1. Introduction
In December 2019, a hitherto unknown virus infection spread in Wuhan, China [
1] and caused severe pneumonia-related diseases. On 12 March 2020, the World Health Organization declared this “COVID-19 disease” a global pandemic. Hitherto, the COVID-19 pandemic has resulted in the loss of human lives as well as economic prosperity. In 2020, the disease ranked as a leading cause of death in the United States [
2]. Social regulations were necessary to save human lives.
These regulations, necessarily enforced to lower infection rates during the COVID-19 pandemic, led to numerous restrictions to peoples’ behavior in everyday life. In fact, during the “lockdown”, when people were instructed to stay at home whenever possible, their work and leisure habits changed. For example, travelling, shopping, and sauntering in cities and bars was restricted (e.g., [
3]). Many shops closed for a couple of weeks in most European countries during spring 2020. Thus, COVID-19 regulations not only encouraged “social distancing”, but also affected consumer behavior and the economy. Economic effects were first seen in China, where it is likely that SARS-CoV-19 first started to spread [
4]. China’s economic growth already reached a 30-year low during the first quarter of the year 2020 [
5] and the total per capita expenditure in China decreased by 12.5% nationwide compared to the first quarter in 2019 [
6]. Whereas the amount of money spent on food and residence increased, other expenses decreased, for example the expenditure on clothing, transportation, culture, and recreation [
6]. Similar or even worse economic impacts were expected for European countries. Indeed, by the end of April 2020, the consumption sentiment of private households reached an all-time low in Germany [
7].
During these various economic influences of the pandemic, environmental issues were not as prevalent. In Austria, for example, worries about environmental pollution as an issue people were “mostly concerned about” before the pandemic decreased by 21% during the period of COVID-19 restrictions in March and April 2020 [
8]. One possible reason for this is that the media reported on how decreasing economic trends had positive effects for the environment (e.g., [
9,
10]). For the first time in several years, fish were sighted in the canals of Venice; ozone gaps reduced in size; wild animals exploited new habitats [
11]; and the air became measurably cleaner [
12]. Germany announced that it would be able to reach the national climate goals in 2020 [
13]. Thus, nature seemed to recover during the lockdown. However, the short-term environmental benefits could be followed by increasing levels in carbon emissions, exceeding even the initial levels, and a long-term change for the worse. Moreover, human-made environmental impacts could rebound.
Why? As we know from previous crises, the decrease in carbon emissions can be short-lived. For example, after withstanding the global economic crisis in 2010, the highest total annual growth of carbon emissions was recorded [
14]. Another objective measure is the gross domestic product, which showed how the economy recovered while the purchasing power in society heavily increased post-crisis (e.g., [
15]). Indeed, in October 2020, the media already publicized the following title: “China’s economy continues to bounce back from virus slump” [
16].
As economic trends depend on human behavior, we assume that psychological variables matter. In concordance with patterns of consumer behavior, the importance of materialistic values might decrease during a moment of crisis but increase afterwards. That is, materialism might “rebound”. One possible psychological explanation for this trend is how consumption relates to customers’ benefit [
17]. Personal goals and benefits are not naturally limited since, once they are satisfied, they might rebound [
18]. Materialism values spur on personal materialistic goals belonging to different individuals, as the possession of material objects brings temporary happiness and conveys success and status [
19]. An accompanying effect is expected to be the decrease in pro-environmental behavior post-crisis. Indeed, people sacrificed a lot during the lockdown period and (incidentally) this benefited the environment. This supposed “buffer” could then be made up as soon as the lockdown is over. In addition, increased consumption and materialism values contradict environmentalism [
19].
In the present study, we tested these assumptions by comparing pro-environmental behavior and materialism values before (T1), during (T2), and after (T3) COVID-19 lockdown restrictions from April to July 2020. Due to the unpredictability of the COVID-19 lockdown restrictions, T1 measurements (concerning “a typical week before the lockdown’s onset”) were assessed retrospectively.
