Term Paper, 2006, 54 Pages
Research Paper, 163 Pages
Master's Thesis, 79 Pages
Master's Thesis, 141 Pages
Diploma Thesis, 119 Pages
Term Paper (Advanced seminar), 33 Pages
Research paper, 27 Pages
Particularities of the New Zealand tourism market
Definition of data sets
The calendar effect
The capacity effect
Definition of seasons
Table of Figures
Chart 1: Timeline Room Nights all Season
Chart 2: Adjusted room nights, seasonal comparison by month
Chart 3: Timeline of room nights / day
Chart 4: Room nights / day – development of Average, Max, Min and Max-Min
Chart 5: Shares of total - Guest nights
Chart 6: Shares of total - arrivals
Chart 7: Shares of total - room nights
Chart 8: Comparison of Gini-Coefficients - Guest Nights
Chart 9: Comparison of Gini-Coefficients - Guest Arrivals
Chart 10: Comparison of Gini-Coefficients - Room Nights
Seasonality in tourism:
A review of seasonality of hotel accommodation in New Zealand
Student of Leisure and Tourism Management,
University of Applied Sciences Stralsund, Germany
This paper outlines the main facts about the topic “seasonality” in general with special consideration given to the hotel industry of New Zealand. The aim of this article is to recognise and understand the patterns of seasonality in this industry. For a better understanding the term “seasonality” and the different ways of measuring it are briefly described. Particular attention is paid to the definition of the relevant data sets and the capacity effect. The findings are based on reliable statistical data taken from a specialised institution in New Zealand.
Key words: seasonality, capacity, calendar effect, capacity effect
Seasonality is an important factor in the tourism industry. Economically, ecologically and socially it can have many negative influences on a destination, but at the same time it may provide a time for both the environment and the host communities to recover from the stress of the peak season(s). This paper aims to show the pattern of demand regarding the hotel industry of New Zealand. Data of visitor arrivals, guest nights and room nights is analysed in order to clearly present the patterns and strength of seasonality of this industry. In this analysis, various kinds of measurements are being used and the findings are illustrated by graphs.
This paper first explains the term “seasonality” and commonly used methods to measure seasonality are briefly described. This shall give the reader a basic understanding of the topic. The paper then progresses to describe the particular characteristics of the destination New Zealand with regards to seasonality and to describe the statistical data available. The specific methods used to measure seasonality of the hotel industry sector of New Zealand are discussed, the findings are presented and conclusions are drawn.
Seasonality is, according to Butler (1994) “a temporal imbalance in the phenomenon of tourism”. It can be expressed in the number of visitors, the expenditure of the visitors, the traffic on highways and other forms of transportation, the employment or in the number of admissions to attractions. Causes of seasonality may be natural or institutional. Natural causes can be variations in the weather like temperature, rainfall, snowfall, daylight and sunlight, but also the characteristics of the location such as latitude, altitude and distance from the equator or accessibility. Institutional causes on the other side are man-made, as for example actions and policies regarding culture, religion or social life as well as public holidays or specific events. The most important institutionalised factors are school holidays and industrial holidays. Apart from that, also social pressure or fashion, inertia and tradition or even sporting seasons may cause seasonality. Some of the causes are stable over a long period like Christmas holidays, some change at certain intervals such as school holidays, some vary predictably like the timing of Easter and others are unpredictable as for example the weather (Hylleberg, 1992). According to Butler and Mao (1996) there are four different patterns in which seasonality can occur:
- Single peak seasonality, which is the most common and characterised by a single, clearly identifiable and relatively fixed, time-span of peak demand,
- Two peak seasonality, characterised by two clearly identifiable and fixed time-spans of peak demand
- Non peak seasonality, where no time-span of peak demand can be identified
- Dynamic seasonality, characterised by a single or multiple time-spans of demands, which are, however, not fixed.
