Introduction

Buildings account for 40% of Europe’s final energy consumption (COM 2019; ODYSEE-MURE 2018). Non-residential buildingsFootnote 1 of different types, including hotels, account on average for 25% of energy consumption from the total European building stock, which makes them major contributors to overall CO2 emissions (WBG 2020). Moreover, non-residential buildings are on average 40% more energy-intensive than residential buildings; they also require more electrical energy than residential buildings (286 kWh/m2 compared to 185 kWh/m2) (D’Agostino et al. 2017).

Energy consumption in buildings varies from one European Union (EU) country to another. Differences may be explained by climate conditions, by building characteristics (e.g. building envelope, insulation level) and by social and cultural factors (lifestyle, habits, etc.) among others. Energy demand in non-residential buildings may also be influenced by economic growth and employment growth (Bertoldi et al. 2018).

Energy efficiency (EE) provides an opportunity to substantially reduce energy consumption and consequently GHG emissions from the services sector (Hrovatin et al. 2016; Liang et al. 2019; Schleich 2009; Schlomann and Schleich 2015) and also to reduce energy-related running costs in that sector (Mavrotas et al. 2003; Patel and Guedes 2017; Sakshi et al. 2020). Several studies have analysed the potential energy savings, avoided CO2 emissions and profitability of EE investments (Cattaneo 2019; Fleiter et al. 2012). However, despite its significant monetary benefits and environmental advantages, EE adoption levels are generally low, as shown by the EE gap (Jaffe and Stavins 1994; Linares and Labandeira 2010; Shama 1983). A better understanding of barriers regarding large investments such as HVAC systems is particularly important in addressing this EE gap and encouraging more efficient purchases. In fact, studies exploring what barriers are most relevant to EE in the hotel industry provide important information for the design of effective policies for the promotion of energy EE (Schleich 2009).

Spain provides a very interesting case study for this sector due to the importance of tourism in the country (INE 2019; UNWTO 2019) and its significant energy consumption. The wide range of climates and types of hotel industry establishment in the country also provide a chance to explore the connection of these factors with the adoption of energy-efficiency measures, thus providing insights for other countries where tourism is also a significant part of the economy.

In Spain, 73% of the energy demand from buildings in 2017 (including residential and non-residential buildings) came from heating (40.4%) and electrical equipment (32.5%). Hot water accounted for 12.5%, air conditioning for 9%, and cooking for 5.5% of consumption (ODYSEE-MURE 2020). Focusing on non-residential buildings, 7% of the country’s final energy consumption was attributable to restaurants and accommodation establishments (IDAE 2017). Most of the energy used came from electricity (74%), followed by natural gas (17%) and petroleum products (7%), while renewables had a direct contribution of 2.4% (IDAE 2017). The National Integrated Energy and Climate Plan 2021-2030 (PNIEC) (MITECO 2020), which defines the priorities for the energy transition, proposes measures to promote renewable energies and EE in heating systems among other points.

In that context, this study sets out (i) to analyse barriers that influence the consideration of EE by hotels in Spain; and (ii) to explore the effect of several drivers on the valuation of EE. The data used were collected expressly for this study from 200 hotels throughout Spain. The survey seeks to (i) assess the importance attributed to EE achieved through HVAC installation; and (ii) identify the most common barriers that prevent tourist establishments from investing in energy-efficient HVAC systems. A binary response model is estimated to explore whether these barriers and a set of other factors influence the consideration of EE in the purchasing decision, i.e., the consideration of EE as a very important attribute in the purchase of HVAC systems. Note that these barriers are analysed in isolation and no trade-off is allowed between them. That is, the analysis is ceteris paribus, so that it provides and understanding of the impact of one attribute when the rest remain constant. The results provide insights for policy makers for effective incentives to encourage the adoption of energy-efficient HVAC systems so as to decarbonise energy consumed in heating and cooling at hotels.

The rest of paper is organised as follows: the "Barriers to energy efficiency" offers an analysis of the barriers to the wider use of efficient HVAC systems, with a particular focus on non-residential buildings. The "Methodology" section describes the survey and the econometric model. The "Results and discussion" section presents and discusses the results. Finally, the "Conclusion" section concludes and outlines policy implications.

Barriers to energy efficiency

Barriers to the adoption of EE measures have been analysed previously in the relevant literature for both residential (Gerarden et al. 2017; Linares and Labandeira 2010; Solà et al. 2020) and non-residential buildings (Cagno et al. 2013; Sorrell et al. 2004). Some of the main findings are described below.

