The Impact of COVID-19 to Millennials' Consumption Behaviour and Halal Preference: Does Religiosity Matter?

- COVID-19 and the threat of economic crisis may change consumption behaviour and halal preference. Millennials in the surrounding cities of the capital in Indonesia currently contributes up to 28 percent from the total population. This study aims to analyse the influence of COVID-19 pandemic to the consumer behaviour and halal preference of the Millennials group in Jakarta, Bogor, Depok, Tangerang, and Bekasi. Theory of Planned Behaviour is modified and implemented in preparing the frameworks of the dependent-independent variables. Using partial least square method, 316 Muslim Millennial respondents from Jakarta and the surrounding cities as areas which reported first cases of COVID-19 in Indonesia. The result shows that crisis magnitude, fear of missing out, and social media have significant influence on consumer behaviour. Religiosity is, however also found to moderate the effect of crisis magnitude to consumer behaviour. On the other side, halal preference is found to be influenced only by religiosity.


INTRODUCTION
Pandemic which started at the end of 2019 had resulted in changes to several aspects, including economy and social. Restrictions in terms of time and space to conduct activities including economic activities has affected national and individual personal income. It is obvious that this pandemic has brought economic crisis. Badan Pusat Statistik (BPS) reported that Indonesia's economic growth in the second quarter of 2020 was minus 5.32 percent. Previously, in the first quarter of 2020, BPS reported that Indonesia's economic growth only grew by 2.97 percent, down far from growth of 5.02 percent in the same period in 2019 (Badan Pusat Statistik, 2020).
Crises create feelings of uncertainty and anxiety, thus perceptions of crises and beliefs about potential future impacts can lead to changes in consumer behaviour (Einarsdottir, 2016). Among the factors determining consumption is real income and expectation of the future economic condition. Some people, especially those affected by this condition may change their lifestyle or spending pattern. Urbonavicius and Pikturniene (2010) explain that in responding to crises, some consumers may not change their consumption at all, some others may drastically reduce spending to survive, and or increase savings as protection against potential losses. The severity of the crisis could be then seen from how many of the population who decrease their daily consumption. The decrease in consumption may happen voluntarily or by design because of limited in movement such as decreasing demand for the tourism service (Wong et al., 2022). Experience in certain products, services, and events can have a major impact on fear of consumption in times of crisis (Argan & Argan, 2018). Some researchers distinguish between necessities and extras (McKenzie, 2003). And others distinguish between "durable items, such as furniture, semi-durable items, such as clothing and food" (Zurawicki & Braidot, 2005). According to elasticity concept, those categorized as necessities like clothing and food are considered as inelastic commodities. There might be changes in the demand because of the changes in income, but it will not be significant because those are basic needs that must be fulfilled.
Millennials are the first generation to be widely exposed to the use of the Internet (Murphy, 2018). They are typically defined as those who were born between the 1980 up until 2000 (Rauch, 2018). Currently, the composition of Millennials ranges between 26 percent to 36 percent from the total population of Jakarta and the surrounding cities, as the area with high internet connection and online economic activity (Badan Pusat Statistik, 2021).
Due to lockdown, partially or totally, the movement of people and commodities were restricted. Since then, online activities are increasing especially because economic transaction, educational, and social activities are now mostly conducted virtually. This has also potential to change consumption preference, including consuming halal product.
Normally, Muslims depend on halal certification to ensure whether foods or drinks that they consume are halal and safe. For some commodities which are not halal certified yet, such as street vendors and homemade products, people can make some inquiries either directly to the producers or the sellers. Online purchase has somehow restricted this inquiry since food delivery are usually mediated by a third party, which is the delivery service provider. Halal inquiry is then limited. Kowalczyk et al. (2020) examines whether COVID-19 enhances faith and religiosity. This study conclude that this crisis strengthens the religiosity as one of methods to survive.
According to United Overseas Bank (UOB) Ltd, from income levels of various population segments between 2010 and 2019, the real income of Millennials grew by 8.6 percent per year on a compound annual growth rate. This growth rate is higher than income growth of other demographic groups between 3 and 5 percent (Ryo, 2019). Therefore, the spending behaviour is expected to be the highest compared to other range of ages. This study aims to see the relationship between crisis and Millennials' consumption behaviour and halal preference, and whether religiosity has significant effect on that matter.

