis

Jaffry

(2013) An analysis of monetary voluntary contributions for cultural resources: the case of the British Museum. Tourism Economics 19(3): 631–651.

Arbrecht

(2014) Use value of cultural experiences: a comparison of contingent valuation and travel cost. Tourism Management 42: 141–148.

Ashworth

Johnson

(1996) Sources of “value for money” for museum visitors: some survey evidence. Journal of Cultural Economics 20: 67–83.

Azar

(2007) The social norm of tipping: a review. Journal of Applied Social Psychology 37(2): 380–402.

Balan

(2014) “Pay what you want”: a participative price setting mechanism. International Journal of Economic Practices and Theories 4(5): 952–963.

Bateson

Callow

Holmes

, et al. (2013) Do images of ‘watching eyes’ induce behaviour that is more pro-social or more normative? A field experiment on littering. PLOS One 8(12): 1–9.

Becker

(1974) A theory of social interactions. Journal of Political Economy 82(6): 1063–1093.

10.

Bedate

Herrero

Sanz

(2009) Economic valuation of a contemporary art museum: correction of hypothetical bias using a certainty question. Journal of Cultural Economics 33: 185–199.

11.

Bertacchini

Santagata

Signorello

(2010) Loving cultural heritage – private individual giving and prosocial behavior. FEEM, Working Paper No.11.2010. Available at: https://papers.ssrn.com/sol3/papers.cfm?%20abstract_id=1557776 (accessed 20 December 2017).

12.

Bhattacharya

Hayagreeva

Glynn

(1995) Understanding the bond of identification: an investigation of its correlates among art museum members. Journal of Marketing 59(4): 46–57.

13.

Brida

Dalle Nogare

Scuderi

(2016) Frequency of museum attendance: motivation matters. Journal of Cultural Economics 40(3): 261–283.

14.

Carlsson

Frykblom

Lagerkvist

(2005) Using cheap talk as a test of validity in choice experiments. Economics Letters 89(2): 147–152.

15.

Carson

Groves

(2007) Incentive and informational properties of preference questions. Environmental and Resource Economics 37: 181–210.

16.

Champ

Bishop

Brown

, et al. (1997) Using donation mechanisms to value nonuse benefits from public goods. Journal of Environmental Economics and Management 33: 151–162.

17.

Chao

Fernandew

Nahata

(2015) Pay-what-you-want pricing: Can it be profitable? Journal of Behavioral and Experimental Economics 57: 176–185.

18.

Christie

(2007) An examination of the disparity between hypothetical and actual willingness to pay using the contingent valuation method: the case of red kite conservation in the United Kingdom. Canadian Journal of Agricultural Economics 55: 159–169.

19.

Deb

Gazzale

Kotchen

(2014) Testing motives for charitable giving: a revealed-preference methodology with experimental evidence. Journal of Public Economics 120: 181–192.

20.

Diamond

Hausman

(1994) Contingent valuation: Is some number better than no number? Journal of Economic Perspectives 8(4): 45–64.

21.

Ekström

(2012) Do watching eyes affect charitable giving? Evidence from a field experiment. Experimental Economics 15: 530–546.

22.

Fathi

Bateson

Nettle

(2014) Effects of watching eyes and norm cues on charitable giving in a surreptitious behavioral experiment. Evolutionary Psychology 12(5): 878–887.

23.

Fielding

Knowles

(2015) Can you spare some change for charity? Experimental evidence on verbal cues and loose change effects in a Dictator game. Experimental Economics 18: 718–730.

24.

Foster

Bateman

Harley

(1997) Real and hypothetical willingness to pay for environmental preservation: a non-experimental comparison. Journal of Agricultural Economics 48(2): 123–138.

25.

Grammatikopoulou

Olsen

(2013) Accounting protesting and warm glow bidding in contingent valuation surveys considering the management of environmental goods – an empirical case study assessing the value of protecting a Natura 2000 wetland area in Greece. Journal of Environmental Management 130(30): 232–241.

26.

Gravert

(2017) Pride and patronage – pay-what-you-want pricing at a charitable bookstore. Journal of Behavioral and Experimental Economics 67: 1–7.

27.

Hackl

Pruckner

(1999) On the gap between payment card and closed-ended CVM-answers. Applied Economics 31: 733–742.

