The transformation on the Dutch car market
"What are the new preferences of the Dutch car consumer?"
Tom Wierda s1747711
'Master thesis'
Business Administration:Marketing management
The transformation on the Dutch car market
"What are the new preferences of the Dutch car consumer?"
University of Groningen Department of Marketing 29 august 2012 First supervisor : Dr. J.E.M.van Nierop Department of Marketing University of Groningen
P.O. Box 800, 9700 AV Groningen The Netherlands j.e.m.van.nierop@rug.nl Second supervisor: D.A. Naydenova MSc. Department of Marketing University of Groningen
Preface
This report is my master thesis for the conclusion of my Master program of Business administration; specialization Marketing management at the University of Groningen.
First of all, I want to make a special thanks to my supervisor Erjen van Nierop for the smooth cooperation. Despite of the distance between the current working place of Erjen van Nierop in Leusden and Groningen, I experienced a very good support.
Furthermore, I want to make a special thanks to my second supervisor Daniela Naydenova for reading and judging my thesis.
Third, I want to thank all my respondents, who took time and effort to fill in my questionnaires.
Table of content
Summary ...4
1.Introduction and problem statement ...5
1.1 Introduction ...5
1.2 Dutch car market information. ...6
1.3 Introduction literature ...7 1.4 Research questions ...9 2.Theoretical framework ... 10 2.1 Gasoline prices ... 10 2.2 Buying behavior ... 12 2.2.1 Eco labels ... 12 2.2.2 Environmental concern... 13 2.2.3 Car attributes ... 14 2.3 Tax incentives ... 18 2.4 Conceptual model ... 21 3 Research design ... 22 3.1Data collection ... 22
3.2 Population and research method ... 22
3.3 Method of analysis ... 22
3.4 Survey design and rationale ... 23
3.4.1 Input Conjoint analysis ... 23
3.4.2 Consumer characteristics ... 26
3.5 Attributes questionnaire ... 27
4 Description sample ... 29
4.1 Response on survey ... 29
4.2 Analysis consumer characteristics ... 29
4.3 Analysis of current car characteristics ... 30
4.4 Constructing new variables ... 31
5 Results ... 31
5.1 Explanatory analysis consumer characteristics ... 31
5.2 Conjoint analysis ... 33
5.3 Comparing mean values of segments ... 35
5.3.1 Gender ... 35 5.3.2 Income ... 37 5.3.3 Status ... 38 5.3.4 Degree of usage... 39 5.3.5 Car class... 40 5.3.6 Product involvement ... 41 5.3.7 Environmental beliefs ... 42
5.4 Moderating effects of environmental beliefs... 43
6 Conclusions ... 44
6.1 Main findings... 44
6.2 Managerial implications ... 45
6.3 Limitations and future research ... 46
Literature ... 46
Summary
In recent years the Dutch car market has radically changed. Cars with green energy labels have multiplied almost 6 times their market share within 4 years time. The Dutch car consumer is being motivated by the Dutch government to buy green labelled cars using tax incentives, reducing the purchase prices (by not charging BPM) or running costs (by charging less road tax). Running costs are further reduced, as these green labelled cars are also more fuel economical. Importantly, these cars are also environmentally friendly and thus helping the Dutch car consumer to create a greener world.
It is not yet clear which of these drivers are most important for the Dutch car consumer. This study aims to reveal the underlying motivations of the Dutch car consumer to buy green labelled cars using a conjoint analysis. In addition to the study-specific variables "Tax
incentives "and "Energy labels". the conjoint analysis includes more traditional variables like Country of origin, Gas prices, Performance of the car, in order to create a good view of the importance of the new variables. On top of this, the influence of the consumer’s
environmental beliefs is included to monitor its moderating role in the buying process of the Dutch car consumer.
The outcomes of this study reveal a major influence of "Tax incentives" (28%) and "Energy labels"(24%). The Dutch car consumer is highly influenced by tax incentives and green energy labels in the car buying process. Strikingly, comparison of segments based on
1.Introduction and problem statement
1.1 Introduction
The Dutch car market has changed radically the past decades. Many new Asian brands have entered the market or have managed to capture a decent market share. Furthermore, the total owner costs for driving a car rose to an all-time peak. Fuel prices have never been this high and the road tax have peaked historically.
Moreover, the crisis has a big impact on the consumers’ buying behaviour. Consumers are more and more cautious before making big expenses. On the other hand, the Dutch
government is stimulating consumers to buy low emission cars. Every new car is provided with an energy label, reflecting the car’s emission. The cars with the lowest level of emission get a “Green” energy label (A) and the cars with the most emission get a “Red” energy label (G).
Energy label A cars provide the consumer many tax advantages; the purchase prices of these cars are lower compared to similar cars and the cleanest cars are free from road tax for the consumer. This tax benefit has become a very important extrinsic product cue for the Dutch car consumers. The market share of these clean cars, which benefit from this tax cue, have multiplied over 5 times in the last 4 years (see figure 1). This fact demonstrates the influence of this cue on the buying behaviour of the Dutch car consumer. Importantly, the tax cue also has a positive effect on the environment. Energy label A cars have a low emission and are thus less polluting than label B-G cars.
So, buying an A label car enables consumers to (i) be gentle on the environment, (ii) take advantage of the considerable tax advantages and (iii) benefit from the lower fuel
1.2 Dutch car market information.
In 2011 more than 8 million cars were registered in the Netherlands. In the year 2010 more than 483.000 new passenger cars were sold in the Netherlands.
Figure 1 shows the car sales based on energy label. The energy labels are based on CO2
emission of a car. The labels A, B and C are known as the “Green” labels and provide the consumer access to specific tax advantages. Label A provides more advantages than label B and label B more than label C. Cars with high polluting emission receive a “yellow”/”red” label and have to pay an extra fine on top of the regular tax incorporated in their purchase price, to make it less attractive for the customer to buy polluting cars. This extra tax is based on the level of emission: More emission results in a higher purchase price. Furthermore, the cleanest cars (with an emission below the 95 gram CO2 per kilometre for diesel engines and
109 gram for petrol engines) are free from road tax. As can be seen in figure 1, cars with label C and D were the most sold cars in 2007. Over the years the sales numbers of the “Green” label cars have grown. In 2007 only 5% of the sold cars had an A label, in the year 2010 this number was multiplied almost six times to a market share of 29.8%. This shows that
consumers are sensitive for environmental friendly cars in combination with the tax advantages. Figure 1 Source;CBS/RDC 0 20 40 60 80 100 120 140 160 180 Numbers x 1000
Type of Energy label
Sales per energy label
Another explanation for this raise in sales numbers is shown in figure 2, the price of fuel. As shown, the fuel price has risen to a record high. The latest record price was on 7 May 2012: 1,81 euro for one litre of euro95. In just three years time fuel prices increased with almost 50%, as can be seen in figure 3. The green label cars are more economical in fuel
consumption, which helps to reduce the costs of possessing and using a car.