LITERATURE REVIEW Theory of Planned Behaviour, Crisis, Consumer Behaviour, and Halal Preference
Theory of Planned Behaviour (TPB) is used to explain changes in consumer behaviour. There are three basic factors used in TPB, namely attitude, subjective norms and perceived behavioural control (Ajzen, 1991) which have an impact on purchase intention or behaviour.
Theory of Planned Behaviour (Ajzen, 1991) is originally used to explain the determinants of consumer purchase intention and behaviour. In conditions of natural disasters or epidemics, Daellenbach et al. (2018) also used this model as a basic theory for understanding consumer purchase intentions at uncertain times. This theory could also be used to explain the characteristics of purchasing halal food products.
As Muslims migrate to Western countries, the needs of halal foods start to arise and increase in those Muslim minority populated countries. Halal started to influence the global meat trade networks and the food supply chain (Soon & Wallace, 2017). One example explained earlier explains how migration by Muslims can change global consumption trends. This is an example of the application of the theory of planned behaviour, which specifically explains the intensity of halal food consumption behaviour (Ajzen & Fishbein, 2000). Previous research using theory of planned behaviour has concluded that behavioural control and subjective norm have a significant influence on halal food selection decisions (Alam and Sayuti, 2011). Lada et al. (2009) also explained that subjective norms have a relatively higher influence when tested in multicultural communities.
Despite having conclusions that are aligned, none of these studies have focused on the object of consumers in the younger generation (Khalek, 2014). In addition, there are conflicting results from previous findings, where subjective norms actually do not have a significant effect on the consumption of halal food products (Liou & Contento, 2001). Basev (2014) explores how the economic crisis affects food purchases on consumer behaviour in the UK. The results of this study indicate that there is no total neglect in overall shopping habits. The research of Zurawicki and Braidot in 2005 aimed to examine the responses of Argentina middle class consumers to the 2001-2002 economic crisis. Using SEM, the results show that high-income households reduce and eliminate their expenditure to a lower level than middleclass families. Koos et al. (2017) also found significant relationship between economic crisis and consumption.

Crisis, Fear of Missing Out, Social Media and Consumer Behaviour
In the context of responding to the sense of crisis, Good and Hyman (2020) mention that fear of missing out (FOMO) has become a widely recognized phenomenon recently. FOMO is considered as one of determinants of consumer behaviour. This could be a personal trait but may also induced by social factor such as family member, friends, or economic condition. Abel et al. (2016) studies on how FOMO influences decision making and behaviour. They found significant relationship between variables. Results suggest significant differences in social media consumption across levels of FOMO. Elhai et al. (2016) investigates on how FOMO, depression, and anxiety influence consumer behaviour. The study focuses on smartphone users. They found that there is a significant relationship between the exogenous variables and the frequency of consumption.
Social media can have a big impact on challenges to consumption crisis (Abel et al., 2016). Some people show a strong tendency and willingness to change their behaviour to follow and imitate a group, reflecting a desire that is inseparable from the mainstream and must be the same as others (Kang et al., 2019). One is the desire to have in someone who experiences the fear of losing a strong need between persons as a fundamental person's motivation, such as the need to identify with their status and recognition of others (Przybylski et al, 2013).

Religiosity and Consumption Behaviour
Islamic consumption requires cultural values to shape one's values, attitudes, intentions and behaviour, especially in strong beliefs and religions (Abutaleb et al, 2020). The impact of religiosity on consumer behaviour is also studied by Felix and Braunsberger (2016), Hassan (2011), Ramasamy et al. (2010), and Rice (2006). Religion is one important aspect in determining consumption patterns especially Muslims (Essoo & Dibb, 2004). Religion becomes one of the factors that can control the goals, decisions, satisfaction and motivation of individuals (Zimbardo & Ruch, 1975). While religion also influence individuals to avoid things that are not desired by religious guidance (Masnono, 2005). Religiosity is very personal so therefore it has different levels (de Run et al., 2010). Religiosity is one of the determinants of halal products consumption (Mukhtar & Butt, 2012;Weaver & Agle, 2002). Several findings from previous studies show that religiosity is one of the variables that influences alternative choices of consumers for consumption. While the findings from Delener (1994) show that the exact role in consumer decisions to choose food products is still uncertain, findings from Awan et al. (2015) show that religiosity has the lowest effect compared to marketing and awareness. Ali et al. (2021) investigates the effect of COVID-19 pandemic on the environmental awareness, sustainable consumption, and social responsibility. All dependent variables are found to be positively influenced by COVID-19. They also found that religiosity moderate the relationship between variables. Kowalczyk et al. (2020) examine whether COVID-19 enhances faith and religiosity and conclude that this crisis strengthens the religiosity as one of methods to survive. Sabir et al. (2020) analyses the effect of demographic factors on herding behaviour from the investors in Pakistan stock exchange. They conclude that religiosity moderates age and income in influencing the herding behaviour. Razzaq et al. (2018) conclude that religiosity moderates the positive relationship between fashion involvement and pro-environmental attitude and sustainable clothing consumption. Religiosity enhances the influence from the exogenous variables toward the endogenous variable.