28.

Hamlin

Jennings

(2011) Expressive political behaviour: foundations, scope and implications. British Journal of Political Science 41(3): 645–670.

29.

Horowitz

McConnell

(2003) Willingness to accept, willingness to pay and the income effect. Journal of Economic Behavior and Organization 51(4): 537–545.

30.

Huntington

(1993) Ticket pricing policy and box office revenue. Journal of Cultural Economics 17(1): 71–87.

31.

Institute of Museum and Library Services (2008) Exhibiting public value: Government funding for museums in the United States. Report, IMLS Office of Policy, Planning, Research, and Communications, Washington DC, December.

32.

Jacobsen

Hanley

(2009) Are there income effects on global willingness to pay for biodiversity conservation? Environmental Resource Economics 43: 137–160.

33.

Kahneman

Knetsch

(1992) Valuing public goods: the purchase of moral satisfaction. Journal of Environmental Economics and Management 22(1): 57–70.

34.

Kallgren

Reno

Cialdini

(2000) A focus theory of normative conduct: when norms do and do not affect behavior. Journal of Personality and Social Psychology 26: 1002–1012.

35.

Kim

Kaufmann

Stegemann

(2014) The impact of buyer-seller relationships and reference prices on the effectiveness of the pay what you want pricing mechanism. Marketing Letters 25(4): 409–423.

36.

Kim

Wong

KKF

Cho

(2007) Assessing the economic value of a world heritage site and willingness-to-pay determinants: a case of Changdeok Palace. Tourism Management 28: 317–322.

37.

Klein

WMP

Goethals

(2002) Social reality and self-construction: a case of “bounded irrationality”? Basic and Applied Social Psychology 24(2): 105–114.

38.

Lampi

Orth

(2009) Who visits the museums? A comparison between stated preferences and observed effects of entrance fees. Kyklos 62(1): 85–102.

39.

Lee

(2016) Income and distance-decay effects on willingness to pay estimated by the contingent valuation method. Journal of Environmental Planning and Management 59(11): 1957–1981.

40.

Loomis

(2014) Strategies for overcoming hypothetical bias in stated preference surveys. Journal of Agricultural and Resource Economics 39(1): 34–46.

41.

Martin

(1994) Determining the size of museum subsidies. Journal of Cultural Economics 18: 255–270.

42.

Montemagno

(2002) Heritage and tourism. In: Rizzo

Towse

(eds.), The Economics of Heritage: A Study in the Political Economy of Culture in Sicily. Cheltenham: Edward Elgar, pp. 79–87.

43.

Mourato

Ozdemiroglu

Hett

, et al. (2004) Pricing cultural heritage: a new approach to managing ancient resources. World Economics 5(3): 95–113.

44.

Murphy

Allen

Stevens

, et al. (2005) A meta-analysis of hypothetical bias in stated preference valuation. Environmental and Resource Economics 30: 313–325.

45.

Murphy

Campbell

Drew

(2018) Examination of tourists’ willingness to pay under different conservation scenarios: evidence from reef manta ray snorkelling in Fiji. PLoS ONE 13(8): 1–15.

46.

Natter

Kaufmann

(2015) Voluntary market payments: underlying motives, success drivers and success potentials. Journal of Behavioral and Experimental Economics 57: 149–157.

47.

Noonan

(2002) Contingent valuation studies in the arts and culture: An annotated bibliography. The Cultural Policy Centre at the University of Chicago Working Paper. Available at: https://culturalpolicy.uchicago.edu/sites/culturalpolicy.uchicago.edu/files/Noonan11.pdf (accessed 8 August 2018).

48.

Noonan

(2003) Contingent valuation and cultural resources: a meta-analytic review of the literature. Journal of Cultural Economics 27: 159–176.

49.

O’Hagan

(1995) National museums: to charge or not to charge. Journal of Cultural Economics 19: 33–47.

50.

O’Hagan

(1998) Art museums: Collections, deaccessioning and donations. Journal of Cultural Economics 22(2–3): 197–207.

51.

Perfecto

Jung

Nelson

, et al. (2013) Default effects under pay-what-you-want: evidence from the field. Advances in Consumer Research 41: 265–266.