Figure 2
Source: CBS
1.3 Introduction literature
Luchs et al. (2010) discovered that environmental cues can create negative or positive associations towards a product. Previous conception was that environmental and sustainable product cues have a positive contribution towards the products in the evaluation by the consumer (Leonidou 2010; Irwin and Naylor 2009). Luchs et al. (2010) elucidated that this conception is not uniformly useable to all products, but it differs among product categories. Another finding was that these environmental cues could even have a negative effect among other product cues. Luchs et al. (2010) used "Strength" related attributes in search for consumer preference and found out that consumers judged "Strength" related products significantly lower when the cue sustainability was involved.
Environmental cues could have a positive or negative contribution to the perceived quality evaluation of the consumer. Perceived quality is described in the literature as an important purchase predictor of consumers (Monroe & Dodds 1985). Consumers are willing to buy a product which has the best price/quality ratio. The several product cues of a product influence directly the perceived quality view of the consumer. Product cues can be divided into two
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2 2006 2007 2008 2009 2011 2012 € p e r lit re
Gasoline costs Dutch market
groups, the physical, intrinsic cues, and the extrinsic cues so non-physical attributes of the product itself (Jacoby & Olsen 1972). Intrinsic product cues are for example colour, engine and resolution of a TV screen. Examples of extrinsic cues are for example price and brand name, but in this case also the energy labels of cars belong to these extrinsic cues.
When looking at the sales numbers of the Dutch car market, it becomes clear that the environmental/tax cue has a big impact on the product evaluation by the consumer. The consumers could have become really environmentally sensitive, meaning that a green energy label is perceived as a hallmark of a good quality car. Another driver could be the price. The green label cars have attractive prices and are also economical in fuel consumption. The purchase prices of these environmentally friendly cars are significantly lower because of the tax advantage. Also the costs of use are lower because of the combination of tax advantages and the lower fuel consumption.
In contrast, the low purchase price could have a negative effect on the consumer perceived quality of the car (Zeithaml 1988).The consumer would judge it as a cheap car with lower quality perceptions, belonging to a cheaper product. Especially consumers with little knowledge of intrinsic cues use extrinsic product cues like price and brand name as a
surrogate for quality. Interestingly, previous studies (Royne 2011; Trudel & Cotte 2008) have shown that consumers are prepared to pay the similar price (or more) for a “Green” product compared to a regular product. These findings suggest it is not necessary to reduce the purchase price to stimulate green label car sales.
Another interesting question is how this environmental cue affects other cues, for example the cue “Performance”. Luchs et al. (2010) claim that in certain product categories the
environmental cue has a negative affect on other opposite product cues. In the case of cars, an environmental, fuel economical car is the opposite of a sports or premium car. For this
1.4 Research questions
Based on the findings mentioned above, the following main research question with sub questions has been developed:
Main research question:
- “What is the influence of the environmental cue on product preference for the Dutch car consumer?”
Sub questions:
1. “Which product attributes of a new car influence the Dutch car consumer's purchase intention the most?”
2. “What are the findings of past research on the topic environmental buying behaviour?” 2.1: “What is the influence of eco labelling on the purchase intention of the Dutch car consumer”
2.2: “What is the influence of pro environmental beliefs on purchase intention of the Dutch car consumer?”
3. “What are the findings of the past research on the topic effect of gasoline prices?” 3: ” What is the influence of gasoline prices on purchase intention of the Dutch car consumer?”
4. “What are the findings of the past research on the topic tax incentives for automobiles?” 4 : “What is the influence of tax incentives on purchase intention of the Dutch car
consumer?”
5.” Which way of communication, environmental oriented or economical oriented, is the most effective in influencing the purchase intention of the Dutch car consumer?”
Mentioned before, the main goal in this research is to get insight in the driving forces of the radical changes of the Dutch car market in the last 4 years.
In chapter 2 a theoretical framework will be employed. Different past studies will be collected and discussed with subjects which will be related to the problem statement. This theoretical framework will result in the development of hypotheses. These hypotheses create the input for our research design. In chapter 3 the research design will be outlined, describing how the hypotheses will be tested. Also, choice of research design will be substantiated. The
hypotheses will subsequently be tested and the results will be described and discussed. The last chapter will exist of conclusions, implications and suggestions for future research.
2.Theoretical framework
2.1 Gasoline prices
Evidence can be found in literature that consumers are not very price sensitive in relation to gasoline. Willenborg et al. (1977) describes gasoline as an extremely price-inelastic product. Willenborg et al. (1977) observed a price increase of 40-50% for gasoline in a two year period. During this period the number of registered cars even increased.
The current situation of the Dutch market (figure1, section 1.3) is extremely similar, with gasoline prices currently at a record high and a peak in the total number of cars registered in the Netherlands, with 8 million registrations (2011, see figure 3).
Figure 2
Source: RDC/CBS
Car ownership Dutch market
Figure 4
Price development public transport vs. car transport
Source:CBS
A good additional explanation for the raise in car ownership is shown in figure 4, illustrating the divergence of cost increase for public transport vs. private car use. Public transport prices relatively raised much more then car transport has in the last years. So it becomes more attractive from a price perspective to choose for car transport over public transport.
However, there was a small increase in sales of compact cars and cars which are more fuel economical, in the research of Willenborg et al. (1977). Willenborg et al. concluded that the price mechanism was relatively ineffective in reducing the consumption of gasoline. Findings of Pitts, Willenborg & Sherrel (1981) support this conclusion. Only radical price increases of 40% or more made a big difference in the behavior of households. In a period of 4 years during which the price increased 20%, the total mileage driven by the households had even increased. The average engine capacity dropped only with 5% (allowing a more economical use) over a 4 year period. Thus, the households see the gasoline prices as standard and adapt to these prices. Only a big increase of 40% or more in a one year period made an impact on behavior. Households decreased their amount of driven miles significantly with more than 15%. In the one year period with a 40% price increase, the average engine capacity only dropped 1,7%. Thus, households first reduce their total mileage rather than replacing their car with a more economical model. Only specific segment groups whose financial condition will not accommodate the higher price increases in a short time period showed a strong reaction. These findings therefore underline the inelastic nature of the consumers’ response to
fluctuations in gasoline prices. Price increases of gasoline do not discourage driving nor do they support changes in car ownership; More recently, also Brons et al.(2008) came to similar
conclusions on the topic of price sensitivity of gasoline. On the long and short run consumers are not very sensitive to gasoline price fluctuations. On the long run the price-sensitivity is slightly higher compared to the short run. According to Brons et al.(2008), consumers need time to adjust to the price increases. The consumers will first change their way of driving, so will drive more economical, and will reduce their mileage. Thus in general consumers become more critical in using their car. Consumers are to a lesser degree willing to change their car ownership in response to gasoline price increases. Furthermore, when people are more dependable on automobile transport they get less sensitive to fluctuations of gasoline prices. According to Brons et al. (2008), the pricing policy of gasoline could be more effective in combination with road tax and charges on vehicle purchases. This suggestion of Brons et al. (2008) is identical to the current situation on the Dutch market.