In the field of economics, the Jevons paradox [
20] describes how energy consumption increases while technological progression enhances energy efficiency. This paradox was called the rebound effect. Khazzoom [
21] questioned whether energy efficiency truly lowered the net energy demand. For example, even though the energy used to heat a room or light a bulb was significantly reduced, people did not use less energy in total. One explanation for this is the consumption of gained reserves in the manner of even more units, due to more frequent or more intense usage [
22]. Indeed, Greening, Greene, and Difiglio [
23] summarized the empirical evidence for the existence of a low to moderate rebound effect of energy use. Specifically, there are three different types of rebound effects (e.g., [
24]). First, a weak rebound effect means that efficiency measures do not work to save energy. Second, the definition of a strong rebound effect is that savings are spent on even more frequent or intense consumption. The reduction of costs justifies further expenses in a row. As a third type, a backfire effect means that consumption will even exceed its initial level after savings were made.
Regarding the occurrence of rebound and backfire effects, psychological variables are crucial ([
22,
25]). Therefore, we assume that rebound effects are not limited to energy efficiency. From a psychological perspective, it is known that people are mentally bookkeeping financial costs [
26] but also various other values and behaviors. For example, Thøgersen and Crompton [
27] speak of negative spillover effects in environmental behavior. They argue that lowering one’s personal carbon footprint might lower the probability to show even more pro-environmental behaviors in return. An accompanying finding is the idea of moral licensing. Previous research (although yielding inconclusive results) investigated the idea that pro-environmental behavior, as a form of moral behavior, decreases the likelihood of subsequent pro-environmental behavior or, more generally, triggers immoral behavior [
28,
29].
It is likely that mental accounting has an environmental impact. For example, Girod and de Haan [
30] proposed the concept of mental rebound. “This concept postulates that households apply mental accounting (bookkeeping) for the environmental impacts caused” [
30]. As an example, the authors found higher amounts of flights per year for households without car availability (controlling for income). They assumed that households that do not own a car have a higher mental environmental budget. In turn, this budget allows them to fly more often. Thus, in the context of COVID-19, after people saved on their carbon footprint during a lockdown, they could spend the resulting subjective “buffer” in the way of increased consumption afterwards.
In previous research, rebound effects are limited to behaviors. However, we postulate that even a temporary change of values could accompany behavior changes. People might reassess their values under specific circumstances. For example, materialistic values fade into the background during a deadly global pandemic. Rather than pursuing materialistic values, one’s own health, and that of one’s beloved ones, job-security and relationships might then be more important. In return, the desire for materialistic happiness could rise again after the lockdown has ended and the situation improved. Therefore, the present study investigated rebound effects in terms of pro-environmental behavior and accompanying materialism values by comparing times before, during, and after the COVID-19-lockdown.
As a secondary goal, we were interested in whether some people were more likely than others to exhibit rebound effects in terms of pro-environmental behavior and a change of materialism values. There are numerous variables influencing pro-environmental behavior, among which pro-environmental attitude is one of the most important positive predictors [
31,
32]. We assumed pro-environmental attitude to moderate rebound effects in the sense of a diminishing effect. Thus, a lockdown is expected to affect participant’s pro-environmental behavior to a lesser extent when they endorse a strong pro-environmental attitude. In line with pro-environmental values, pursuing materialistic wishes are expected to be less important, even after withstanding the lockdown.
Further, we considered openness to experiences, honesty-humility, extraversion, conscientiousness, and agreeableness as possible moderation variables. According to the HEXACO model of personality [
33], honesty-humility is characterized by being sincere, fair, and unassuming [
34]. Extraversion describes talkative and highly energetic people. Agreeableness is linked to attributes such as being good-natured, cooperative, and trustful. Conscientiousness is characterized by being orderly, responsible, and dependable. Openness goes along with being intellectual, imaginative, and independent-minded. These personality traits are known to positively influence pro-environmental behavior [
34,
35]. Therefore, participants who are scoring high on these personality scales are expected to vary less in their pro-environmental behavior due to the lockdown restrictions. As a result, diminishing effects of honesty-humility, extraversion, agreeableness, conscientiousness, and openness to experiences on the extent to which rebound effects occur were assumed.