Studies analysing the empirical relevance of barriers for EE in residential buildings have identified cognitive limitations, a lack of financial resources, the principal-agent problem and imperfect information problems as the most important barriers (Ramos et al. 2015; Stieß and Dunkelberg 2013).

Given that the building sector is characterised by dualities in the use of buildings (residential vs. non-residential), some of the barriers for residential buildings also apply to non-residential ones. So, as proposed by Ramos et al. (2015), references to barriers for residential buildings are included in order to draw analogies and provide a more integrated view of barriers to achieving optimal EE levels in residential and non-residential buildings. Further research would be needed to find differences among how these, and other barriers may affect residential and non-residential consumers. The heterogeneity of the households that make investment (or purchase) decisions and of the managers making equivalent decisions suggests that a whole new direction of research would be needed to answer this question.

Barriers to energy efficiency in non-residential buildings

Econometric assessments of barriers in non-residential buildings show that EE investment is driven not only by market and behavioural factors but also by the characteristics of firms, such as organisational factors, the products and services offered by a company, location and profitability. Specific sectors such as industry, commerce and services have been analysed previously (Fleiter et al. 2012; Schleich 2009, 2004; Schleich and Gruber 2008). Three categories of barriers to EE in organisations have been proposed in the literature: (i) market barriers; (ii) behavioural barriers; and (iii) organisational barriers (Cagno et al. 2013; Sorrell et al. 2004). A summary of these barriers is presented in Table 1. Nonetheless, some barriers may fall under more than one category (Sorrell et al. 2011). These barriers are linked below to variables included in this analysis.

Table 1 A summary of the major barriers in non-residential buildings
Full size table

Market barriers include two important groups: informational failures and other market failures. Informational failures are led by imperfect and asymmetric information. These terms refer to a lack of information on costs and energy saving equipment, unclear information by technology providers or poor-quality information as to the energy performances of different technologies. Hidden costs (low perceived profitability of EE investments, additional costs associated with energy-efficient technologies, etc.) and risk and uncertainty (uncertainty about future energy prices, technical risks, economic and technological uncertainties concerning the business, etc.) may also be part of the problem (Cagno et al. 2013). Other market failures are related to access to capital and the principal-agent problem.

Among behavioural barriers, consumers tend to use simple calculations because of bounded rationality (e.g. individuals and companies ignore changes in real energy prices and energy savings that can be made over the lifetime of a good) (Blasch et al. 2019; Cattaneo 2019; Gillingham and Palmer 2014). In addition, Blasch et al. (2018) indicate that consumers need not only specific energy-related and financial knowledge but also the cognitive skills to apply that knowledge, in what they call “energy-related financial literacy”. Moreover, the characteristics of information (specific, simple, personalised, updated, credible or trustworthy) comprise another significant barrier to energy-efficiency investment (Cagno et al. 2013).

Among organisational barriers, power barriers comprise a lack of power and/or influence by those in charge of energy management (lack of energy experts and skills, lack of energy audits, conflicts of interest between individuals and departments with different ideas and values influencing decision-making, or time pressure leading to concentrate solely on core business) and culture barriers refer to a lack of organisational culture leading people to ignore energy issues (Cagno et al. 2013; Hrovatin et al. 2016; Palm 2009).

Fleiter et al. (2012) summarise the main findings of empirical studies addressing the role of barriers for non-residential buildings. They find that the most frequent barriers are lack of information about energy consumption patterns and about EE measures, lack of time to analyse potentials for EE, priority setting within organizations and the principal-agent problem (Schleich 2009; Schleich and Gruber 2008).

Barriers to energy-efficient HVAC systems in the hotel industry

Although the barriers indicated in Table 1 appear to be the most significant at organisations, there are differences by sector and size (Olsthoorn et al. 2017). Most barriers are more pronounced in less energy-intensive firms and in smaller firms (Schleich 2004). The main barriers for small and medium enterprises (SMEs) are a lack of capital and the technical risk of halting production for energy-intensive SMEs. For less energy-intensive SMEs, they are a lack of information and a lack of staff time (Fleiter et al. 2012). Moreover, the significance of these barriers varies from one type of EE measure and technology to another. For example, installing an HVAC system calls for higher investment costs and a greater degree of complexity and customisation than measures involving lighting. This in turn is associated with higher hidden costs (Olsthoorn et al. 2017; Trianni et al. 2014). Specifically, Olsthoorn et al. (2017) conclude that the main barriers for heating replacement are the principal-agent problem, hidden costs (such as financing costs and other investment priorities) and lack of capital. As a result, many firms fail to perceive that investing in energy-efficient appliances reduces operating costs in the future by lowering energy costs.