In contradiction, Diamond et al. (2009) suggests prices of gasoline would be a good predictor of sales of hybrid and fuel efficient cars. According to Diamond et al. (2009), future prices of gasoline are very uncertain for consumers, so they choose a fuel-efficient car as insurance for future volatility and price spikes.
Based on the findings mentioned above, the following hypothesis has been developed:
H1: Raising gasoline prices has no influence on the purchase intention for green labeled cars.
2.2 Buying behavior
2.2.1 Eco labels
In 2001 eco labelling for passenger cars has been introduced on the Dutch car market. These labels indicate how much gasoline the car uses and how much CO2 the car produces. These
figures tell the consumer how green and environmentally friendly a car is. The use of these labels make it easier for the consumer to see if a car is environmentally friendly or not. By law these eco labels have to be visible for the consumer in all promotion material and in the showrooms of new car sellers, .
According to Thøgersen (2000), the use and attention paid to these eco labels is strongly influenced by pro-environmental behaviour. These consumers have the belief they are protecting the environment. Furthermore, the trust factor of these labels is very important. When consumers conceive these eco labels as reliable, they will develop more pro
Thøgersen (2000) claims with this study that consumers believe in pro-environmental purchase behaviour to create a greener world, but how strong these believes are depends on personal pro-environmental attitude and partly on personal traits. Interestingly, Grankvist et al. (2004) discovered that not every consumer is sensitive to environmental information through eco-labels. Consumers without or with only very weak pro-environmental beliefs are not sensitive to the information of eco-labels.
In practice eco-labels constitute extra information and extra products in the perspective of the consumer. What is more, the eco-label information is often experienced as too much
information and hard to understand. Furthermore, eco-labels produce extra products, which create too much choice in the perspective of the consumer (Horne et al., 2009).
In the case of the Dutch car market it can be concluded that the eco-labels of cars could have a positive influence on the sales of environmentally friendly cars. How strong this influence of these labels is, is still hard to say. The eco-labels for the Dutch car market are government regulated which is favourable for the acceptance, as this increases the trust factor of the labels (Horne et al., 2009). Due to these labels more pro-environmental beliefs and behaviour is stimulated, and consumers who already posses a pro-environmental attitude will create stronger beliefs when confronted with these eco-labels.
Based on the findings mentioned above, the following hypothesis has been developed:
H2: Eco-labels have a positive influence on the purchase intention of green labeled cars.
2.2.2 Environmental concern
According to Gupta et al. (2009), green consumers are high “trusters” and expect that other people also engage in green buying. High “trusters” believe that other people also have a green buying behaviour and because of this belief they will themselves buy green products sooner.
Individuals with strong pro-environmental beliefs will buy green products sooner (Mainieri et al., 1997), but pro-environmental beliefs will not lead directly to green buying behaviour. Also Leonidou et al. (2010) has shown that there are different levels of pro-environmental behaviour. Such behaviour varies between active buying behaviour and being
pro-environmental but not being active in consuming these products. Hence, when general and green products are interchangeable, the consumer will choose for the green products. Also Wong et al. (1996) and Kim and Choi (2005) found out that environmental concern is positively correlated with green purchase behaviour. Eco-information can trigger this concern according to Siriwardena et al. (2012).Also Choi et al. (2011) showed that sustainable
information has a positive impact on the consumers’ product evaluation and purchase
intention. This information positively effects green buying behaviour, but the eco-information has to be repeated over and over, in order to stay effective. Education and creating consumer awareness for environmental concerns positively triggers green buying behaviour (Teisl et al., 2008).
Based on the findings mentioned above, the following hypothesis has been developed:
H3: Pro-environmental beliefs have a positive influence on the purchase intention of green labeled cars.
2.2.3 Car attributes
In contradiction, according to Griskevicius et al. (2010) green consumers are able to use green hybrid cars as status symbol. This section of fanatic green consumers chooses a pro-social green product over more luxurious same-priced general products. These green products are especially desirable for these particular green consumers when the products are more expensive than the traditional non-green products, further reinforcing the owner’s status. According to Noblet et al. (2006) eco-attributes like air quality and pollution are to be
considered in the purchase decision of cars by the majority of consumers, but are on their own not strong enough to change the market. Noblet et al. (2006) further suggests that incentives to stimulate the purchase of “green” and “environmental” cars could be an effective option. In conclusion, if green labelled cars want to succeed, green labelled cars will have to be available in every vehicle class, allowing every consumer to choose for a green label in his preferred class.
Fetscherin et al. (2009) investigated which product attributes influence the price paid by the consumer when purchasing a new car. The categories which were found to be positively related are ‘chassis’, interior’, ‘comfort’, ‘engine’ and ‘safety’. These categories drive the price up when they match the consumers’ needs. However, the product attributes
‘environment’ and ‘economical’ were found not to be significant. These attributes were not important enough for the consumer to pay more for a new car.
According to Teisl et al. (2008), vehicle purchases are for 75% influenced by habit/loyalty and consumer needs. This means that consumers want to buy the same class of vehicle that they own at the moment when buying a new vehicle in the future. The majority of the
consumers does not want to switch to another class car, for example from a small city car to a MPV class car or otherwise. This is because the primary driver of vehicle class choice is related to the uses and needs of the vehicle. Hoen et al. (2011) came to the same conclusion when investigating the Dutch car consumer and their detailed preference for passenger cars: The Dutch car consumer in general has a strong preference for similar cars to the car which they drive at the moment. Specifically switching to another car class and engine capacity are the product attributes which possess the most negative values. In another words, the Dutch car consumer is not prepared to switch to another class of car or give in on engine performance in comparison to the current car which they possess.
Based on the findings mentioned above, the following hypothesis is developed:
Another strong cue influencing the purchase intention of the car consumer is the Country of origin (COO) effect, which indicates the car’s COO most preferred by the consumer.
Figure5
Source; RDC
As shown in figure 5, most cars for the Dutch car market are imported from Germany.
European cars are the most preferred cars of the Dutch car consumer, followed by Asian cars. The three countries Germany, French and Japan are good for almost 75% of the car import. The largest brands are Volkswagen, Opel and Ford for Germany. Peugeot, Renault and Citroën for France. Toyota, Suzuki and Nissan for Japan. Countries who’s market share in the Netherlands is growing in the last decade are South Korea, with the brands KIA and Hyundai, and the Czech Republic with the brand Skoda.