In contrast, narcissism is expected to enhance rebound effects of pro-environmental behavior and accompanying materialism values. We considered narcissism as a personality trait (instead of a clinical disorder), which can be described as a pattern of grandiosity, self-importance, and self-focus [
36]. Narcissism is known to diminish altruistic daily pro-environmental behavior [
37]. Moreover, given that narcissists do not value environmental issues in general [
38] and there are positive relations between narcissism, materialism, and compulsive buying [
39], we expected that after the restrictions of a lockdown, especially narcissistic participants show an increase of materialistic values and care even less about environmental issues.
2. Materials and Methods
Hypotheses. Assuming that rebound effects are not limited to behavior, and values could also be temporarily adapted to specific circumstances, our study investigated rebound effects in pro-environmental behavior and accompanying materialism values comparing the times before (retrospectively), during and after lockdown-regulations during the COVID-19 pandemic.
Figure 1 displays the materials and methods used in this study.
- (1)
Compared to the time before the lockdown, we expected people to report less pro-environmental behavior after the lockdown (1.1) and higher values of materialism after the lockdown (1.2).
- (2)
Moderating effects of pro-environmental attitude, personality (Hexaco, Ashton, and Lee, 2009) and narcissism (NPI; Ames, Rose, and Anderson, 2006) were expected concerning the relations specified in the first hypothesis (pre vs. post). In particular, we expected pro-environmental attitude to have a diminishing effect on the change of pro-environmental behaviors and materialism values from pre to post lockdown restrictions (2.1). We expected openness to experiences, honesty-humility, extraversion, conscientiousness, and agreeableness to have a diminishing effect on the change of pro-environmental behaviors and materialism values from pre to post lockdown restrictions (2.2). In addition, we expected narcissism to have an enhancing effect on the change of pro-environmental behaviors and materialism values from pre to post-lockdown restrictions (2.3).
Participants. During the time between 8 April and 20 April 2020 (T1), 654 participants took part in an online study, which we announced via newsletters and social media (e.g., on Facebook). The study was conducted using the online platform SosciSurvey [
40]. On June 23, we contacted the same participants again. Data were collected until 25 July 2020. Overall, 378 participants filled out the survey both times (42.20% dropout). As a reward, these participants could take part in a raffle to win one of five 50 Euro cash prizes. Eight participants had been excluded due to failed attention checks. Among the remaining sample of 370 participants, there were 243 women (65.68%), 126 men (34.05%), and 1 participant stating a diverse gender. On average, participants were 37 years old (
SD = 16.28). Three-hundred-and-twenty-five (87.84%) participants lived in Germany, thirty-eight (10.27%) in Austria, and seven (1.89%) in Italy (mainly South-Tyrol). Lockdown restrictions were similar at the time of the study in these countries. Restrictions included closed shops, closed cultural and sport institutions, and prohibited sauntering in cities and bars during the first ever experienced COVID-19-lockdown in April 2020. At the second period of measurement of this study, restrictions on daily life were significantly loosened (in June until the end of July 2020). By this time, all shops re-opened, and one was allowed to meet friends and spend time outdoors again (especially in Austria, where this was forbidden for some weeks). In Germany, mouth and nose masks still had to be worn in shops, but people met for sporting activities and in cities.
Around 38% of participants (N = 141) were employed and 28% were students at a University (N = 105). Other occupations were self-employment (5.95%, N = 22) and retirement (8.12%, N = 30). As a consequence of the COVID-19 pandemic, the occupation of 140 participants (37.83%) changed (e.g., by cut hours or short-time work). Fifteen participants even lost their job. The monthly financial net income ranged from 250€ to 1500€ in 146 cases (39.50%), from 1500€ to 3000€ in 112 cases (30.03%), and exceeded 3000€ in 63 cases (17.03%). A few participants were without income (8.11%, N = 30), and others did not want to answer this question (3.24%). During the time of the lockdown (T2), 29.46% of participants (N = 109) were officially or voluntary self-quarantined.
