1. www.seo.nl - secretariaat@seo.nl - +31 20 525 1630
The effects of the Dutch museum pass on museum visits and museum finances
Presentation for the 18th International Conference on Cultural Economics
Montreal, June 25, 2014
Siemen van der Werff

2. Conclusions
Dutch museum pass holders visit a museum 3 times as much because of the museum pass (5 million visits instead of 1.6 million visits in 2012)
This estimate is based on a count model that takes endogeneity into account
Museums earned € 14 million in extra revenue because of the museum pass
In addition, museums generate extra income in museum shops and cafes
This calculation assumes marginal costs of extra visitors are 0, which can only be true when museums operate under their maximum capacity

3. What is the Dutch museum pass
Museum pass gives access to 400 Dutch museums for a fixed fee
Fee in 2012 was €40 for adults, €20 for under 18
0.9 million pass holders in 2012 (out of 16 million inhabitants)
They made 5.0 million museum visits with their pass (out of 20 million total visits to participating museums)
Pass is owned and administered by participating museums united in a separate foundation
Revenues are distributed under participating museums based on number of visits and regular entrance fees of museums
Revenues per visit from pass holders are about 65% of the revenues per visit from regular visitors

4. Distribution of museum visits by pass holders

5. Research questions Does the museum pass stimulate museum visits?
To which amount?
What is the financial effect of the museum pass for museums?
More visitors versus lower revenues per visitor: what is the balance sheet?
Research originally commissioned by Dutch museum pass

6. Literature (1)
Most literature considering museum entrance fees tackle the question whether it should be free or not
Furthermore, demand for museums is inelastic
Estimate of elasticity for Dutch museums is (Goudriaan et al)
It has been estimated that Dutch museum visits would increase with 30% if museums would have free entrance (Goudriaan et al)
Entrance fees make up around 25% of total costs of visiting a museum (Goudriaan et al)

7. Literature (2)
Frey and Meier model museum finances, they assume that museums are utility maximizing with respect to the number of visitors and the quality of exhibitions, within a budget restriction
Revenue is based on the number of visitors is gained from entrance fees and ancillary goods
Museums have high fixed costs and marginal costs close to 0
But marginal costs are close to zero only in the short run when a museum is not operating at full capacity (Bailey & Falconer)

8. Contribution to field and literature
We look at empirical effects of a national museum pass
Applying econometric methods to revealed preference data
Museum subsidies throughout Europe are dropping
So museums need more other revenues, museum passes might be an option
And museums might want more visitors, which could be reached by a museum pass

9. Data Survey under 2,070 museum visitors (May 2013)
A museum visitor is a Dutch citizen who has visited a Dutch museum at least 1 time in the last 3 years
Includes museum pass holders, non-holders and holders of other type of discount passes
Administrative data of Dutch museum pass
Museum visits of museum pass holders
Characteristics of museum pass holders
Entrance fees of museums
Benefits of museum pass fees distributed to museums

10. Museum visits pass holders and non-pass holders
Museum visits in 2012 (if <= 20)
Number of observations
Pass holders
5,2
291
Holder of other type of discount pass
2,1
344
Other museum visitors
1,3
1419
Total
2,0
2054
Museum pass holders visit a museum 4 times as much as other museum visitors
This does not only shows the effect of the museum pass…
… but also differences in characteristics of card holders and other visitors
So endogeneity is a problem, therefore an econometric analyses is needed

11. Econometric model (1)
Outcome variable is of count nature (number of visits per visitor)
Therefore a count model is needed, in this case the negative binomial model
The treatment is endogenous, and there are multiple treatments
So a multinominal endogenous treatment model is needed
This model
Respondent owns museum pass
Respondent owns other type of discount pass (ie. 65+, youth, students, low income)
Respondent does not own a discount pass

12. Econometric model (2)
Background variables used for estimation of number of museum visits are gender, age class, type of household, level of own education, level of parents education, region
Extra background variables used for estimation of probability of owning a museum pass are household income and an indicator for students
These variables do not have a direct effect on museum visits (empirically tested), but have an effect (both theoretical and in practice) on probability of owning a museum pass or another type of discount pass (low income respondents and students)