USA Car brands are mostly lacking from the Dutch car market with a market share below 1% in 2010. A plausible explanation could be the higher costs of ownership for the consumer, related to the fact that these cars are usually bigger than European and Asian cars. A major downside of these bigger cars is that gasoline consumption is higher and the size makes them heavier than most European and Asian cars. As road tax in the Netherlands is based on vehicle weight the consumer will pay more road tax, further increasing costs of ownership. The Dutch consumers’ aversion to American car brands could thus be an indication of the price sensitivity of the Dutch car consumer. Furthermore, the weight and gasoline
consumption also has its impact on the energy labelling of these cars: Only a few models of American cars obtain a green label.
Based on the findings mentioned above, the following hypothesis has been developed:
H5: COO effect has negative influence on the purchase intention of green labeled cars.
Carsales based on COO
2.3 Tax incentives
The Dutch government is currently using tax incentives to stimulate the sales of green labelled cars. As mentioned before, the tax incentives give a reduction on the sales price of a car and an extra incentive on the road tax, depending on the car’s CO2 emission. The main
advantages for the consumer are the reduction of the cost price and a reduction on the periodical costs. The consumer could be price sensitive for these tax incentives.
Bijmolt et al. (2005) has shown that consumers have become more price sensitive in the last decades. Especially with durable goods consumers are more price elastic. In the growth stage of product categories the price elasticity is most strong. This interesting finding could explain the high demand of green labelled cars in the Netherlands.
In contrast, according to Johansson et al. (1985) the price of cars has a positive effect on the quality perception of the consumer. Higher priced cars are perceived to possess more quality and vice versa high quality cars are perceived to be higher priced. However, price has a negative effect on purchase intention because of the budget constraint. This last observation is interesting in relation to the tax incentives. The tax incentives create a bigger group of
consumers that can afford a (more expensive) new green labelled car, as budget constraints are in part elevated.
Diamond et al. (2009) investigated the influence of an incentive system on the sales of fuel economical hybrid cars in Mexico. Local policy in Mexico used incentives in different stages of strength to stimulate the consumers to buy fuel economical hybrid cars. Diamond et al (2009) concluded that none of these forms of incentives were significantly related to a positive change in market shares of these cars. Sales and excise waivers did have a marginal positive effect and in this study were found to be more effective than rebates and tax credits. Furthermore, the monetary incentives would for the most part preferentially benefit the higher incomes, as these consumers can afford the more expensive new fuel efficient hybrid cars. By nature, this segment of consumers is less price sensitive.
respondents were significantly more willing to switch to a smaller engine, so to choose a car with a green label.
The smaller engines give the car a green label because of the use of less gasoline and the reduction of CO2 emission. On the other hand the smaller engines result in lower
performance. To acquire a green car, the consumer thus has to accept a car with less power, which negatively affects the car’s acceleration and top speed. The majority of the consumers in the study of Haan et al.(2009) was found to be prepared to give in on this performance attribute in turn for the financial gain related to the sales price and gasoline costs. In contrast, they were found to be less willing to switch to a smaller car, except when previously owning the higher classes like SUV, sport cars and luxury cars, because these cars do not have an engine earning them a green label. Importantly, Teisl et al. (2008) shares the conclusions made by Haan et al. (2009): The consumer is willing to switch to green labelled cars, if there is a financial benefit for the consumer they are willing to switch to a smaller engine, but the majority is not prepared to switch to another (lower) car class.
However, when examining the situation of the last few years on the Dutch car market we see a different situation: In the last 5 years there has been significant movement in market shares of vehicle classes (see figure 6).
Figure 6
Source; CBS/RDC
Segment A (mini- class) and segment B (compact- class) which are the smallest cars of all segments, have increased their market share while all other segments declined or stabilized
Marketshare per segment
0% 20% 40% 60% 80% 100% 2007 2008 2009 2010 2011 Year Others Segment L/M (SUV) Segment J/K (MPV) Segment E (Higher mid-class)
Segment D (Mid-class) Segment C (Smaller mid-class)
Segment B (Compact-class)
their market share. Especially segment A registered a major increase of almost 100% in market share between 2007 and 2011. Segment A was also the most sold vehicle class on the Dutch car market in 2010 with a market share of 26%. These figures show that the Dutch car consumer is prepared to switch to a smaller vehicle class.
The correlation of the number of A label cars in every segment with the increase in market share could be an explanation for this movement. In the smallest car segments like the mini class and compact class, the number of A label cars is larger than in other and larger car segments. This is because smaller cars contain smaller engines and weigh less compared to the other segments, allowing them to consume less gasoline and emit less CO2. In other words
the smaller cars segments benefit more and faster from green energy labelling in comparison to other segments.
Based on the findings mentioned above, the following hypotheses are developed:
2.4 Conceptual model
Figure 7 shows the conceptual model which has been constructed for this research and that will be implemented in chapter 3. As can been seen, the main purpose is to find out which variables influence the PI (Purchase Intention) of Dutch consumers for green labeled cars the most. Also we assume a moderating role exists for the variable “Environmental beliefs”. Furthermore, the control variable "Consumer characteristics" will be used in explanatory analysis. Consumer characteristics consist of demographic characteristics, current car characteristics, product involvement and environmental beliefs.
The purpose is to find out which variables are the most important in the perspective of the Dutch car consumer.
3 Research design
3.1Data collection
The data required for our statistical analyses was collected by means of an internet survey. An internet link directing to the online questionnaire was spread using email, internet-fora and social media. The program used to generate the survey and to collect all data is called “Thesistools.com”.
3.2 Population and research method
The population studied consists of potential Dutch car consumers: Dutch nationals of 18 years or older, holding a driving license. These people all were considered potential car consumers, as they are able to drive a car.
We had a useful response of 118 respondents. We were aiming for a total respondents around 120. This is because we segmented the population in different groups and tried to get a size of each group near or larger of 30 respondents to create a equal distribution. We have worked with a confidence interval of 95% or 90%. This are the most commonly used values (Burns, Bush & Swart, 2006). The research was conducted by using an online questionnaire. The purpose was to get a wide range of respondents in context of demographic characteristics, car characteristics, environmental beliefs and product involvement. The questionnaire has been construct based on a conjoint design. Through this questionnaire we collected our data. Furthermore we investigated the environmental beliefs of each respondent. To measure this we used the scale of Kim & Choi (2005). Also have we investigated the product involvement of the respondents using the scale of Zaichkowsky (1985).