METHODOLOGY Data
This study uses primary data collected through a survey of respondents who are Muslim Millennials in Jakarta, and cities surrounding the country's capital: Bogor, Depok, Tangerang, Bekasi. These cities were selected as the first places where COVID -19 cases were reported. The total sample size of 316 is expected to meet the adequacy of the data for statistical testing in SEM modelling.

Method
This study applied Partial Least Square-Structural Equation Modelling (PLS-SEM). This analysis is chosen to see the dynamics of consumption behaviour and the dependent variables given the COVID-19 Pandemic. The model specification is as follow: The hypotheses are as follow: H1: Crisis has significant influence on consumer behaviour. H2: Fear of Missing Out has significant influence on consumer behaviour. H3: Social Media has significant influence on consumer behaviour. H4: Religiosity moderates relationship between crisis magnitude and consumer behaviour. H5: Crisis has significant influence on halal preference. H6: Fear of Missing out has significant influence on halal preference. H7: Social Media has significant influence on halal preference. H8: Religiosity has significant influence on halal preference.  After the COVID-19 Pandemic, I reduced the consumption of secondary goods CB 3 After the COVID-19 Pandemic, I reduced the consumption of tertiary/luxury goods The questionnaire is in the form of Likert 6 scale statements. 1 means strongly disagree and 6 means strongly agree.
Over 52% of respondents are females and 54% are over thirty years old but within the range of Millennials. Most of respondents reported an income reduction of 65%; 25 % responded that there were no changes in income, and 10% even reported an increase in income. This is supported by the next questions inquiring the income before and after pandemic. After pandemic, the percentage of those who previously had income more than IDR 10 million has dropped from 48% to 41%. On the contrary, the percentage of those who had income less than IDR 1 million has increased significantly from 16% to 34% after pandemic. The result of the model specification is depicted on Figure 2 below  Figure 2 shows the dynamics of the variables. Fear of Missing out, crisis magnitude, and social media are functioned as determinants for both consumer behaviour and halal preference. Religiosity is placed as a mediating variable on consumer behaviour model, and as endogenous variable on halal preference model.  Table 3 shows that The Cronbach's Alpha coefficients are all above 0.6. The Rho_A coefficient for all variables are more than 0.6. The composite reliability and average variance extracted for all variables are all bigger than 0.7. By those values, the instrument measurement is consistently acceptable as explained by Fornell and Larcker (1981).  Table 4 shows that the VIF value of all the exogenous latent variables are less than 5 which indicates that there is no multicollinearity in the structural model, as according to Hair et al. (2017). The result as captured by Table 5 shows that crisis magnitude, fear of missing out, and social media are found to influence consumer behaviour. The first exogenous variable is significant at alpha 1% while the two later variables are significant at alpha 5%. Millennials are affected by this pandemic. The bigger the perception of the crisis magnitude and the fear of missing out of commodities, and the more they are exposed to COVID-19 news on the social media, the more they are cautious on spending and tend to reduce their consumption on primary, secondary, and tertiary goods. Those which are considered as primary, secondary, and tertiary goods may also change during this pandemic situation. Primary needs are related mainly to basic needs which ensure the life survival only, such as staple food, clothing, and housing. More than before, medicine has become increasingly important not only to survive but also to prevent from decreasing health immune. Before, internet connection might be considered as secondary need; the need which is the fulfilment to beautify the life. Since education and economic transaction are now shifted online, this might be now considered as primary needs. During crisis, what is considered as tertiary or luxurious commodities may shift as well. Because of the decrease in global economy, as also shown by the majority of this study's respondents, standard of living might now change, and what has been considered as expensive might now change as well.
The influence of crisis magnitude to the decrease in spending is however moderated by religiosity. The moderating effect is statistically significant. Religiosity is found to be reducing the inclination of spending decrease during this COVID-19 pandemic. On the other side, crisis magnitude, fear of missing out, and social media are not significant in influencing halal preference. Only religiosity is positively significant at alpha 1%.

CONCLUSION
COVID-19 and the threat of economic crisis has been proven to influence the consumers' behaviour, including Millennials' halal preference and consumption behaviour. Using Theory of Planned Behaviour as the framework to the structural equation models, this study concludes that crisis magnitude, fear of missing out, and social media are found to influence Millennials' consumption behaviour in the capital of Indonesia and the surrounding cities. This research also proves that religiosity moderates the effect of crisis magnitude to consumer behaviour. On the other side, halal preference is also influenced by religiosity.
This study suggests that religiosity has an important role to consumption behaviour. Therefore, paying attention on programs that could increase the religiosity is important. It could prevent the failing confidence of the society towards the economic condition which could call for recession. For halal industry, religiosity also plays significant function. Promoting halal from religious perspective is urgent in order to boost the awareness and the development of this industry.