This paper analyses the factors that influence the adoption of energy-efficiency HVAC systems, focusing on the impacts of different barriers. A distinguishing feature of this study is its focus on behaviour regarding energy consumption at establishments and on investment in green and energy-efficient equipmentFootnote 2. All this contributes to a better understanding of the EE gap in the hotel industry.

As regards behavioural failures, Fadzli Haniff et al. (2013) conduct a detailed review of HVAC scheduling techniques for buildings, analysing energy and cost savings obtained by changing practices. They demonstrate that advanced scheduling techniques (e.g. a combination of “ON” and “OFF” statuses and precooling or preheating techniques) may be able to save up to 20% in energy and 20% in cost for HVAC systems used for heating and cooling buildings. Chedwal et al. (2015) estimate energy savings by applying advanced energy-efficiency HVAC systems in different categories of hotel buildings in India. They find that the payback period for replacing existing HVAC systems with grand source heat pumps (GSHP) is 5–7 years, and that such investments increase profits and make establishments more competitive in the tourism market (Cingoski and Petrevska 2018). Considering the current context of climate change, increases in extreme temperatures will have consequences for energy demand in this sector, not only for heating but also for cooling (Biardeau et al. 2020), thus making these practices even more important.

To explain the EE gap, it is also important to consider energy consumption behaviour at hotels. Owners believe that energy consumption depends not only on the EE of the HVAC system but also on the behaviour of customers. Liang et al. (2019) find that there are barriers to changing behaviour towards energy savings due to attitudes such as inattention and myopia among customers who do not pay the marginal cost of their energy consumption. To address these failures, Tiefenbeck et al. (2019) conduct a field experiment that provides information on energy and water consumption in every shower taken through a smart meter fitted to the hotel room shower unit. They show that real time feedback is a cost-efficient policy instrument for promoting resource conservation among hotel guests.

Related to this, several studies find no significant link between proenvironmental attitudes and investment in EE or energy-saving actions (Hornsey et al. 2016; Kollmuss and Agyeman 2002; Ramos et al. 2016; Schleich et al. 2016; van der Linden 2017), in other words that proenvironmental attitudes may not always translate into substantial proenvironmental action (Cattaneo 2019). Moreover, Nauges and Wheeler (2017) suggest that proenvironmental attitudes may have a negative effect on environmental concerns because agents who invest in energy mitigation behaviour may feel less concerned about climate change. Other behavioural explanations include lifestyle categories that capture the energy culture of a company (i.e. how energy is perceived and what habits and routines are in place). This allows for a deeper understanding of how and why companies improve EE (Palm 2009; Trianni et al. 2017). For the specific case of the hotel industry and investment in energy-efficient HVAC systems, Ramos et al. (2016) find that environmental concerns appear to be significantly less relevant for high-cost energy-efficient investments (such as HVAC systems), suggesting that there may be a trade-off between environmentally friendly behaviours and monetary costs.

Methodology

Survey deployment

Two hundred telephone interviews were conducted in December 2017 and January 2018 by CPSFootnote 3, a specialist polling company. Establishments were recruited so as to provide a representative sample of types of accommodation, climate areas, geographical locations and other characteristics such as star rating and number of rooms. Decision-makers from three types of accommodation—individual hotels, hostels and cottages—were interviewed to explore their attitudes to HVAC systems. A preliminary test of the survey questionnaire was conducted. It included the three types of accommodation indicated plus international hotel chains, but the latter were left out of the final version because their decisions about HVAC systems were found to be centralised at their main offices and often unrelated to their geographical locations. The respondents targeted were individuals in charge of purchasing and investment decisions at the establishments. Seventy percent of the respondents were building owners and the rest had lease agreements. The establishments included in the survey were drawn from five main climate regions in Spain (Mediterranean, Atlantic, Continental, Subtropical and Mountain) and two types of surroundings (coast and inland). They also represented different star ratings. Establishments were also drawn from municipalities of different sizes. For a more detailed explanation of the characteristics of the sample, see Appendix A, Table 4.

The final questionnaire was designed based on the results of eight in-depth interviews conducted with Spanish accommodation operators to identify the key factors in their purchasing decisions. The analysis of these in-depth interviews revealed that consumers do not focus solely on the energy intensity of goods but are influenced by many other factors (de Ayala et al. 2020). Consequently, the goal of this analysis is to explore the interactions of these factors with the consideration of EE in purchasing decisions and to test a large number of factors. A binary response model (probit model) is thus used to explore the effect of several drivers on the weight given to EE. This enables drivers or factors to be analysed in isolation, i.e. kee** the rest of the factors unchanged (ceteris paribus). Several earlier studies have used probit models for this type of approach in the energy context (Hrovatin et al. 2016; Liang et al. 2019; Olsthoorn et al. 2017; Schlomann and Schleich 2015). Other approaches in the literature, such as discrete choice experiments, do not keep other factors constant in order to analyse possible trade-off between a number of factors (Fleiter et al. 2012; Michelsen and Madlener 2012; Schleich 2009; Schleich and Gruber 2008). Both types of approach are useful and complementary for shedding more light on the topic.