3.3 Method of analysis
We employed the most commonly used conjoint analysis "full profile" (Malhotra, 2007) in combination with regression analysis to provide insight in the relative importance of each attribute in the purchase intention of the Dutch car consumer for green labeled cars. To construct the conjoint analysis stimuli, we have chosen to use a "full profile approach", to make the profile as realistic as possible. The attributes:
- Tax incentives
- Country of origin (COO) - Performance
All 5 attributes are represented in the profiles. Furthermore, as 5 attributes with each three levels are used, the number of profiles can be reduced further with the "full profile"method in comparison to the "pair wise" approach (Malhotra, 2007).A total of 3x3x3x3x3= 243 profiles can be constructed. To reduce the total numbers of profiles a fractional factorial design called the orthogonal design was used, resulting in a set of 16 profiles that constitute the stimuli. To measure the utility per attribute multiple regression analysis was used. Prior to using this analysis the data was dummy coded. Furthermore, utility scores per individual were
calculated to create segments in the data. The mean scores of these segments has been compared with the independent sample T test or ANOVA test, depending on the amount of segments. To calculate the importance for each attribute and to identify which attribute is the most important for the Dutch car consumer the part-worth’s alphas were computed.
Furthermore, to calculate the interaction effect of the variable "Environmental beliefs" we have multiplied the items of the variable "Energy labels" with the variable "Environmental beliefs" and took these new variables within the existing multiple regression analysis
3.4 Survey design and rationale
3.4.1 Input Conjoint analysis
To measure the importance of each attribute of the purchase intention of the Dutch car consumer each attribute was divided in different levels which were used in the conjoint
analysis. We have told the respondents in the questionnaire that they have to imagine buying a new car, in the same car class and size as they currently own. The engine will be a petrol engine.
Explanation per attribute: Gasoline:
For gasoline the current price of 1,79 per liter euro 95 was used. This is the current average price (date 15-07-2012) for all gas stations in the Netherlands. The other price variants will be 10% and 28% higher. The values will be 1.98 per liter Euro 95 and 2,29 per liter Euro95. We have chosen to make one variant a little bit higher with a increase of 10% . The last variant is more extreme, with a price of 2.29 per liter Euro 95. Like mentioned in the theoretical
Eco-labeling:
For eco-labeling we selected the following variance; A, C and E labels. The labels are in practice also colored in three segments: Green, yellow and red, respectively. Green (A) means environmentally friendly and economical in gasoline usage, yellow (C) means average in environmental friendliness and gasoline consumption, and red (E) is not environmentally friendly and not economical in gasoline usage.
Tax incentive:
The Dutch government makes use of three different variants of tax incentives, that are also used in our survey. These incentives are in the real world setting related to the amount of emission and this way linked to the energy labels. Only the low emission vehicles and in this way only the green labeled cars profit from these tax incentives. Variant 1 is called BPM free, this is an incentive on the purchase price. The incentive varies between €2000,- for a small city car and €3500,- for a midsize car. Variant 2 is a combination of BPM free and road tax free. The latter means the consumer does not have to pay road tax. Road tax for a midsize car amounts to around 700-800 euro a year. The last variant (3) is no tax incentive. There will be no benefits for the Dutch car consumer in purchase price nor on road tax.
Performance:
We used three variants for the attribute performance: (i) the same performance as the
Country of Origin (COO) effect
Country of origin (COO) was used in three stages. We have selected three continental profiles; Europe, Asia and North America. We have knowingly not chosen for separate countries in our profiles, because of the large number of countries which produce cars. In Europe alone there are already more than 10 different car-producing countries. Which would make our conjoint plan to large. Importantly, no large Dutch car brand is available on the market, reducing the value of including individual countries in this particular survey.
The respondents were to judge 16 different profiles on a scale from 1-10, were 1 represents “I do not like this at all” and 10 represents “I like this very much!” Every profile represents the 5 attributes, but the profiles all contain different levels of the 5 attributes.
Example of a profile used in the questionnaire (figure 8):
3.4.2 Consumer characteristics
To enable a good description of our sample, questions in relation with consumer
characteristics were asked. First, demographic questions to register age, gender, income, education and family situation were included. Then subjects were questioned about the car they currently own. This included questions concerning the car class and their mileage during last year. The mileage provided insight in the frequency of car use of the respondents. User frequency was used as a segmentation tool, segregating the respondents in three types of users:
- Heavy users - Moderate users - Light users
Furthermore, the environmental concern of each respondent was measured, as explained in the theoretical framework outlined above. To measure the environmental concern items of the research of Kim & Choi (2005) were used (listed below) and measured on a 7 points Likert-scale.
1: I am extremely worried about the state of the global environment and what this will mean for my future.
2: Mankind is severely abusing the environment.
3: When humans interfere with nature it often produces disastrous consequences. 4: The balance of nature is very delicate and easily upset.
To measure the product involvement of each respondent, the scale of Zaichkowsky (1985) was used. These items (listed below) were measured on a 7 points Likert-scale.
1: I would be interested in reading information about how the product is made. 2. I would be interested in reading the Consumer Reports article about this product. 3. I have compared product characteristics among brands.
4. I think there are a great deal of differences among brands. 5. I have a most-preferred brand of this product.
3.5 Attributes questionnaire
All attributes and their levels which were used in the online questionnaire are described below, with an explanation for the respondents of every attribute. The complete questionnaire is showed in appendix A.
Please fill in your gender. 1: Gender:
Male or female
Please fill in your age. 2: Age:
Between 18-99 years
The average gross income for a Dutch household is 2700 euro per month. What is your gross income per month (or together with your partner)?
3: Income per year: - < 2700
- Average 2700
- Between 2700 and 5400 euro - > 5400
- I do not want to answer this question
What is your current family situation? 4: Family situation:
- Single - Cohabiting - Single with kids - Cohabiting with kids
How much kilometers did you drive last year with your own private car? If you do not own a private car fill in the kilometers of the car you drive the most?
5: Kilometers per year with private car: - <10.000
- >30.000
Which car class do you currently drive? If you cannot find the right class, make an estimation of your own car and choose a car class which fits the best in your view.
6: Current car class:
- Small city car (For example: Renault Twingo) - Compact car (For example: Peugeot 206)
- Middle class (For example: Volkswagen Golf, Opel Astra) - Larger middle class (For example: Volkswagen Passat) - Upper class (For example: BMW 5 serie)
- SUV (For example: Honda CRV)
- MPV 7 person (For example: Renault Espace) - Sports car (For example: Coupe, Convertible) - Midi MPV 5 persons (For example: Opel Zafira) - Van (For example: Volkswagen Caddy)
The next session will consist of several statements. Please indicate to which level you agree with the statement.
7: Items for environmental beliefs:
1: I am extremely worried about the state of the global environment and what it will mean for my future.
2: Mankind is severely abusing the environment.