As for the tourism industry Hartmann (1986) stated that the “reliable and predictable recurrence of tourists has formed the economic base for the development of the tourist industry”. Therefore, he argued, tourism is naturally seasonal. Additionally, the WTO (1984) argued that the more specialised a destination is, the more seasonal it is and that large urban centres experience less seasonality due to a diversified demand.
Seasonality can be measured in different ways. Firstly, it is necessary to decide for a category of figures. Usually, it is measured in number of visitors. Other units could be the number of arrivals or departures, the number of overnight stays, the length of stay or the expenditures of the visitors. This then needs to be measured on a daily, weekly, monthly or quarterly basis. For the measurement there are different possibilities. One is the coefficient of seasonal variation in terms of standard deviation. Other possibilities can be the seasonal ratio (the largest value divided by the average) or the seasonal indicator (the average divided by the largest value), which could be seen as the average occupancy rate. Some simple additional calculations would be the difference between maximum and minimum or the maximum divided by the minimum. One can also apply the maximum or the minimum on the total share. Another possibility is the Gini coefficient.
The Gini coefficient measures inequalities, in other words, the deviation from a uniform distribution of demand. The data is arranged from the smallest to the highest and accumulated to a so-called Lorenz curve. The Gini coefficient is then calculated by dividing the area between the perfect diagonal and the Lorenz curve by the area between the diagonal and the x-axis. The coefficient can lie between zero and one. The higher the coefficient, the more unequal the distribution, therefore the higher the seasonality. Another highly intuitive way of measuring seasonality is the plotting of data in graphs. To illustrate the findings, data can be shown in absolute figures or in relative figures, with the resulting curves indicating the pattern and changes in the pattern.
New Zealand is located in the southern hemisphere and has a temperate climate. Due to its geographical location a marked summer season, with warmer temperatures and lower rainfall than average, occurs in the months of December to March. As in destinations in the northern hemisphere, the climatic summer months coincide by both tradition and design with the main school and business holidays, which start in mid-December and end on the 6th of February. This can be assumed to be one of the main causes for seasonality. In New Zealand these two key factors of seasonality coincide with a third major factor, the Christmas period. When three causes of seasonality coincide, an increased level of seasonality is to be expected.
New Zealand is an island nation and visitors to New Zealand have the choice of only two modes of arrival and departure, by sea and by air. Indeed, “99.66 percent of all New Zealand’s arrivals and departures are by aircraft. Only a few people arrive by cruise boats.” (Landvogt, 2006) As Landvogt argues in his 2006 paper “Seasonality of Tourism in New Zealand ”, the economic imperatives of the airline industry, in particular the imperative of spreading demand evenly throughout the year, should exercise a moderating influence on the seasonality of New Zealand. Similarly, Landvogt argues that the high share of visitors to New Zealand from the northern hemisphere with their different patterns of demand caused by main travel times of the source countries being diametrically opposite to the main travel times of New Zealand should be in favour of a lower seasonality in New Zealand.
For a more in-depth discussion of the particularities of the New Zealand tourism market the reader should refer to Landvogt’s 2006 paper “Seasonality of Tourism in New Zealand”.
In comparison with other countries, New Zealand possesses exceptional data material. As New Zealand is an island situated at a distance from the neighbouring landmasses auf Australia and Asia, the means of arrival are exclusively aeroplanes and ships. The exact number of arrivals and departures is collected at the airports and ports of New Zealand. Additional data comes from other surveys such as the International Visitor Survey (IVS), the Domestic Tourism Survey (DTS) and the Commercial Accommodation Monitor (CAM). Data is easily accessible on the Internet as New Zealand runs a policy of publishing this data as a resource for decision makers in the tourism industry.
The data of this paper has been extracted from the file “Regional Tourism Organisation (RTO) Area by Accommodation Type ” published on the Statistics New Zealand website. This file contains data from July 1996 to February 2006, with the number of establishments, capacity, occupancy rates, guest nights, guest arrivals, length of stay and some ratios of these variables listed.