The questionnaire contains sections on (i) socio-demographic and economic characteristics; (ii) the characteristics of HVAC systems; (iii) the attributes of the decision whether to purchase HVAC systems; (iv) barriers to EE investment; (v) understanding and use of existing labels and simulated monetary labels and (vi) environmental behaviour, including energy-saving habits and investments in green and energy-efficient equipment. More detailed information on the questionnaire, including all the questions, is presented in Appendix B.

Barriers considered in the survey

The survey conducted here captures most of the relevant barriers for non-residential buildings identified in ‘Barriers to energy efficiency in non-residential buildings’. Specifically, survey participants were asked to indicate their attitudes and beliefs concerning different market, behavioural and organisational barriers to EE. Respondents were asked to answer using a 4-point Likert scale ranging from “strongly disagree” to “strongly agree”. Table 2 provides an overview of the fourteen barriers addressed in the survey.

Table 2 Overview of the barriers addressed in the survey (N = 191)
Full size table

Regarding market barriers, “EE does not vary” and “More energy consumption” reflect the low perceived profitability of EE. “More EE goods are less reliable” and “Uncertainty as to energy prices hinders EE investment” are expected to capture the risk and uncertainty related to investing in energy-efficient heating systems. “Loan access limits my purchases” and “Cannot afford to upgrade” capture the importance of external access to capital due to high interest rates but also internal access to capital because EE investments are often classified as less urgent than essential maintenance projects or strategic investments. And “Effectiveness of energy consumption information” refers to measures to make guests aware of energy consumption, aligning the incentives for energy savings between managers and guests. Behavioural failures are measured using several barriers that account for bounded rationality in EE investment decisions. The survey also inquired about organisational barriers such as willingness to take a chance on new technologies and the comfort and environmental values of the establishments, thus capturing information on the scale of EE investments in the accommodation sector.

Econometric model

A binary response model is built up using a probit model (Greene 2003; Wooldridge 2002). The dependent variable represents the probability of hotels rating EE as a very important attribute of their HVAC investment decision. Based on the literature review in the previous section, we use explanatory variables from seven different categories referring to (i) geoclimatic areas; (ii) socio-economic characteristics; (iii) technical characteristics of HVAC systems; (iv) specific attributes of HVAC systems such as price, brand reliability, performance and noise; (v) barriers to investing in energy-efficient HVAC systems; (vi) knowledge and perception of monetary information labels; (vii) environmental behaviour and (viii) habits for energy savings and investment in green equipment. The general specification of the probit model applied can be expressed as follows:

$$ \Pr\ \left(\mathrm{Y}=1\ \right|\ \mathrm{X}\Big)={\upbeta}_1+{\upbeta}_2\mathrm{GeoClimatic}\ \mathrm{areas}+{\upbeta}_3\mathrm{Socioeconomic}\ \mathrm{characteristics}+{\upbeta}_4\mathrm{HVAC}\ \mathrm{technical}\ \mathrm{characteristics}+{\upbeta}_5\mathrm{HVAC}\ \mathrm{specific}\ \mathrm{attributes}+{\upbeta}_6\mathrm{Barriers}\ \mathrm{to}\ \mathrm{EE}+{\upbeta}_7\mathrm{Monetary}\ \mathrm{information}\ \mathrm{label}+{\upbeta}_8\mathrm{Environmental}\ \mathrm{behaviour}+{\upbeta}_9\mathrm{Habits}\&\mathrm{Investment}+\varepsilon $$
(1)

where Y is “Energy Efficiency is a very important attribute in the purchasing decision”, X contains explanatory variables and ε~N(0, 1) is the error term. The elements of the vector of parameters [β1, …, β9] are reported as the marginal effects of the explanatory variables on the consideration of EE, with all other attributes remaining unchanged (ceteris paribus). With binary independent variables, marginal effects measure a discrete change, i.e. how predicted probabilities change as the binary independent variable changes from 0 to 1. For a continuous independent variable, it measures the amount of change in the probability of considering EE produced by a 1-unit change of the independent variable as very important.