52.

Perrine

Heather

(2000) Effects of picture and even-a-penny-will-help appeals on anonymous donations to charity. Psychological Reports 86: 551–559.

53.

Ponzini

(2010) The process of privatisation of cultural heritage and the arts in Italy: analysis and perspectives. International Journal of Heritage Studies 6: 508–521.

54.

Porter

(2011) The Price of Everything: Solving the Mystery of Why We Pay What We Do. New York: Portfolio/Penguin.

55.

Porter

Kramer

(2002) The competitive advantage of corporate philanthropy. Harvard Business Review 2002: 5–16.

56.

Queensland Government (2009) Valuing the Queensland Museum. A contingent valuation study 2008. Report, Queensland Museum, Australia, April.

57.

Roberts

(1998) Competitive altruism: from reciprocity to the handicap principle. Proceedings of the Royal Society B 265: 427–433.

58.

Saayman

Krugell

Saayman

(2016) Willingness to pay: Who are the cheap talkers? Annals of Tourism Research 56: 96–111.

59.

Santagata

Signorello

(2000) Contingent valuation and cultural policy: the case of “Napoli Musei Aperti”. Journal of Cultural Economics 24: 181–204.

60.

Sanz

JÁ

Herrero

Bedate

(2003) Contingent valuation and semiparametric methods: a case study of the national museum of sculpture in Valladolid, Spain. Journal of Cultural Economics 27: 241–257.

61.

Savas

(2016) Factors affecting donations in U.S. retail stores: a conceptual framework. Journal of Retailing and Consumer Services 33: 178–185.

62.

Schultz

(1998) Changing behavior with normative feedback interventions: a field experiment on curbside recycling. Basic and Applied Social Psychology 21(1): 25–36.

63.

Seaman

(2006) The relationship among regional economic impact models: contingent valuation versus economic impact in the case of cultural assets. Andrew Young School of Policy Studies Research Paper Series, Working Paper 07-05. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=975773 (accessed 8 August 2018).

64.

Snowball

(2013) The economic, social and cultural impact of cultural heritage: methods and examples. In: Rizzo

Mignosa

(eds.), Handbook on the Economics of Cultural Heritage. Cheltenham: Edward Elgar, pp. 438–455.

65.

StatsNZ (2017) Household income and housing-cost statistics: year ended June 2017. Available at: http://archive.stats.govt.nz/browse_for_stats/people_and_communities/Households/HouseholdEconomicSurvey_HOTPYeJun17.aspx (accessed 20 December 2017).

66.

The Economist (2002) The state and the arts: patrimony and privatisation. Available at: http://www.economist.com/node/1465796 (accessed 21 December 2017).

67.

Thompson

Berger

Blomquist

, et al. (2002) Valuing the arts: a contingent valuation approach. Journal of Cultural Economics 26: 87–113.

68.

Tian

Holst

(2011) Applying the payment card approach to estimate the WTP for green food in China. IAMO Forum 2011, No. 23. Available at: http://hdl.handle.net/10419/50786 (accessed 20 December 2017).

69.

Togridou

Hovardas

Pantis

(2006) Determinants of visitors’ willingness to pay for the national marine park of Zakynthos, Greece. Ecological Economics 60: 308–319.

70.

Tohmo

(2004) Economic value of a local museum: factors of willingness-to-pay. Journal of Socio-Economics 33: 229–240.

71.

UK Behavioural Insights Team (2013) Applying behavioural insights to charitable giving. Report, Cabinet Office, UK, May.

72.

Venkatachalam

(2004) The contingent valuation method: a review. Environmental Impact Assessment Review 24(1): 89–124.

73.

Voyager New Zealand Maritime Museum (2017) Annual report 2015/2016 financial year. Available at: www.maritimemuseum.co.nz/sites/default/files/2015%20-%202016%20Annual%20Report%20FINAL.PDF (accessed 15 December 2017).

74.

Willis

(1994) Paying for heritage: what price for Durham Cathedral. Journal of Environmental Planning and Management 3: 267–278.

75.

Zawojska

Czajkowski

(2017) Re-examining empirical evidence on stated preferences: importance of incentive compatibility. Journal of Environmental Economics and Policy 6(4): 374–403.