All figures contained in the table are collected from the target population on the last day of the month and the ratios of these variables are computed. The target population of the survey is all 'geographic units' ,called 'establishments' in the file, that are classified as short-term commercial accommodation providers operating in New Zealand and belonging to an economically significant 'enterprise'. Economic significance is generally determined as being GST (Goods and Services Tax) registered and having a turnover of at least $30,000 per annum.
The survey aims for 100 percent coverage of the population. In practice, however, an overall response rate of between 76 and 80 percent is usually achieved. The remaining units are given imputed values based upon the characteristics of similar establishments in the same or similar regions. Imputation introduces unknown errors into the estimates, whose size is difficult to quantify and which have therefore been disregarded.
Only accommodations fulfilling the following criteria are included in the survey:
- Establishments are not temporarily closed for more than 14 days during a month
- Establishments must have a GST turnover of at least $30,000
- Establishments are not primarily offering accommodation for periods of one month or more.
As this paper is concerned with only with the accommodation type hotels, it is important to note that the figures in the table include the figures for both hotels and resorts. This does, however not pose a problem, as hotels and resorts are sufficiently similar from an operational point of view that they can be grouped and evaluated together.
Definitions concerning the data published by Statistics New Zealand follow ; additional comments by the authors are highlighted in italics:
The smallest statistical unit operating within a single physical location and owned by a single enterprise. The term is used to represent what is usually called the 'geographic unit' in other Statistics New Zealand publications.
A guest night is equivalent to one guest spending one night at an establishment. For example, a motel with 15 guests spending two nights would report provision of 30 guest nights of accommodation.
The term used to describe the unit of accommodation that is available to be charged out to guests (e.g. a powered site in a caravan park, a bed in a backpackers, a room in a hotel or motel).
Stay unit nights / Room nights:
A stay unit night is equivalent to one stay unit of one establishment occupied for one night. In the case of hotels and resorts one stay unit is one room, hence the term room night.
This is the basic measure of an establishment's accommodation capacity. It is defined as one stay unit multiplied by one night. For example, 10 units in a motel available for guest use (whether occupied or not) for the full 31 days in July would have an accommodation capacity of 310 stay unit nights. However, in practice the daily capacity is reported on the last day of a given month and the total capacity for the month is calculated by multiplying the capacity on the last day of the month by the number of days in the month. Thus, if the motel from the above example has on all but the last day of July 10 units available, but only 9 units available for guest use on the last day of the month, for instance due to maintenance work in one unit, this figure will be used as the daily capacity for the month. By multiplying the so derived daily capacity of 9 units by the 31 days of July a monthly capacity of 279 rather than the actual figure of 309 stay unit nights will be entered in the data set.
This derived variable is calculated by dividing stay unit nights occupied by stay unit nights available. In the case of the motel above, if six of its 10 units were occupied every night in July, it would have 6 x 31 = 186 stay unit nights occupied, and its occupancy rate would be 60 percent. If, however, the monthly capacity has been miscalculated due to the aforementioned reasons as being 279 stay unit nights, the occupancy rate would thus be miscalculated at 66 percent rather than the actual 60.19 percent. In actual practice the occupancy rate for one month seems to be derived directly from occupancy rates reported by the establishments, rather than being computed in the way described by Statistics New Zealand. For further information on this topic the reader should refer to the chapter “Methodology” of this paper.
Average length of stay (stay length):
This derived variable is calculated by dividing total guest nights by total guest first nights
First, the data for this paper was extracted from the file “Regional Tourism Organisation (RTO) Area by Accommodation Type” published on the Statistics New Zealand website and examined for the maximum and minimum values. The averages and the variation of the maximum and the minimum from the average in percent were computed. The extracted raw data contained the following items:
The name of the month and the year for all months from July 1996 to February 2006 are listed here. All other data is sorted by month.
The number of establishments surveyed in any given month is listed here. The maximum number of establishments was recorded in December 2001 with 585 establishments surveyed, the minimum with 539 in July 1997. The average number of establishments surveyed is 566, with a variation of 4.66 percent of the maximum and the minimum from the average over the observed period.
