ABSTRACT
Many localities are able to survive without competitive export industries. The economic base of regional development rests upon various wage and non-wage external incomes. However, suitable regional data are often unavailable, and particularly those related to external consumer spending in urban regions and short-stay tourism. Three geo-localized databases are here combined into a circular-flow template from former consumer spending to the final wage income. By doing so, the paper contributes to identifying ‘transient custom’ as a new inflow driving local development and it provides a first estimate of its magnitude, using the Paris Region as a case study.
ACKNOWLEDGEMENTS
I express many thanks to Richard Shearmur (McGill University) and Andree Woodcock (Coventry University) for their valuable advice.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author.
SUPPLEMENTAL DATA
Supplemental data for this article can be accessed at http://dx.doi.org/10.1080/00343404.2017.1364842.
Notes
1. The definition of the World Tourism Organization constitutes an internationally recognized standard for statistics and tourism satellite accounts.
2. In France, départments (departments) are local authorities legally competent to act for social solidarity, territorial cohesion or even spatial planning. Yet, many of them support economic development and encourage the economic policies of small communities.
3. The year 2006 is the closest year of the BDF survey available with a Paris Region representative sample suitable for this research paper.
4. Using a generalized least square estimation is also an alternative to consider.
5. The log transformation enhances the explanatory power of the model and its capability to provide predicted values as close as possible to observed ones (see Appendix A in the supplemental data online). Indeed, the distributions of both total annual income (explanatory variable) and total annual expenditure (endogenous variable) are highly positively skewed and log transformation is a common fix to reduce skewness (normalize the distribution and partly corrects heteroskedastic biases). Both variables are strictly positive and the transformation also ensures strictly positive predicted values. Lastly, the log transformation made the sensibility of total annual expenditure to the change in total annual income relative rather than absolute; what is more accurate.
6. The following are a few examples of excluded candidate explanatory variables because of a high risk of error: every class with a living area less than 150 m²; every class where the age of the head of the household is less than 65 years; living in an inner suburb area; the head of the household is a manager or executive/a merchant or artisan/being a graduate of a high school/university.
7. In the case of such large-scale surveys that deal with topics that are as sensitive as income and expenditure, recurring biases may be caused by the underestimation of actual amounts or the non-reporting of certain expenditure.
8. When applied to the 2001 BDF survey, the model has similar properties, with an adjusted R² = 0.53 and a similar distribution of observations around the regression line.
9. Such effects are commonly discussed by geographers through the modifiable areal unit problem (MAUP).
10. According to the accounts of the European Union, France received €13.1 billion in 2010 as a result of European policies.
11. The amount of tourist revenue was obtained from the annual statements made by the regional tourism committee.
12. According to the estimates, once individuals’ expenditure in their residential municipality has been removed.