1. More or Better- The Effect of Quality on Income Elasticity in Tourism Consumption Aliza Fleischer Department of Agricultural Economics Hotel, Food Resources and Tourism Management Program The Hebrew University of Jerusalem

2. Trends in Tourism Consumption The quality of tourism goods is on the rise. Boutique hotels, unique spas, luxurious B&B, and spiritual tours with modern amenities are some examples of the emerging new high quality products This means that tourists, in recent years, decided to increase the quality of their vacations

3. Objectives In this paper I estimate the income elasticity with quality choice The estimations take into consideration the fact that some of the households have zero expenditure I show that income elasticity of demand is smaller when quality decision is considered

4. Definition of Income Elasticity Income elasticity= % change in quantity demanded % change in income If prices are constant: % change in expenditures % change in income

5. Important points in income elasticity study Income elasticity shows the sensitivity of tourism consumption to changes in income In using cross-section data such as Households Expenditure Survey one assumes constant prices As a result the demand function collapses into the Engel curve; price elasticities cannot be obtained

6. Estimations of Income Elasticities in the Literature Davies and Mangan (1992) used the UK family expenditure survey. The income elasticity in mid point is 2.1. Poor households have elasticity of 4 and wealthy households elasticity of 1.5. Van Soest and Koorman (1987) and Melenberg and Van Soest (1996) studied the factors determining vacation expenditures in Dutch households. They also used cross section data but unlike Davies and Mangan (1992) they took into consideration the fact that only a fraction of the households have non-zero expenditures. In the first paper it was found that vacation abroad is a luxury good with income elasticity of 2.1 while domestic vacations are a basic good with income elasticity of only 0.7. In the latter using parametric and semi- parametric modeling income elasticity was found to be 1.7. In works analyzing leisure and recreation expenditures based on cross section data (Costa, 1997 and 1999; Weagley, 2004) similar results were found. Income elasticity was found to be significantly larger than one.

7. Problems with Existing Estimates Assuming Constant Prices Across Households 1.Income elasticity larger than one implies that if a household enjoys an increase in income it will increase its expenditures even more on vacation. However, since prices are constant it means that the increase in expenditures reflects an increase in the number of vacation days only. It does not reflect changes in the quality of the vacations. 2.Polinsky (1977) shows that the econometric estimates of the expenditure function using cross section data are biased if prices do vary across households.

8. Unit Value The problem of obtaining prices from household expenditures surveys was overcome in recent works on demand for food. Additional data on quantity consumed enabled researchers to obtain prices (unit values). This is done by dividing expenditure by the number of units. Using it enables the estimation of price elasticity and an unbiased estimation of income elasticity. The unit value differs from the price of homogenous good if we use aggregate commodity such as vacation.

9. Unit Value Expenditure on vacations include hotel nights in Israel and abroad, travel abroad and related recreation activities. Thus, vacation is a heterogeneous commodity whose unit price reflects differences in quality This means that unlike consumer theory where prices of homogenous goods are exogenous the unit value as a price is endogenous. A higher unit value per day of vacation reflects a decision of the household to stay, for example, in a luxury boutique hotel rather than a two-star hotel. It is a function of income and not independent of income like prices of homogenous goods.

10. Unit Value A vacation day during the high season is a better quality product than off season because the weather is better or it is more convenient in terms of vacation policy from school or work. In tourism, unit value can vary between households because of differences in the quality of information-seeking skill. Unit values vary across households

11. Theoretical Model Consumer maximizing its utility subject to budget constraint Demand is a function of price and income Price of vacation as a function quality units and price of unit of quality

12. Theoretical Model Vacations as aggregate commodity Rewriting of the maximization problem Demand of aggregate commodity

13. Theoretical Model Expenditures Unit value

14. Empirical Model 1 without unit value with Selectivity Decision equation Expenditure share equation If I >0 then the following model holds:

15. Empirical Model 2 Simultaneous Equations Model with Selectivity Decision equation Unit value equation Expenditure share equation If I >0 then the following model holds:

16. Income Elasticity

17. Data Household Expenditure Survey 1999 Additional data provided by the CBS on number of vacation days in Israel and abroad enabled calculation of unit value Total of 5987 out of which 699 reported non zero expenditures

18. Income Elasticity Family size elasticity is -1.44 in both models Price elasticity is -0.33

19. Conclusions Increase in income leads to increase in vacation expenditures This increase is affected more from increase in the quality of the vacation than from the number of vacation days This explain the rapid growth we see in quality of tourism products