The daily capacity, as defined before, in any given month is listed here. The maximum daily capacity was recorded in November 2005 with 30814 stay unit nights available to guests, the minimum with 21634 in July 1996. The average daily capacity is 26244, with a variation of 17.57 percent of the maximum and the minimum from the average over the observed period.
The monthly capacity, as defined before, in any given month is listed here. The maximum monthly capacity was recorded in January 2006 with 952134 stay unit nights available to guests, the minimum with 643888 in February 1997. The average monthly capacity is 798550, with a variation of 19.37 percent of the maximum and the minimum from the average over the observed period.
Occupancy Rate (%):
The occupancy rate, as defined before, in any given month is listed here. The maximum occupancy rate was recorded in February 2005 with 74.01 percent, the minimum with 36.79 percent in June 1998. The average occupancy rate is 53.61 percent, with a variation of 31.38 percent of the maximum and the minimum from the average over the observed period.
The number of guest nights recorded in any given month is listed here. The maximum number of guest nights was recorded in March 2005 with 1016491 guest nights, the minimum with 376782 guest nights in June 1998. The average number of guest nights is 696377, with a variation of 45.89 percent of the maximum and the minimum from the average over the observed period.
The number of guest arrivals, (number of first night stays) recorded in any given month is listed here. The maximum number of guest arrivals was recorded in February 2005 with 580372 guest arrivals, the minimum with 203376 guest arrivals in June 1998. The average number of guest arrivals is 389097, with a variation of 47.73 percent of the maximum and the minimum from the average over the observed period.
The stay length, as defined before, in any given month is listed here. The maximum stay length was computed in August 1998 with an average length of stay of 1.99 nights, the minimum with 1.65 nights in March 1997. The average stay length is 1.8 nights, with a variation of 8.4 percent of the maximum and the minimum from the average over the observed period.
Stay Unit Nights:
The number of stay unit nights recorded in any given month is listed here. The maximum number of stay unit nights was recorded in February 2006 with 614689 stay unit nights, the minimum with 257470 stay unit nights in June 1997. The average number of stay unit nights is 429392, with a variation of 40.04 percent of the maximum and the minimum from the average over the observed period.
Guests per Stay Unit Night:
The number of guests per stay unit night in any given month is listed here. It is computed by division of the guest nights by the stay unit nights. The maximum number of guests per stay unit night was computed in January 2005 with 1.81 guests per stay unit night, the minimum with 1.46 in June 1998. The average number of guests per stay unit night is 1.62, with a variation of 9.65 percent of the maximum and the minimum from the average over the observed period.
Stay Units per Establishment:
The average number of stay units per establishment in any given month is listed here. It is computed by division of the daily capacity by the number of establishments. It therefore reflects the average size of hotels in New Zealand during the observed period. The average number of stay units per establishment was computed in June 2005 with 53.95 stay units per establishment, the minimum with 38.50 in August 1996. The industry average of the average number of stay units per establishment for the observed period is 46.43, with a variation of 17.06 percent of the maximum and the minimum from the average over the observed period.
Guest Night % of January Guest Nights:
As a simple measure of seasonality, the guest night figures of the month of January of each calendar year is taken as the base-line against which the guest nights of all other months of the same calendar year are measured as a percentage. The maximum percentage computed was recorded in November 2000 with 117.96 percent, the minimum with 58.25 percent in June 2003. The average percentage is 89.86 percent, with a variation of 35.17 percent of the maximum and the minimum from the average over the observed period.
The extracted raw data was further examined in order to establish which data sets were most suitable for measuring the seasonality of the hotel industry of New Zealand.
Daily and monthly capacity:
As discussed before, the available data contains some systematic errors, which are mainly due to the calculation of the daily and monthly capacity figures. As the daily capacity is measured only on the last day of the month and the monthly capacity is computed from this figure, the monthly capacity listed in the data set may be higher or lower than the actual monthly capacity measured on a day-to-day basis. It is impossible to compute the size of the error exactly. Although the capacity offered to guests might show seasonal patterns, those have not been researched in this paper, as the reliability of the data is too low for the discussed reasons.