Dependent variables. As just noted, the survey was conducted at two points of time. During COVID-19-lockdown-regulations in April 2020, we retrospectively assessed people’s pro-environmental behavior and materialism values concerning a typical week before the pandemic offset (T1). At the same time, the dependent variables during the COVID-19-lockdown (T2) were assessed. At the end of the first survey, participants registered their e-mail address so that we could contact them after lockdown-restrictions were loosened. After strict restrictions on daily life were significantly loosened (in June until the end of July 2020), we assessed the dependent variables again (T3).
To assess pro-environmental behavior, two types of behavior were considered. On the one hand, a 12-item scale asked for the frequency of ecologically compatible actions [
41]. Answers were given on a seven-point scale (1 “never”, 4 “sometimes”, 7 “always”). Ecologically compatible actions contain daily behaviors, such as buying ecologically grown vegetables or the use of biodegradable detergents. On the other hand, the tendency to abandon certain behaviors for the sake of the environment was assessed (pro-environmental sacrificing actions; [
42]). A sample item is: “I have saved water in household for environmental reasons (e.g., short showering instead of bathing etc.)”. Answers were given using a six-point scale (1 “not at all”, 6 “entirely”). To reduce social desirability effects, some items were inversed. We defined concerning periods for both types of pro-environmental behaviors by “a typical week before the lockdown restrictions due to the COVID-pandemic” (T1), currently, which meant the time during these restrictions (T2) as well as a typical week after the restrictions were loosened (T3).
As a second key dependent variable, materialism values were assessed. To this end, we adapted the materialism questionnaire of Richins and Dawson [
43] in order to measure state instead of trait values by adding time specifications. Other authors (e.g., [
44]) previously chose this method. The periods specified were identical to the assessment of pro-environmental behavior (T1 before, T2 during, T3 after the lockdown). One of the eighteen items, which was answered by using a five-point scale (1 “fully disagree”, 2 “rather disagree”, 2 “undecided”. 4 “rather agree”, 5 “fully agree”), is: “Buying things gives me lots of pleasure”.
Potential moderating variables. Environmental attitude was assessed at the first point of measure, using the 12-items scale of a validated questionnaire [
45]. The items asked for the participant’s agreement by a four-point Likert scale (1 “not at all”, 4 “entirely”). One sample item is: “Growth has natural limits that are already reached in our industrialized world”.
In the second part of the study, participants stated their current pro-environmental behavior and materialism values again. Afterwards, the HEXACO-PI-60 questionnaire [
46] was employed to assess basic personality traits. Participants stated their agreement using a five-point scale (1 “not agree at all”, 5 “fully agree”). A sample item, corresponding openness to experiences (reversed), is: “I would be quite bored by a visit to an art gallery”. Narcissism was also assessed in the second survey, using the NPI-16 questionnaire [
47]. In a forced choice format, participants had to make a decision between two counterparts in 16 cases. An example is the decision between the following statements: “I don’t want to stand out in the crowd” vs. “I love to be the center of attention”, whereas the latter statement represents the narcissism pole. Answers were coded by giving one point for each narcissism pole selected.
3. Results
Data Analysis. The reliability of scales was investigated by Cronbach’s alpha and 95% confidence intervals (
Table 1). Following Koning and Franses [
48], providing confidence intervals for Cronbach’s Alpha can help to test for significant differences for alpha values across different studies. Some scale reliabilities were relatively poor. This issue will be later discussed. Further, Shapiro–Wilk tests indicated a normal distribution for only two scales, namely ecologically compatible actions during and after the lockdown, whereas normal distribution was violated for all other scales. Hence, we relied on non-parametric spearman-rho rank correlations (
Table 2).