3: When humans interfere with nature it often produces disastrous consequences. 4: The balance of nature is very delicate and easily upset.
8: Items for product involvement:
1: I would be interested in reading information about how the product is made. 2. I would be interested in reading the Consumer Reports article about this product. 3. I have compared product characteristics among brands.
4. I think there are a great deal of differences among brands. 5. I have a most-preferred brand of this product.
Next the respondent is asked to imagine the following situation:
presented with several profiles of cars with different characteristics. Please judge the profiles on a scale from 1-10 to indicate how much you like the car (1 = dislike, 10 = like).
4 Description sample
4.1 Response on survey
The survey has been spread using an internet link. This link has been sent to an e-mail list, placed on different social media like Facebook and has been posted on different internet sites, such as Fok forum and Autoweek.
After one week reminders were sent by e-mail and social media to realize an increase of the response rate. The total number of respondents was 173but not every respondent was useful. After removal of incomplete responses we were able to use 118 completed questionnaires for our analyses.
4.2 Analysis consumer characteristics
In table 1 the demographic characteristics of the study population are presented. The average age of the responders was 41 years (not presented in table) – a bit lower than the average age of the Dutch car consumer, which is around 51 years (RDC/ CBS). The majority of the responders were male, and over 50% of the responders was Cohabiting. This latter item may also reflect the younger than average age among responders.
Furthermore, the sample is spread nicely in terms of income. Ten respondents didn`t wanted to answer the question what their monthly income is. The education level is rather high with almost 45% representing HBO and lower educated people aren`t represented very well in our population with only 15 respondents.
Table 1
Study population: Descriptive demographic consumer characteristics
Items Frequency Percent Gender= Male 74 62,7 Female 44 37,3 Status= Single 27 22,9
Single with children 3 2,5
Cohabiting 60 50,8
Cohabiting with children 28 23,7
Income per month= <€2700,- 23 19,5 Average = €2700,- 24 20,3 Between €2700,- and €5400,- 42 35,6 >€5400,- 19 16,1 No answer 10 8,5
Highest education= VMBO 3 2,5
HAVO 8 6,8
VWO 4 3,4
MBO 28 23,7
HBO 53 44,9
WO 22 18,6
4.3 Analysis of current car characteristics
In this part the characteristics of the respondents’ current cars and its usage are shown (see table 2). The distribution of each car class in our population is remarkably balanced, with a small bias towards the larger cars. The degree of usage based on kilometers per year shows a bigger divergence, with the majority of our population driving between 10.000 and 30.000 km per year.
Table 2
Current car of responders: Descriptive car characteristics
Items
Frequency Percent
Car class= Smaller cars 36 30,5
Midsize cars 38 32,2
Larger cars 44 37,3
Degree of usage in Kilometers per year=
Less than 10.000 Km 35 29,7
Between 10.000 and 30.000 Km 63 53,4
More than 30.000 Km 20 16,9
4.4 Constructing new variables
For the constructs “Environmental beliefs” (4 items) and “Product involvement” (5 items), a Cronbach's alpha test has been done, to determine if they can be treated as new variables. Both constructs have alpha scores above the requisite 0,7 score (see table 3), meaning the separate items of the two constructs can be treated as two new variables in our analyses The two new variables are calculated by taking the mean of the five items for "Product involvement" and the mean of the four items of "Environmental beliefs".
Table 3
Outcomes Cronbach's Alpha
Items Cronbach's Alpha N of Items
Product involvement ,749 5
Environmental beliefs ,704 4
5 Results
5.1 Explanatory analysis consumer characteristics
In this section multiple regression is applied on the variables "Environmental beliefs" and "Product involvement". We would like to see if there are variables in terms of demographic and car characteristics explaining the degree of "Environmental beliefs" and "Product
involvement". These outcomes have been compared with later segmenting outcomes to see if they match with each other in terms of importance values. These outcomes explain if the variables "Environmental beliefs" and "Product involvement" also actually been taking in account in the purchase intention of the Dutch car consumer.
Table 4
Outcomes Multiple-regression based on "Environmental beliefs" Items T Sig. 95% B Std. Error Beta (Constant) 3,949 ,558 7,076 ,0008 Age ,006 ,255 2,848 ,005 ,017*
Education level middle
,117 ,343 ,045 ,341 ,734
Education level high
,069 ,304 ,030 ,227 ,821
Degree of usage middle
,028 ,226 ,013 ,123 ,902
Degree of usage high
-,672 ,297 -,229 -2,262 ,026*
Gender
,091 ,213 ,040 ,427 ,671
Size car class midsize
-,244 ,249 -,104 -,981 ,329
Size car class bigsize
-,533 ,246 -,234 -2,165 ,033* *P≤0.05;
Table 5 below illustrates that only the item "Gender" is significant influencing the degree of product involvement. Male respondents own more product involvement in comparison to females. The size of car class is also influencing product involvement. Respondents with a midsize or a bigsize car hold more product involvement compared to respondents with a small car. Noted that the significance level for these outcomes is 90%.
Table 5
Outcomes Multiple-regression based on "Product involvement"
T Sig. 90% B Std. Error Beta (Constant) 5,067 ,680 7,455 ,000** Age -,005 ,007 -,065 -,716 ,475
Education level middle
-,152 ,417 -,049 -,363 ,717
Education level high
Degree of usage middle
,334 ,275 ,126 1,214 ,227
Degree of usage high
,456 ,362 ,130 1,261 ,210
Gender
-,612 ,259 -,225 -2,361 ,020**
Size car class midsize
,584 ,303 ,207 1,929 ,056*
Size car class bigsize
,522 ,300 ,192 1,740 ,085*
*P≤0.10; **P<0,05
5.2 Conjoint analysis
As mentioned in chapter 3 have we conducted an orthogonal design of the variables which has been used in the questionnaire. To calculate the utility scores and importance values of the variables and their items we have conducted a conjoint command for the total population in SPSS. The results of these analyses are presented in table 6.
Table 6
Outcome conjoint plan with utilities
Variables:
Items: Utility Estimate P-valueEnergy label
Energy-label A ,761 0,00
Energy-label C ,033 0,29
Energy-label E -,793 0,00
Gas price
Gas price of €1.79 per litre ,559 0,00
Gas price of €1.98 per litre ,097 0,03
Gas price of €2.29 per litre -,655 0,00
Performance
Same performance as current car ,725 0,00 15% less performance compared to current
car -,150 0,01
30% less performance compared to current
car -,576 0,00
Tax incentive
No tax incentives -,867 0,00
BPM free -,183 0,00
BPM free and roadtax free 1,050 0,00
COO
European manufactured car ,481 0,00
American manufactured car -,276 0,00
Asian manufactured car -,204 0,00
Figure 8
The importance values of each variable conducted in percentages is represented in figure 8. These results provide us the insight that the variable "Tax incentives" is the most important variable in creating purchase intention among the Dutch car consumer, thus partially
confirming our hypotheses H6. Especially the item "Road tax free" makes a car very attractive for the Dutch car consumer. In contrast, the item "BPM free" is a tax incentive with a negative value in the buying process of cars. Although the Dutch car consumer profits from a lower purchase price with this tax incentive, the rating towards this incentive is negative. "No tax incentives" has been rated as the most negative item of this variable.