The explanatory variable “Barriers to EE” includes the market, behavioural and organisational barriers described above in ‘Barriers considered in the survey’. All the barriers described in Table 2 were tested and the selection of the final model presented in Table 3 is based on the Akaike Information Criterion.

Table 3 Marginal effects of the EE attribute for hotels and similar establishments in Spain
Full size table

Results and discussion

Descriptive statistics

Descriptive statistics are used to describe and discuss the findings for the full sample in more detail regarding energy-efficiency considerations for the purchase of HVAC systems and to explain their socio-demographic background.

Socio-demographic and economic characteristics of establishments

The socio-demographic and economic characteristics of an establishment may influence the adoption of EE HVAC. The establishments surveyed have been running on average for 17 years, with a range from 1 to 86 years. The average number of rooms is 26, but the range is wide: from 1 to 434. The average occupancy rate of rooms is 80% during the high season and 40% during the low season. In 2017, most establishments considered that their businesses were financially sound, and they did not envisage or anticipate financial problems in the futureFootnote 4. On a scale from 1 (“I am having major financial difficulties”) to 10 (“My financial situation is very comfortable”), the average score is 6 for the present situation and 7 for expectations for the following 5 years (Table 5 in Appendix).

Technical characteristics and specific attributes of HVAC

The HVAC systems used vary from one establishment to another, with more than one type of system in some cases (e.g. cottages with reverse cycle air conditioning system and fireplaces or wood stoves). 27% of establishments have combined HVAC systems, 72% have separate heating systems and 18% also have separate cooling systems. The energy sources used for separate heating systems are heating oil (39%), electricity (18%), biomass (17%), natural gas (13%), propane (9%) and renewable energies such as photovoltaic and geothermal energy (4%). Combined systems and separate cooling systems use electricity. On average, HVAC systems were installed 10 years ago, with a range from 1 to 65 years.

HVAC systems are not a recent innovation in the Spanish hotel industry and the willingness to upgrade the EE of systems is low. Only 6% of respondents reported that they had upgraded to a more energy-efficient HVAC system in the last 5 years and 94% reported that they did not plan to upgrade their HVAC systems in the next 5 years. They stated that their current HVAC was working properly and assigned less importance to other reasons (e.g. building infrastructure problems, limits on access to loans, the current HVAC covers the needs of the establishment). This may be considered as a limitation in the analysis.

Energy efficiency is the attribute most frequently rated as important in choosing an HVAC system (Fig. 1). 67% of respondents rate it as a very important attribute. Other attributes less highly rated than energy efficiencyFootnote 5 are also classed as important in purchasing decisions: noise level (decibels) is ranked second (64% rate it as very important), followed by price (62%). Brand reliability (i.e. durability and technical and maintenance support) and performance (such as automatic control, temperatures adjustable according to outside temperature and humidity level, heat recovery and integrated heating and cooling functions) are also rated as very important by at least 50% of respondents.

Fig. 1
figure 1

Importance of the attributes of HVAC systems in purchasing decisions for establishments in Spain

Full size image

Barriers to energy-efficient HVAC systems

The three families of barriers identified in the literature were investigated in the survey and in the preliminary in-depth interviews (see Fig. 2) using fourteen statements.

Fig. 2
figure 2

Agreement with drivers and barriers for energy-efficient HVAC systems for hotel establishments in Spain. Note: The respondents indicated their agreement to several statements on a scale with the four ordered response categories “strongly agree”, “slightly agree”, “disagree” and “strongly disagree”

Full size image

Regarding market barriers, the vast majority of respondents were found to believe in the reliability of energy-efficient equipment: about 83% disagreed or strongly disagreed with the idea that more energy-efficient HVAC systems were less reliable. This is supported by Peruzzi et al. (2014), who reveal the importance of reliability for HVAC systems, especially for those systems that must guarantee uninterrupted service. 67% of respondents answered that lack of access to loans was an important barrier limiting their energy-efficient HVAC purchases. It is also important to consider that about 61% of respondents strongly or slightly believed that specific measures could make customers more aware of and more responsible regarding their energy consumption. As for behavioural barriers, 38% of respondents knew how much energy their equipment consumed, and 34% were aware of energy prices, but only 7.5% strongly agreed that they understand how much money they would save if they bought a more energy-efficient HVAC system. Concerning organisational barriers, about 39% of respondents strongly agreed that they were willing to take a chance on new technology to reduce their energy consumption and only 30% of respondents stated that they could not afford to buy a new, energy-efficient HVAC system. It is also important to highlight that about 43% of respondents strongly agreed that buying a HVAC system with more energy-efficient properties would reduce their environmental impact.