Test of possible rebound effects. Our main hypotheses (‘compared to the time before the lockdown, people state less pro-environmental behavior (1.1) and higher values of materialism afterwards (1.2’) were investigated by a Wilcoxon signed-rank test for paired samples. With regard to ecologically compatible actions, the difference before (T1, M = 4.66, SD = 0.79), compared to after the lockdown (T3, M = 4.62, SD = 0.83) was not significant (p = 0.315, r = −0.05). Concerning pro-environmental sacrificing actions, the difference when comparing T1 (M = 4.66, SD = 0.84) to T3 (M = 4.47, SD = 0.91) yielded a significant difference (p < 0.001, r = −0.23). As hypothesized, participants reported less pro-environmental behavior at T2. In terms of materialism, the comparison of T1 (M = 2.43, SD = 0.53) and T3 (M = 2.26, SD = 0.54) yielded a significant difference as well, p < 0.001, r = −0.43. However, contrary to our hypothesis, the level of materialism values was still lower after the lockdown-restrictions, compared to a typical week before the restriction’s onset
Further differences were tested on an exploratory basis. For the time before the lockdown (T1) ecologically compatible actions were significantly higher than during the lockdown (T2,
M = 4.40,
SD = 0.92),
p < 0.001,
r = −0.47. Comparing T2 with T3, ecologically compatible actions significantly increased again,
p < 0.001,
r = −0.35. sacrificing actions did not change from T1 to T2 (
M = 4.61,
SD = 0.91),
p = 0.303,
r = −0.05, but they decreased from T2 to T3,
p < 0.001,
r = −0.18. Considering materialism, values significantly changed from T1 to T2 (
M = 2.11,
SD = 0.54),
p < 0.001,
r = −0.66, as well as from T2 to T3,
p < 0.001,
r = 0.28. All means are illustrated in
Figure 2 and
Figure 3.
Moderation analyses. To test the potential moderating effects of pro-environmental attitude, personality, and narcissism (hypotheses 2.1–2.3), an analysis of variance with repeated measures was conducted for each dependent variable. Times of measurement (T1 before, T2 during, and T3 after the lockdown) were considered in a repeated measurement design. In a first step, zero-order models without moderation effects were interpreted. Afterwards, the following covariates were added: the HEXACO scales honesty humility, extraversion, emotionality, agreeableness, conscientiousness and openness, the NPI score (narcissism), and pro-environmental attitude.
Zero-order models revealed a significant effect of time (T1 vs. T2 vs. T3) for pro-environmental sacrificing actions,
F(2, 368) = 45.90,
p < 0.001,
ƞ2 = 0.20, for ecologically compatible actions,
F(2, 368) = 10.17,
p < 0.001,
ƞ2 = 0.05, as well as for materialism values,
F(2, 368) = 113.71,
p < 0.001,
ƞ2 = 0.38 (see
Figure 2 and
Figure 3).
In a next step, we added variables that were assumed to moderate the findings. For pro-environmental sacrificing actions as a dependent variable, no significant effects were found (Openness: p = 0.121, ƞ2 = 0.012; Conscientiousness: p = 0.363, ƞ2 = 0.005; Agreeableness: p = 0.068, ƞ2 = 0.015; Extraversion: p = 0.382, ƞ2 = 0.005; Emotionality: p = 0.219, ƞ2 = 0.008; Honesty-Humility: p = 0.251, ƞ2 = 0.008; NPI: p = 0.139, ƞ2 = 0.011; environmental attitude: p = 0.326; ƞ2 = 0.006).
Considering ecologically compatible actions as a dependent variable, no effects were found either (Openness: p = 0.177, ƞ2 = 0.010; Conscientiousness: p = 0.119, ƞ2 = 0.012; Agreeableness: p = 0.089, ƞ2 = 0.013; Extraversion: p = 0.711, ƞ2 = 0.002; Emotionality: p = 0.625, ƞ2 = 0.003; Honesty-Humility: p = 0.059, ƞ2 = 0.016; NPI: p = 0.584, ƞ2 = 0.003; environmental attitude: p = 0.222; ƞ2 = 0.008).