13. Estimation results (1) Effect on museum visits
Effect on probability of owning museum pass
Effect on probability of owning other type of discount pass
Females (ref males)
0.096**
0.097
-0.036
Age class: (ref 15-19)
-0.056
-1.293**
-0.954***
Age class: 25-29
-0.475***
-1.847**
-1.768***
Age class: 30-34
-0.429***
-0.630
-1.687***
Age class: 35-39
-0.198
-0.895
-1.504***
Age class: 40-44
-0.257**
-1.021
-1.615***
Age class: 45-49
-0.161
-0.847
-1.513***
Age class: 50-54
-0.226*
-1.110*
-1.714***
Age class: 55-59
-0.169
-0.693
-2.435***
Age class: 60-64
-0.058
0.536
-1.412***
Age class: 65-69
-0.084
0.618
-0.958**
Age class: 70-89
0.152
1.051
-1.123**
Type of Household: with partner without children (ref: single)
-0.021
-0.695***
0.105
Type of Household: with partner and children
-0.049
-0.654**
0.233
Type of Household: single with children
-0.118
-0.441
0.189
Type of Household: other
0.116
-0.207
-0.083
Household income: between € and € (ref: <€12.500)
1.045*
0.137
Household income: between € and €26.200
0.869
-0.193
Household income: between € and €32.500
1.146**
-0.269
Household income: between € and €38.800
1.172**
-0.671*
Household income: between € and €51.300
0.808
-0.749*
Household income: between € and €65.000
0.983*
-0.473
Household income: above €65.000
1.718***
-0.419
Household income: won’t say
1.180**
-0.378

14. Estimation results (2) Effect on museum visits
Effect on probability of owning museum pass
Effect on probability of owning other type of discount pass
Highest level of education: higher professional (ref: university)
0.030
0.011
0.222
Highest level of education: senior general secondary / pre-university
-0.086
-0.825**
0.417
Highest level of education: secondary vocational
-0.233***
-0.993***
0.026
Highest level of education: pre-vocational
-0.173*
-0.811**
0.552
Highest level of education: lower pre-vocational
-0.309**
-0.655
0.198
Highest level of education: won’t say
0.095
0.839
0.487
Highest level of education parents: higher professional (ref: university)
0.014
-0.896***
-0.166
Highest level of education parents: senior general secondary / pre-university
-0.040
-0.847**
-0.601*
Highest level of education parents: secondary vocational
-0.075
-1.013***
-0.303
Highest level of education parents: pre-vocational
-0.055
-0.728**
-0.445
Highest level of education parents: lower pre-vocational
-0.127
-0.943***
-0.788**
Highest level of education parents: won’t say
-0.148
-1.504***
-0.667**
Dummies for regions (40)
included
Dummy for students
-0.368
0.661**
Dummy for museum pass owner
1.101***
Dummy for owner other type of discount pass
-0.051
Constant
0.431*
-0.922
-0.093

15. Estimation results Effects on probability of owning a museum pass:
Under 30 (-), singles (+), high income (+), highly educated (+), highly educated parents (+)
Effects on probability of owning other type of discoutnt pass:
Students (+), under 20 (+), above 65 (+), low income (+), highly educated parents (+)
Effects on museum visits:
Females (+), between (-), highly educated (+), museum pass owners (+)

16. Effect museum pass on number of visitors
So museum pass holders visit a museum 3.0 (=exp(1.10)) times as much because of the museum card controlling for a large number of visitor characteristics
95%-confidence interval is 2.1 to 4.0
In million visitors
Regular
Museum pass
Other
Total
Total number of visits to participating museums
7.1
5.0
7.3
19.5
Estimated number of visits to participating museums if museum pass would not exist
8.8
0.0
16.2
Effect museum pass
-1.7
+5.0
+3.4