The second most important variable is "Energy labels". A green energy label, energy label "A" , makes a car more attractive. Non-green labels, so cars with high CO2 emission, own
negative value and holds the strongest influence towards this variable. These findings underpin hypotheses H2, indicating that "Energy labels" are positive related towards green labeled cars creating purchase intention among Dutch car consumers.
The third variable "Performance" has an importance value of 19%. Lower performance is being judged negatively. The item "Same performance as current car" owns the highest value of this variable and is positive. This is in line with our hypotheses H4, but performance is not judged as the most important variable. The research of Hoen et al. (2010) discovered a large impact on "Performance" under Dutch car consumers, which does not correspond with our results. One explanation for this discrepancy could be the lack of the variables "Tax incentives" and "Energy label" in the research of Hoen et al. (2010).
24%
18% 28%
19% 11%
Importance values Dutch car consumer
Furthermore, the variables "Gas price" (18%) and "Country of origin"(11%) are the least important for the Dutch car consumers. A high gas price of €2.29 per liter has been judged negatively, partially confirming our hypothesis (H1). Remarkably, the second item "Gas price € 1.98 per liter" presents a small positive value in importance. Only when the price increases with almost 30% it receives a negative value. These results are in line with the studies of Willenborg et al. (1977), Pits et al. (1977) and Brons et al. (2008), which showed that
consumers are not very price sensitive for gas. Interestingly, the variable with the lowest score is Country of origin(COO). This variable has often been described in the literature as a very important and strong cue, influencing the purchase intention of the car consumer. In this research it holds the lowest importance value of only 11%. "European manufactured car" contributes a positive value to the purchase intention and is the most important item of this variable. These results refute our hypotheses H5 according to the COO effect. A plausible explanation for this discrepancy could be the lack of a general domestic brand on the Dutch car market.
5.3 Comparing mean values of segments
In this chapter different segments are compared based on "Gender", "Income", "Status", "Car class", "Degree of usage", "Product involvement" and "Environmental beliefs". Comparing the mean results of the different segments gives us insight were the groups significantly differ from each other in strength of purchase intention of a car. A table for the variable "Age" and "Education level" is not included. Segments in these variables did not reveal significant differences. In chapter 5.1 we revealed the remarkable variable "Age", that had a positive relationship towards the variable "Environmental beliefs". This means that older and younger people differ in their environmental beliefs, but not as significantly in the purchase intention of a car.
5.3.1 Gender
importance value (26%) and creates the most purchase intention. Figure 9 shows per variable the differences of importance values between men and females.
In contrast, for men the variable "Tax incentives" holds the highest importance (27%). Men and women significantly differ on this variable (Tax incentives). The item "BPM free plus road tax free" hold the highest positive value for men. Women also judge this positively but not as high as men. The item "No tax advantages" is the most negative value of all items for the male population. Especially the addition of "Road tax free" is important for men.
Remarkable is the negative appreciation of men towards the item "BPM free". The last item where gender differences are seen is on the COO item 'European cars'. Men judge these cars more positively than women. The explanation for this is given in chapter 5.1 and in the coming section 5.3.6, were we show that men have a higher product involvement.
Table 7
Outcomes independent sample T-test on gender
Items Male Female Sig. 95%
N=74 N=44
Constant 3,89 4,23 0,06
Energy-label A 0,57 1,09 0,00*
Energy-label C 0,02 0,05 0,42
Energy-label E -0,59 -1,14 0,00*
Gas price of €1.79 per litre 0,50 0,65 0,08
Gas price of €1.98 per litre 0,12 0,07 0,35
Gas price of €2.29 per litre -0,62 -0,72 0,23
Same performance as current car 0,68 0,79 0,29
15% less performance compared to current car -0,13 -0,18 0,35
30% less performance compared to current car -0,56 -0,61 0,38
No tax incentives -0,89 -0,82 0,35
BPM free -0,33 0,06 0,00*
BPM free and roadtax free 1,22 0,76 0,00*
European manufactured car 0,57 0,33 0,02*
American manufactured car -0,32 -0,20 0,19
Asian manufactured car -0,25 -0,12 0,20
Figure 9
5.3.2 Income
Below the ANOVA table based on income is shown. Income is divided in 4 groups (A) "below €2700,- per month", (B) " Average = €2700,- per month", (C) "Between €2700,- and €5400,- per month" and (D) "More than €5400,- per month". This ANOVA test analyses the groups seperately from eachother based on the mean scores per item. The variable
"Performance" holds the most differences. The group with the highest income (D) judge the item "Same performance" most importantly. All other groups (A,B,C) judge this item
positive, but significantly lower as the group with highest income (D). This variable is for the group with the highest income (D) also most important. Furthermore, the variable "Tax incentives" is the most important is for the group with lowest income (A). Group (A) differs especially significantly from group (B) and (D).
Table 8
Outcomes ANOVA test based on income
Items Below €2700,- (A) Average = €2700,- (B) Between €2700,- and €5400,- (C) More than €5400,- (D) Sig. 95% N=23 N=24 N=42 N=19 Constant 4,66 3,71 3,92 4,01 A>BC* Energy-label A 0,65 0,84 0,72 0,85 Energy-label C 0,21 -0,12 0,10 -0,07 Energy-label E -0,86 -0,72 -0,82 -0,78
Gas price of €1.79 per litre
0,52 0,51 0,57 0,68
Gas price of €1.98 per 0,26 -0,03 0,15 0,01 A>B* 0,0% 5,0% 10,0% 15,0% 20,0% 25,0% 30,0%
Importance values based on gender
litre
Gas price of €2.29 per litre -0,78 -0,48 -0,72 -0,69 Same performance as current car 0,35 0,67 0,70 1,33 D>ABC* 15% less performance compared to current car
0,07 -0,21 -0,14 -0,32
30% less performance compared to current car
-0,43 -0,46 -0,56 -1,01 D>ABC*
No tax incentives -1,23 -0,77 -0,86 -0,62 A>BD*
BPM free -0,13 -0,17 -0,21 -0,21
BPM free and roadtax free 1,35 0,94 1,07 0,83 European manufactured car 0,39 0,44 0,59 0,42 American manufactured car -0,03 -0,33 -0,39 -0,23
Asian manufactured car -0,36 -0,11 -0,20 -0,19
Bold is P<0.05 different to 0 * P< 0,05 between groups
5.3.3 Status
This ANOVA (Table9) is based on the status situation. Status is divided in three groups; (A) "Single", (B) "Cohabiting" and (C) "Cohabiting with children". The group which differs from the rest are the singles (A). Singles don`t judge the variable "Energy labels" very important, significantly lower as the other groups. Furthermore, for the group "Cohabiting with children" (C) the item "Same performance" is more important compared to "Singles"(A). Explanation could be the amount of persons (children) and luggage this group (C) has to carry within the car.