The role of energy efficiency labels

Several studies highlight the importance of labelling schemes in preventing informational failures and consequently addressing the EE gap (Carroll et al. 2016; Lucas and Galarraga 2015). We therefore also analyse the role of ecodesignFootnote 6 and energy labellingFootnote 7 regulations used in HVAC systems. According to Figs. 3 and 4, in the case of both heating and cooling systems only half the respondents (i.e. 100 respondents) acknowledged the existence of the energy label and/or technical specifications label. 70% of those respondents who were aware of the ecodesign and energy labelling of heating systems stated that these labels had influenced past purchasing decisions for heating systems (see Fig. 3), and 74% of those aware of ecodesign and energy labelling of cooling systems stated that these labels had influenced past purchasing decisions for such systems (see Fig. 4). In addition, 97% of the respondents agreed with the statement that their company considered energy efficiency labels when purchasing heating and cooling systems. That is, their answers to this question were not directly linked to past purchasing decisions being conditioned by the label. In fact, they may be reflecting future preferences too.

Fig. 3
figure 3

Awareness and influence of ecodesign and energy labelling regulations with 95% confidence intervals. “Influence” data refers to those respondents who are aware of both types of regulation

Full size image
Fig. 4
figure 4

Awareness and influence of ecodesign and energy labelling regulations with 95% confidence intervals. “Influence” data refers to those respondents who are aware of both types of regulation

Full size image

Several studies of the effectiveness of energy labelling suggest different ways of improving such labels, so it is important to understand how consumers use the information on the labels in their purchasing decisions (Heinzle and Wüstenhagen 2012; Stadelmann and Schubert 2018). In this regard, the responses in the study reported here indicate a high level of agreement about understanding of and trust in existing energy-efficiency labels.

This study also analyses the role of labels in HVAC systems with energy-cost information and incorporate those statements into the survey presented in ‘Barriers considered in the survey’. Ninety percent of hotel owners think that labels with additional monetary information are more understandable and trustworthy (85%) than existing EE labels, and 92% of respondents said that these labels would influence their purchasing decisions. In fact, most respondents believed that a label with additional monetary information would be more helpful in understanding how much energy was consumed by an HVAC system (89%) and calculating how much it cost to run (88%). Finally, 30% of respondents believed that there was a risk of labels potentially being manipulated by manufacturers.

Hotel owners’ attitudes towards the environment

Lastly, this study includes attitudes and beliefs about environmental issues. Eighty-six percent of respondents stated that they were concerned or extremely concerned about the environment. Habits for energy savings and investments in green and energy-efficient equipment were measured using a survey question validated by the OECD (2011). Regarding energy saving behaviour, we analyse how often establishments implement the following practices in their business: automatic control of HVAC systems or regular information to promote responsible consumption of energy and water by workers. Answers were classified into the following categories: “Never”, “Occasionally”, “Often” and “Always”. The responses indicate a high level of recognition of these practices. Specifically, 37.5% of respondents stated that they always automatically controlled the use of HVAC systems in rooms (e.g. by using smart key cards, smart thermostats or on/off programming). Seventy-one percent also answered that they always provided information on energy consumption to promote responsible energy consumption among workers.

In regard to other energy-efficiency investments, this study analyses whether establishments have invested or not in green and energy-efficient equipment, with the following response categories: “Yes”, “No”, “Already equipped, more than 10 years ago”, and “Not possible/feasible in my establishment”. It is important to note that 66% of respondents have energy-efficient appliances (e.g. minibars or TVs) installed. All respondents except one were also found to have invested in LED lighting. Seventy percent of establishments stated that they had invested in energy-efficient windows and thermal insulation of walls and roofs. Sixty percent said that they had sensors for controlling lights and temperature in common areas. Finally, respondents were asked to indicate whether they had invested in solar panels for electricity generation or for heating water. Such investments were found to be much less common, with only 23.5% of establishments equipped with this technology. According to Caird et al. (2008), this may be explained by barriers such as uncertainty as to the performance and reliability of the technology.

Factors influencing the importance assigned to EE as an attribute

The factors influencing the rating of EE as a very important attribute were explored in a probit model. The results are presented in Table 3. All categories of factors have an influence on the probability of the value EE very importantly.