17. Financial effects museum pass on entrance revenues
Effect museum pass 2012 in million €
Lost regular entrance fees
-15.4
Contribution for visitors with museum pass
28.7
Commission to museums for selling museum pass
0.7
Total
14.0
The total effect is €6.6 million if lower bound of effect of museum cards on visits is used

18. Financial effect on revenue from ancillary goods
Effect on revenue from ancillary goods depends on profit margin on these goods
Based on survey under museum, we estimate a profit margin of 118% on goods and 102% on food & drinks
Here, personnel costs are not taken into account
When personal costs are taken into account, the literature suggest a profit margin of around 40%
Estimated additional revenue based on high profit margins is €12.7 million
Based on low profit margins, this estimate is €5.5 million

19. Females (ref males)
0.096**
0.097
-0.036
Age class: (ref 15-19)
-0.056
-1.293**
-0.954***
Age class: 25-29
-0.475***
-1.847**
-1.768***
Age class: 30-34
-0.429***
-0.630
-1.687***
Age class: 35-39
-0.198
-0.895
-1.504***
Age class: 40-44
-0.257**
-1.021
-1.615***
Age class: 45-49
-0.161
-0.847
-1.513***
Age class: 50-54
-0.226*
-1.110*
-1.714***
Age class: 55-59
-0.169
-0.693
-2.435***
Age class: 60-64
-0.058
0.536
-1.412***
Age class: 65-69
-0.084
0.618
-0.958**
Age class: 70-89
0.152
1.051
-1.123**
Type of Household: with partner without children (ref: single)
-0.021
-0.695***
0.105
Type of Household: with partner and children
-0.049
-0.654**
0.233
Type of Household: single with children
-0.118
-0.441
0.189
Type of Household: other
0.116
-0.207
-0.083
Household income: between € and € (ref: <€12.500)
1.045*
0.137
Household income: between € and €26.200
0.869
-0.193
Household income: between € and €32.500
1.146**
-0.269
Household income: between € and €38.800
1.172**
-0.671*
Household income: between € and €51.300
0.808
-0.749*
Household income: between € and €65.000
0.983*
-0.473
Household income: above €65.000
1.718***
-0.419
Household income: won’t say
1.180**
-0.378
Highest level of education: higher professional (ref: university)
0.03
0.011
0.222
Highest level of education: senior general secondary / pre-university
-0.086
-0.825**
0.417
Highest level of education: secondary vocational
-0.233***
-0.993***
0.026
Highest level of education: pre-vocational
-0.173*
-0.811**
0.552
Highest level of education: lower pre-vocational
-0.309**
-0.655
0.198
Highest level of education: won’t say
0.095
0.839
0.487
Highest level of education parents: higher professional (ref: university)
0.014
-0.896***
-0.166
Highest level of education parents: senior general secondary / pre-university
-0.040
-0.847**
-0.601*
Highest level of education parents: secondary vocational
-0.075
-1.013***
-0.303
Highest level of education parents: pre-vocational
-0.055
-0.728**
-0.445
Highest level of education parents: lower pre-vocational
-0.127
-0.943***
-0.788**
Highest level of education parents: won’t say
-0.148
-0.667**
Dummies for regions (40)
included
Dummy for students
-0.368
0.661**
Dummy for museum pass owner
1.101***
Dummy for owner other type of discount pass
-0.051
Constant
0.431*
-0.922
-0.093

20. Financial effects related to museum size
39% of financial revenue of museum pass goes to 10 museums, 61% goes to the other 390 museums
The 100 smallest museums get 0.35% of the total revenue
The average contribution to a small museum is €1000, this can still be substantional for them

21. Conclusion and discussion
Museum pass yields large benefits to museums, both financial as in the total number of visits
Financial effect both from museum pass revenue as from additional sales of ancillary goods
Main limitation is that we cannot be sure that all relevant control variables are included
If possible remaining differences are both positively correlated with museum visits as with museum pass ownership, then the analysis might overestimate the true effect of the museum pass