Occupancy rate and stay unit nights:
According to Statistics New Zealand the occupancy rate in percent is calculated by dividing “stay unit nights” by stay unit night’s available (“monthly capacity”). However, closer examination of the data in the hidden cells of the Microsoft Excel worksheet showed no such calculation. In actual fact, the occupancy rate for any given month seems to be a given figure, with the stay unit nights calculated from the occupancy rate figure by multiplying the occupancy rate with the monthly capacity figure. As occupancy rates are not usually contained in publicly available statistical data on accommodation providers collected by organisations comparable to Statistics New Zealand, but stay unit nights usually are, the decision was made to utilise the stay unit night’s figures in this paper in order to insure comparability of the findings on an international level.
Guest nights and guest arrivals:
These figures have a high degree of reliability, as they are reported directly from the establishments to Statistics New Zealand. Errors in these figures may occur due to respondent error or errors in processing. While Statistics New Zealand is making every effort to minimise those errors, they still occur. The effect is not quantifiable. Apart from the high degree of reliability the data possesses, comparable figures are usually contained in statistical data collected by other organisations comparable to Statistics New Zealand, and the results of this research are therefore comparable on an international level. Therefore both guest nights and guest arrivals were used to measure the seasonality of the hotel industry of New Zealand.
Stay length, guests per stay unit night, stay units per establishments and guest night % of January guest nights:
Whilst stay length and guests per stay unit night could potentially show seasonal patterns, no calculated data has been used as the base data for this research.
To briefly summarise, the data used in this paper to measure the seasonality of the hotel industry of New Zealand is the following:
- Guest nights
- Guest arrivals
- Stay unit nights.
These three sets of data all share one characteristic – they are monthly figures. As such they are not directly comparable to each other. Each month contains between 28 and 31 days. Therefore, ceterus paribus, a month with 31 days will automatically show higher figures in the observed categories than a month with 28 days. To improve the comparability of the figures, the daily averages were calculated by dividing the given figures by the number of days in each month. Leap-years were taken into account. Thus, the following derived figures were calculated:
- Average Guest Nights / day
- Average Arrivals / day
- Average Room Nights / day.
For an in-depth discussion of this calendar effect the reader should refer to Landvogt, 2006. Further calendar effects on seasonality mentioned by Koenig & Bischoff (2004) and Frechling (2001) are the number of weekends per month or year and the variability of religious holidays. Bar-On (1975) argues that movable feast days, public holidays, religious festivals etc. should also be taken into account when measuring seasonality. This paper does not address any of these issues.
A further complication when assessing and measuring the seasonality of the hotel industry of New Zealand can be found in the fact that the reported daily and therefore the monthly capacity change on a monthly basis. During the observed period of nearly 10 years the maximum monthly capacity, which was recorded in January 2006, was 19.37 percent higher than the average monthly capacity of 798550 stay unit nights during the observed period whilst the maximum daily capacity, recorded in November 2005, was 17.57 percent higher than the average daily capacity of 26244 stay unit nights during the observed period. Clearly, in a month with a higher recorded capacity the number of guest nights and the number of room nights can be higher than in a month with lower capacity, as the capacity may not be exceeded in any given month. To fill the available capacity to a degree given by the occupancy rate, a certain number of arrivals are necessary.