Lastly, materialism values were considered as a dependent variable. Three variables showed a significant interaction effect with time: Agreeableness (
F(2, 360) = 5.60,
p = 0.004,
ƞ2 = 0.030), NPI (
F(2, 360) = 3.27,
p = 0.041,
ƞ2 = 0.018), and pro-environmental attitude (
F(2, 360) = 5.25,
p = 0.006,
ƞ2 = 0.028). A multiple linear regression model revealed no effect of agreeableness on materialism values before the COVID-19 lockdown (T1:
β = −0.09,
t(368) = −1.67,
p = 0.095) and a significant negative effect meanwhile (T2:
β = −0.10,
t(368) = −1.97,
p < 0.05). After loosened restrictions (T3), agreeableness had no effect on materialism values again (
β = −0.02,
t(368) = −0.45,
p = 0.657). Narcissism traits enhanced materialism values at the first time of measure (T1:
β = 0.03,
t(368) = 2.63,
p = 0.009), as well as during (T2:
β = 0.02,
t(368) = 2.33,
p = 0.021) and after the lockdown (T3:
β = 0.04,
t(368) = 3.62,
p < 0.001). Thus, results showed the strongest enhancing effect of narcissism traits on materialism values after the lockdown. Pro-environmental attitude negatively influenced materialism values before (T1:
β = −0.54,
t(368) = −7.22,
p < 0.001) as well as during (T2:
β = −0.69,
t(368) = −9.51,
p < 0.001) and after (T3:
β = −0.57,
t(368) = −7.54,
p < 0.001) the lockdown. The decreasing effect of pro-environmental attitude on materialism values was strongest during the lockdown. In order to display the effects in
Figure 4,
Figure 5 and
Figure 6, we dichotomized values of narcissism, agreeableness and pro-environmental attitude in groups of “low” (values ≤ mean minus one standard deviation) and “high” (values ≥ mean plus one standard deviation).
4. Discussion
In the present study, we tested whether pro-environmental behaviors (pro-environmental sacrificing actions and ecologically compatible actions) and materialism values would change throughout a “lockdown” (before vs. during vs. after) due to the COVID-19 pandemic from April until the end of July 2020. Based on previous research on rebound effects, we assumed that pro-environmental behaviors would decrease after the lockdown compared to the time before. Indeed, pro-environmental sacrificing actions (e.g., short showering instead of bathing) decreased throughout the study. However, ecologically compatible actions (e.g., the use of biodegradable detergents) also decreased during the lockdown, but increased on its initial level again after restrictions were loosened. Thus, proactive environmentally friendly behaviors, on the one hand, can be differentiated from the tendency to abandon certain behaviors for the sake of the environment, on the other hand. While ecologically compatible actions resumed post-lockdown, environmental sacrificing actions declined continuously. We assume that participants showed this persistent decrease of environmental sacrificing actions, as they additionally had to sacrifice in many other areas of life during that time (for example in terms of work, leisure, and travelling; [
3]). Thus, only habits of ecologically compatible actions “recovered”.
Contradicting hypothesis 1.2, materialism values did not “fire back” after the lockdown. To the contrary, during the lockdown, materialism values decreased and stayed below the initial level post-lockdown. Feelings of insecurity of the social and economic situation might have still been present to this point (in July 2020) and further inhibited materialism values even after restrictions were loosened.
Concerning moderation effects, and contradicting our expectations, personality and environmental attitude variables did not moderate changes of pro-environmental behaviors, but moderation effects on materialism values were found. In line with our reasoning, pro-environmental attitude had a diminishing effect on materialism values, especially during the lockdown-restrictions. Thus, caring about the environment in general went along with lower materialistic values, strongest during the lockdown. Further, agreeableness also had a diminishing effect on materialism values during the lockdown, but not before or after these restrictions. As agreeableness goes along with being good-natured, cooperative, and trustful [
34], this trait might promote acceptance of social regulations, which might more shift the focus away from material desires and towards immaterial concerns during a global pandemic.
In contrast, narcissism traits enhanced materialism values, which were strongest after withstanding the lockdown. Thus, narcissists especially showed a tendency of rebound effects in materialism values after the lockdown-restrictions. This is in line with our reasoning and with previous findings on the link between narcissism, materialism, and compulsive buying. As narcissism goes along with strengthened materialistic desires, as well as with compulsive buying [
39], narcissistic people might be more prone to the expected rebound effects after the COVID-19 lockdown.