Table 9
Outcomes independent sample T-test on status
Items Single (A) Cohabiting (B) Cohabiting with children (C) Sig. 95% N=27 N=60 N=28 Constant 4,30 4,02 3,77 Energy-label A 0,48 0,79 0,95 C>A* Energy-label C 0,32 -0,01 -0,14 A>BC* Energy-label E -0,80 -0,78 -0,81
Gas price of €1.79 per litre 0,49 0,51 0,71
Gas price of €1.98 per litre 0,14 0,06 0,14
Gas price of €2.29 per litre -0,62 -0,57 -0,85
Same performance as current car
0,41 0,78 0,90 C>A*
compared to current car 30% less performance compared to current car
-0,38 -0,63 -0,64
No tax incentives -1,01 -0,82 -0,84
BPM free -0,25 -0,16 -0,18
BPM free and roadtax free 1,25 0,98 1,01
European manufactured car 0,59 0,49 0,37 American manufactured car -0,27 -0,27 -0,29
Asian manufactured car -0,32 -0,21 -0,09
Bold is P<0.05 different to 0* P< 0,05 between groups
5.3.4 Degree of usage
Table 10 shows the differences based on the degree of usage. Degree of usage is divided in three groups (A) "Light users", (B) "Average users" and (C) "High users". The degree of usage is based on the amount of kilometers the respondents drove over the last year. The "High users"(C) group differ the most from the other two groups on the variables
"Performance" and "Gas price". "High users"(C) dislike a high gas price more compared to the other which is logically because this group faces higher costs on gas due to higher degree of usage of a car. Other difference is the variable "Performance". For the "High users"(C) group is "Performance" the most important variable."Same engine performance" is the item with the highest importance value for the high users group. Also the second item "15% less performance" is been judged significantly lower by the other two groups compared to the "High users" (C) group.
Table 10
Outcomes ANOVA test based on degree of usage
Items Light users
(A) Average users (B) High users (C) Sig. 95% N=35 N=63 N=20 Constant 4,28 4,00 3,61 A>C* Energy-label A 0,81 0,77 0,64 Energy-label C 0,02 0,01 0,14 Energy-label E -0,83 -0,78 -0,78
Gas price of €1.79 per litre 0,52 0,52 0,75
Gas price of €1.98 per litre -0,03 0,15 0,15
Gas price of €2.29 per litre -0,49 -0,67 -0,90 A>C*
Same performance as current car 0,45 0,71 1,27 C> AB*
current car
30% less performance compared to current car
-0,42 -0,62 -0,72
No tax incentives -1,00 -0,79 -0,87
BPM free -0,03 -0,27 -0,18
BPM free and roadtax free 1,03 1,06 1,05
European manufactured car 0,28 0,55 0,60
American manufactured car -0,18 -0,36 -0,20
Asian manufactured car -0,11 -0,19 -0,41
Bold is P<0.05 different to 0 * P< 0,05 between groups
5.3.5 Car class
In table 11 below are the significant differs displayed based on the current car class
respondents own. The group is divided in three groups (A) "Smaller cars", (B) "Average cars" and (C) "Larger cars". The variables "COO" and "Performance" contain significant
differences. For the group with larger cars (C) is "Performance" holds the most important variable and judge the items "same performance" and 15% less performance" significantly more important as group "Smaller cars" (A). Furthermore, owners of larger cars judge the item "European cars" more important and do they dislike "Asian cars" significantly more compared to the group with smaller cars(A).Explanation for this is given in chapter 5.1 were we found out that respondents with larger cars hold higher values of "Product involvement" Another plausible explanation could be the lack of larger car models of Asian car brands. In Europe existing more car brands with larger car models for example: BMW, Mercedes Benz and Audi compared to Asia. If the owners of larger cars have to choose for a Asian car brand they also to choose for a smaller car model.
Table11
Outcomes ANOVA test based on car class size
Items Smaller cars
(A) Average cars (B) Larger cars (C) Sig. 95% N=36 N=38 N=44 Constant 4,14 4,28 3,69 A>C* Energy-label A 0,78 0,87 0,65 Energy-label C 0,11 0,06 -0,05 Energy-label E -0,89 -0,93 -0,60
Gas price of €1.79 per litre 0,50 0,57 0,60
Gas price of €1.98 per litre 0,00 0,18 0,10
Same performance as current car 0,44 0,66 1,02 A<C*
15% less performance compared to current car
0,05 -0,23 -0,25 A>C*
30% less performance compared to current car
-0,49 -0,44 -0,77
No tax incentives -0,83 -0,96 -0,81
BPM free -0,19 -0,16 -0,20
BPM free and roadtax free 1,02 1,12 1,01
European manufactured car 0,31 0,47 0,62 A<C*
American manufactured car -0,31 -0,30 -0,23
Asian manufactured car 0,00 -0,17 -0,39 A>C*
Bold is P<0.05 different to 0* P< 0,05 between groups
5.3.6 Product involvement
Table 8 shows the ANOVA results for the groups with (A) "Low involvement",(B) "Average involvement" and (C) "High involvement" on cars. They differ among the variable "COO". People with "Low involvement"(A) don`t hold a strong preference for cars out of a certain country. The people with" High involvement"(C) and "Average involvement"(B) like
European cars more and dislike Asian cars more compared to people with" Low involvement" (A). This can be explained due to the larger group of males in the high involvement group and due to the fact that respondents with higher involvement owning larger car classes. (see chapter 5.1 and 5.3.1)
Table 11
Outcomes ANOVA test based on product involvement
Items Low involvement (A) Average involvement (B) High involvement (C) Sig. 95% N=39 N=44 N=36 Constant 4,01 4,10 3,93 Energy-label A 0,78 0,79 0,70 Energy-label C 0,20 -0,02 -0,08 Energy-label E -0,98 -0,77 -0,62
Gas price of €1.79 per litre 0,51 0,59 0,58
Gas price of €1.98 per litre -0,01 0,19 0,10
Gas price of €2.29 per litre -0,50 -0,77 -0,68
Same performance as current car 0,57 0,67 0,96
15% less performance compared to current car
-0,12 -0,13 -0,20
30% less performance compared to current car
-0,45 -0,54 -0,75