To place all months on a level footing and to make the months better comparable, the figures, which had previously been adjusted for the calendar effect, were further adjusted to allow for this capacity effect. To this end, the figures for average room nights / day were adjusted first. This was done by multiplying the average room nights / day with the period average of 46.43 rooms per establishment and dividing the result by the actual number of rooms per establishment in the month concerned. In other words, it was calculated how many room nights would have been observed in a particular month if the number of rooms per establishment had been 46.43 for the given number of establishments at a given occupancy rate. The results of this calculation are the adjusted room nights. The period average was chosen over an arbitrary number, such as, for instance, 50, in order to stay as close as possible to reality. The effect of choosing the period average rather than an arbitrary number is that in a graphical representation the timeline of the development of room night figures over the period is only rotated around the average point, rather than rotated around the arbitrary point and shifted from the average to the arbitrary point. Chart 1 illustrates this effect. However, choosing an arbitrary number would have the advantage to make seasonality data of other industries or destinations more comparable to the data on the New Zealand hotel industry if the chosen number is the same for all industries or destinations compared. The problem here would lie with the choosing of the arbitrary number, as it should represent all industries and destinations as closely as possible.
Through the adjustment of figures to an even level reality is distorted to a greater or lesser degree. The underlying assumption here must always be the ceterus paribus assumption. In particular it has to be tacitly assumed that if the number of rooms per establishment had been the same throughout the period, the occupancy rate would have remained the same. This is not necessarily the case, as any hotel manager will readily testify.
Chart 1: Timeline Room Nights all Season
In a second step the adjusted guest nights were computed by multiplying the adjusted room nights with the number guests of guests per stay unit night. In order to derive the adjusted arrivals figures the previously calculated adjusted guest nights were divided by the stay length.
Through the calculations described above the following figures were derived:
- Adjusted room nights
- Adjusted guest nights
- Adjusted arrivals.
Although the average variance between the adjusted figures and the daily averages is only 0.8 percent, the extreme values are 20.58 percent and -13.95 percent respectively.
Prior to measuring the seasonality of the New Zealand hotel industry, the data sets were grouped into seasons running from July of one calendar year to June of the following calendar year. Thus the season 96/97 contains the data of the months July 1996 to June 1997. This was done to prevent the natural peak during the aforementioned climatic summer months from being cut in half. The reader should take note that this definition of seasons does not conform to either the calendar year or the New Zealand fiscal year which starts at the beginning of April. For a more detailed discussion of the advantages and disadvantages of defining the season of New Zealand in this fashion the reader should refer to Landvogt (2006).
The previously extracted, calculated and grouped data sets
- Guest nights,
- Guest arrivals,
- Stay unit nights,
- Average Guest Nights / day,
- Average Arrivals / day,
- Average Room Nights / day,
- Adjusted room nights,
- Adjusted guest nights and
- Adjusted arrivals
were investigated with various measures of seasonality. The following measurements have been performed:
- The sums of all data sets were calculated for each season.
- The averages of all data sets for each season were calculated.
- The maximum (Max), that is the figures of the month with the highest figures was determined for all seasons and data sets
- The minimum (Min), that is the figures of the month with the lowest figures was determined for all seasons and data sets.
- The difference between Max and Min was calculated for all seasons and data sets.
- Max was divided by Min for all seasons and data sets, thus giving an indication of seasonality by comparing the weakest with the strongest month
- The share of the Sum was calculated for both Max (Max share) and Min (Min Share) for all seasons and data sets
- The seasonality ratio and seasonality indicator were calculated for all seasons and data sets.
- The standard deviation and the coefficient of variation were calculated for all seasons and data sets.
- The Gini-Coefficient was calculated for all data sets and the seasons 96/97 to 04/05 as well as for the whole observation period.
Research Paper, 163 Pages
Master's Thesis, 79 Pages
Master's Thesis, 141 Pages
Diploma Thesis, 119 Pages
Term Paper (Advanced seminar), 33 Pages
Research paper, 27 Pages
GRIN Publishing, located in Munich, Germany, has specialized since its foundation in 1998 in the publication of academic ebooks and books. The publishing website GRIN.com offer students, graduates and university professors the ideal platform for the presentation of scientific papers, such as research projects, theses, dissertations, and academic essays to a wide audience.
Free Publication of your term paper, essay, interpretation, bachelor's thesis, master's thesis, dissertation or textbook - upload now!