Implications. Our research has important theoretical and practical implications. For the first time, the influence of a “lockdown” on pro-environmental behavior and materialism values was investigated. Although findings of previous research on rebound, negative spillover effects and mental accounting (e.g., [
27]) seemed to be adaptable in the context of a COVID-19-lockdown, our findings did not support this assumption. However, the COVID-19-pandemic and this first-ever lockdown, in spring 2020, was an exceptional situation. It had severe psychological impacts (e.g., increased rates of depression, insomnia, and anxiety [
49]) and might therefore not be comparable to previous settings when rebound effects of pro-environmental behavior were observed (e.g., [
30]). From an evolutionary psychology standpoint, it makes sense that people only seek status and recognition when basic needs for safety and health have been met. Materialistic values, spurred by envy and jealousy [
50], are “an evolutionary adaptation that motivates individuals to take actions in order to improve their material or social status” [
51], p. 214. A pandemic, as COVID-19, might have such a considerable impact on a person’s basic needs so that a further improvement of one’s materialistic status is postponed.
It is important to keep in mind that our study has several limitations (see below) that may preclude strong conclusions, but what could be inferred from the present contribution is that previous theoretical frameworks on rebound effects and mental accounting did not receive support from our data. Further empirical investigations would be welcome that test the appropriateness of these theories to account for psychological effects of a COVID-19-lockdown. The present results show that this is an important issue: in fact, materialism values were adapted from the time before, during, and after lockdown, coupled with changes in environmental behavior. However, these changes might follow another pattern than has been assumed by classical rebound theories of previous research.
With regard to practical implications, the COVID-19-pandemic was previously said to have beneficial effects for the environment (e.g., [
10,
13]). The present study provides mixed results concerning this assumption. While environmental sacrificing actions persistently decreased, habits of environmental compatible actions surprisingly resumed after the lockdown. The latter trend might be beneficial in light of the looming climate crisis. In addition, the decrease of materialism values throughout our observation might foster economic declines but are superficially good for environmental reasons. However, it is not clear if this decline of materialism values through the COVID-19 restrictions is long lasting, as the present study conducted a period of four months only. An important practical conclusion is that we cannot rely on temporary changes for the better when it comes to pro environmental behavior. Rather, people seem to be returning to old patterns and habits even after a lengthy exceptional situation such as the COVID-19 lockdown restrictions were.
Limitations and future research. One limitation of the present study is the retrospective assessment of dependent variables for the time before the lockdown occurred. However, due to the unpredictability of the situation during the COVID-19-pandemic, we had to rely on memorized experiences. In addition, the present results rely on self-reports, which is a typical practice in the research of pro-environmental behavior, but might be problematic in terms of validity (e.g., due to the effects of social desirability) [
52,
53].
Additionally, especially the scale to observe pro-environmental sacrificing actions in this study yielded low rates of scale consistency, although the developers of the scale reported high scale reliability [
42]. As the scale was developed to ask for a two-month period, our adaption to a period of one week could have led to diminished scale reliability.
Based on the present study, which showed narcissism to enhance rebound effects of materialism values, future research could further investigate the role of narcissism in rebound effects of materialism values in particular. Further, objective measures of pro-environmental behavior (e.g., by the actual use of resources) in addition to the measured subjective psychological variables could improve validity in future research. Lastly, long-term effects of the COVID-19 pandemic on environmental behavior and materialism values would be worth to investigate. In addition, at the time of the present study, the COVID-19 pandemic is not yet over. Thereby, comparisons of effects in previous crises, such as the global economic crisis in 2008, can be insufficient. The present results are limited for the time when the first COVID-19-lockdown occurred in April 2020 until restrictions were loosened by the end of June until July 2020 (in Austria and Germany). Future research could further investigate long-term effects of the pandemic.
In sum, this research does not provide support for the occurrence of spillover or rebound effects of pro-environmental behavior and materialism values due to COVID-19 lockdown-restrictions. Moreover, materialism values, pro-environmental sacrificing actions, and ecologically compatible actions decreased during the lockdown. Materialism did not ‘fire back’ and ecologically compatible actions even resumed post-lockdown. In conclusion, not only daily social life of people, but also materialism values and pro-environmental behaviors were “locked down” due to COVID-19 restrictions.