Establishments in areas with a continental climate are 28% more likely to value EE as a very important attribute than those located in a Mediterranean climate. Managers of establishments in areas characterised by hot summers and cold winters are more interested in EE as it might reduce energy-related running costs of air conditioning and heating. The type of establishment also plays a significant role: hostels are 20% more likely to value EE as a very important attribute than hotels. This may be because small establishments are more concerned about energy bills, given that they have lower personnel costs and higher energy costs than larger hotels. Another reason may be that the level of insulation in hostels is lower. Financial soundness was found to be significantly correlated with occupancy rates during the high season. The latter variable was used to avoid collinearity in the regression to capture the income situation. Establishments with higher occupancy rates and thus higher energy consumption are more likely to value EE as a very important attribute. For example, those with an 80% occupancy rateFootnote 8 are 45% more likely to do so. The possible reason is that higher occupancy establishments may have higher energy costs and so they may recoup EE investments more rapidly. We find no evidence of barriers to access to capital in the rating of EE.

Some technical characteristics of HVAC systems have a significant impact on ratings. Establishments with HVAC systems that run on propane are 38% more likely to value EE as a very important attribute. In terms of energy price, a comparison of the energy sources considered in this study reveals that propane costs more than natural gas but less than heating oil and electricity for heating (EIA 2020). The higher price of propane may lead owners to use energy-efficient equipment so as to reduce their HVAC bills. Nevertheless, buildings in Spain are less likely to have propane-fired HVAC systems than buildings in other countries (e.g. the USA) (EIA 2011), which suggests that effects involving this technology should be interpreted with caution. On the other hand, establishments that use electricity are 24% less likely to rate EE as a very important attribute. This may be for two reasons: one is that establishments which use electricity are relatively unconcerned about EE and the other is that energy consumption in these establishments is lower than energy consumption in other establishments.

Attributes of HVAC systems such as price, brand reliability and performance are also important determinants of the decisions made by establishments. Specifically, respondents who consider price as a very important attribute are 19% more likely to value EE as a very important attribute. One interpretation of this positive relationship concerns the budget constraints of consumers. Consumers with a binding budget constraint rate price (namely low prices) and EE as important in reducing running costs. Energy-efficient goods are more expensive, so these consumers are less likely to buy such goods. This budget constraint explains EE campaigns with financial incentives to buy energy-efficient goods. Other studies show that energy-efficient equipment is more price-elastic than regular equipment (Coad et al. 2009; Galarraga et al. 2011). Other attributes, such as brand reliability and performance, also significantly affect the rating of EE, to a similar extent to pricesFootnote 9. Indeed, respondents who rate the brand reliability and performance of HVAC systems as a very important attribute are 24% and 27% more likely, respectively, to rate EE as a very important attribute. Brand reliability and performance are thus as important as price is in the rating of EE. These findings provide evidence that consumers prefer energy-efficient HVAC systems which also have good performance and good brand reliability. This could indicate that they may be considering EE as a proxy for quality, i.e. considering the ability of HVAC systems to fulfil a specific requirement.

The attitude towards specific barriers as regards EE also helps to explain why EE is rated as very important. Respondents who strongly agree with taking a chance on new technologies to reduce their energy consumption are 17% more likely to rate EE as a very important attribute. This result, combined with the intention to upgrade HVAC systems, however, seems to indicate a gap between beliefs and purchasing decisions due to the barriers to EE adoption reviewed above. Indeed, 40% of respondents indicated that they were willing to take a chance on new technologies to reduce their energy consumption, but only 6% reported that they planned to change their HVAC systems. Similar results are observed in household energy-efficiency choices. Damigos et al. (2020) find no evidence that this same belief increases the purchase of energy-efficient refrigerators.

In regard to the role of monetary information labels, we find that hotel owners who state that a label with additional monetary information would be more helpful than the current label are 26% more likely to rate EE as a very important attribute.

Environmental concerns are also a factor in explaining EE in the hotel industry. On average, owners more concerned about the environment are 15% more likely to rate EE as a very important attribute. This finding is consistent with those of other studies such as Damigos et al. (2020) and Shen (2008), who show that the importance of EE is positively affected by the proenvironmental behaviour of the respondents. Energy-saving habits positively influence the probability of rating EE as a very important attribute, as expected from other studies (Palm 2009). Indeed, establishments that always control the use of HVAC systems in rooms automatically (e.g. using smart key cards or on/off programming) are 18% more likely to rate EE as a very important attribute. This result supplements the existing literature on EE measures in the service sector, which finds that factors affecting the adoption of high-cost technologies such as HVAC systems also affect the adoption of low-cost measures (e.g. switching off lights whenever possible or managing and controlling energy use) (Schlomann and Schleich 2015).

Interestingly, we find that establishments with thermal insulation in walls and roofs are 19% less likely to rate EE as a very important attribute. This seems to indicate a negative feedback from nonHVAC-related EE investment in the rating of EE. Establishments which invest in insulation have lower heating and cooling consumption, so their need for energy-efficient HVAC systems, or the savings brought by them, is lower. This may lead them to underrate the EE attribute. Moreover, it would make sense for owners to scale down the importance of this attribute for the future once major action towards it has been taken. Another potential explanation is that establishments which undertake substantial environmental actions such as investing in insulation and energy-efficient windows feel less motivated to adopt further proenvironmental measures. They believe that their investment in thermal insulation of walls and roofs has already helped to mitigate climate change. This result supports previous findings on environmental concerns and energy mitigation behaviour (Nauges and Wheeler 2017; van der Linden 2017). For example, Nauges and Wheeler (2017) find that adoption of mitigation behaviour may have a negative effect on a household’s climate change concerns.

Conclusions

HVAC systems are major consumers of energy in the hotel industry, and reducing and decarbonising energy consumption on heating and cooling is crucial for the energy transition. EE provides an important pathway for reducing energy consumption, generating energy savings and reducing energy expenses. However, it is often observed that investments in EE are lower than they should be. The analysis reported here, based on a survey of the hotel industry, identifies factors that influence the EE choices of Spanish accommodation owners and contributes to the literature exploring the barriers to EE investment in that industry.

The main results of this study are that hotel accommodation owners rate EE highly in their purchasing decisions but that several barriers limit the importance attributed to EE and thus their investment in EE. Those factors are related to the market and to individual behavioural and organisational factors.

There is evidence of lack of information and bounded rationality, because only a third of all respondents know how much energy their equipment consumes and what the price of energy is. Results also show that the decision of whether to purchase HVAC seems to be affected by existing energy labels, and establishments believe that a label with additional monetary information would be more understandable and helpful in understanding how much energy is consumed by HVAC systems and in calculating how much HVAC would cost to run.

The market price of goods also influences how highly EE is rated by owners. Lower prices would help to increase the importance attributed to EE, and other attributes reflecting the quality of goods, such as brand reliability and performance in terms of services provided, also positively influence EE.

Organisational factors such as the importance attributed to new technologies and environmental concerns are also a factor in explaining EE in the hotel industry, although there is a gap between beliefs and purchasing decisions.

Climate considerations, information and the technical characteristics of establishments affect how highly the energy-efficiency attribute is rated. A negative feedback effect is detected from investment in other EE goods but not in HVAC-related equipment. Establishments with thermal insulation in walls and roofs are less likely to rate EE as a very important attribute. However, those that always control the use of HVAC systems in rooms automatically (e.g. by using smart key cards or on/off programming) are more likely to rate EE as a very important attribute.

This research has several policy implications. To design the right EE policy, one must account for the potential responses by agents, and the analysis reported here helps identify the drivers to which they may or may not respond. This is consistent with the findings of Blasch et al. (2018), who analyse the concept of “energy-related financial literacy” (which measures the level of energy-related knowledge and cognitive abilities that consumers need in order to take decisions with respect to investment for the production of energy services and their consumption). One of the points brought to light by our survey is the lack of knowledge among owners of hotel establishments about the energy and monetary savings provided by more energy-efficient equipment. This clearly indicates that it is of interest to use information-based policy instruments such as labels, energy audits and feedback on bills. This is reinforced by the fact that many owners do value the information provided by labels, and would like that information to be included.

This analysis also shows that different responses may be obtained in different climates: policies could be directed first at those areas, such as continental climates, where agents are more responsive to EE concerns. Also, in terms of directing policies, interaction with building retrofitting also needs to be accounted for, given that interest in EE HVACs decreases when buildings have been thermally insulated.

Energy taxes help reduce free-riding and rebound effects, and may also help accentuate pro-energy-efficiency attitudes among current owners of HVAC systems based on fossil fuels. Energy demand measured trough the occupancy rate is found to influence the rating of EE positively. A higher associated cost for energy would therefore reinforce the importance given to EE help owners reconsider their currently low willingness to upgrade HVAC systems, though subsidies may also be required to help overcome the bounded rationality problem. Subsidies may also help overcome the gap between beliefs and purchasing decisions.

Finally, the fact that EE is strongly linked with price, brand reliability and performance means that those who invest in cheaper, low-quality equipment seem less likely to rate EE highly. Here, the introduction of stricter, mandatory EE standards across the board would ensure that energy is saved even in such cases.

However, it is worth remarking once again that our study addresses potential barriers to the adoption of EE but is not able to measure how they translate into underinvestment in EE, given the low replacement rate in the sample. This would be of course a very welcome element for improving policies, and to determine the real consequences of these barriers. Further research is ongoing in this area.