The Influence of COVID-19 Alcohol Ban on Residential Fires in Cape Town, South Africa
Article information
Abstract
COVID-19 pandemic affected everyone’s life. Different measures were implemented around the world to stop the spread. However, South Africa was the only country where an alcohol ban was imposed for several months. Although alcohol consumption has been related to fire casualties, its relationship with the number of fires has been more difficult to establish. This work studies the influence the alcohol ban had on the number of residential fires in Cape Town. The 2020 fire statistics were compared with data from 2015 to 2019. International work that studied fire statistics during the pandemic was used as benchmark, where data from other cities showed almost no change in the number of fire incidents during lockdowns. The analysis was done in terms of residential fires, formal residential fires and informal residential fires. Results show a drop of 32% in the number of fires during April and 12% in May 2020. After that, the values observed were very similar to the 2015-2019 average until the end of the second ban. Finally, a decrease of 10% was seen in December. It was also seen that the alcohol ban had more impact on informal settlements than on formal settlements. Although it is extremely unlikely this work can be replicated again, it provides novel and difficult to capture information.
1. Introduction
It is fair to say the COVID-19 global pandemic affected everyone’s life. Different measures were implemented in different countries to try to stop the spread[1,2]. Strict lockdowns were seen throughout the world for several months, and South Africa (SA) was not an exception. The National State of Disaster started in March 2020, and it was only lifted in April 2022. During 2020, SA put into place five alert levels (alert level 5 being the most restrictive and alert level 1 the least restrictive), each of them with distinctive restrictions as shown in Table 1. Besides the "typical" international COVID restrictions, such as lockdowns and curfews, SA implemented unique, and some even might say controversial[3], measures that not only restricted the movement of people but also interfered with their drinking and smoking habits. South Africa was the only country where an alcohol and cigarette ban was imposed for several months.
The reasons behind the alcohol ban were very clear, as hospitals were crowded with patients whose injuries came from alcohol related situations. Previous studies have analysed the relationship between violence, injuries, and alcohol in SA. Parry et al.[4] studied alcohol use and risky drinking behavior in the South African population. Alcohol consumption was reported by 45% of the men and 17% of the women. Of those, one-third stated that they were involved in risky drinking over weekends[4]. Furthermore, it was found that alcohol problems correlated with low socioeconomic status, and women with no school education. This is particularly interesting for this research considering the socioeconomical differences seen in South Africa. Research has also shown the influence of alcohol on road traffic accidents. Mabunda et al.[5] found that from 4004 pedestrian fatalities that were analysed, 58% of the deceased tested positive for alcohol. Furthermore, in[6] trauma center admissions were studied, and it was found that from the 9236 patients considered, 30% of the trauma admissions were related to alcohol consumption. This was corroborated in[7] where it was seen that the number of trauma admissions in the hospitals decreased dramatically in South Africa during the alcohol bans. On the other hand, the reasons for the cigarette ban are not so evident. Some studies suggest that smoking increases the probability of drinking and binge drinking[8]. Additionally, according to South Africa’s statistics, the number of fires due to smoking between 2015 and 2019 was only 2.5%[9]. Furthermore, informal settlement residents’ fire risk perception indicates that more than 50% believes the number of fires is associated with drinking while only 5% thinks its linked to smoking[10]. For these reasons, the alcohol and cigarette ban will only be analysed in terms of alcohol consumption reduction. It should be noted that the implementation of the alcohol ban does not mean there was no alcohol consumption altogether. It means that during that period the sale of alcohol was not allowed in restaurants, supermarkets, or liquor shops. It is also important to mention that there was not any forewarning on the alcohol ban. Alcohol sales volumes decreased during this period, with losses of more than R36.3 billion in sales revenue[11]. As a direct result of the alcohol bans, domestic wine sales were down by 20% (in volume)[12]. However, alcohol may have been stockpiled prior to the ban or sold illegally[13]. During the alcohol ban period, the media reported several cases of people trying to produce homemade beer, which depicts people were unable to find alcohol during the first ban[14,15]. Furthermore, Google Trends shows that during the period April 1-April 25 the search of "pineapple beer" and "homemade beer" dramatically increased in South Africa. It is noteworthy that there was a decline in alcohol-related trauma[16] and alcohol’s perceived role in crime[17], highlighting the influence of the restriction in different areas.
Alcohol consumption has also been related to fire injuries and fatalities[18]. It has been acknowledged that alcohol abuse can increase the risk of having a fire[10,19]. While it is not a direct fire cause, it creates dangerous situations that could lead to a fire incident[10,19,20]. Firefighters and residents perceive it as one of the main risks[10,21]. However, its relationship with the number of fires is more difficult to establish. In some countries, firefighters are able to indicate in the fire incident report if they suspect alcohol or drugs might be involved in the fire event[22], but only the deceased and injured people can be tested. It is with this backdrop that the authors saw an opportunity to make a unique contribution by studying if the alcohol ban in SA had an impact on the number of residential fires in Cape Town, South Africa. Furthermore, the work studies if there was a difference between formal and informal dwellings. Informal dwellings or areas may be referred to using terms such as slums, ghettos, shanytowns, shacks or similar, although many of these terms are considered derogatory. The reminder of this work presents (a) the methodology used, (b) the results and (c) the discussion.
2. Methodology
This section presents the methodology used in this work. First, the study area is presented, then the analysis period is defined, and information about the statistics data is provided. Finally, the method used in the analysis of the data is explained.
2.1 Study area
Although the alcohol ban was imposed on the whole country, the work only focuses on Cape Town. The main reason for this is the quality and quantity of data available from the municipality. The City of Cape Town has been reporting fire incidents for more than two decades and their data is readily available to the public from 2015. Cape Town is located in the south-western corner of South Africa, as shown in Figure 1. It has an area of 2461 km2 and a population of 4.7 million people. Additionally, by 2020, Cape Town had a Gini coefficient of 0.626, which shows there is a severe inequality in the area[26].
Location of Cape Town municipal area from which data on residential fire incidents was analysed.
2.2 Analysis period
The analysis period was defined as the year 2020. As shown in Table 1, throughout that time several lockdown measures were implemented. The first quarter was a "normal" one, where no regulations were imposed. That changed at the end of March when the National State of Disaster started. Since then, three alcohol bans were implemented. The first, from March 26th to May 31st (during alert level 5 and alert level 4). The second, from July 12th to August 17th. Finally, the alcohol ban was reintroduced on December 28th.
2.3 Fire statistics data
The authors had access to the City of Cape Town Fire & Rescue database. The database provided information for the 2015-2020 period. The databases contained all fires that took place between 2015 and 2020 in different types of occupancy (i.e., educational buildings, hospitals, industries). Only those in the 'Formal' and 'Informal' category were considered. Data from Suzuki and Manzello[27], based on multiple international cities and fire statistic datasets during the pandemic, was used as benchmark.
2.4 Analysis
In order to determine the influence the alcohol ban had on the number of residential fires, the analysis period will be compared with fire statistics from previous years. Considering the methods applied in similar work[27], the work employed statistics from 2015 to 2019 to understand the fire trends in Cape Town before COVID-19. The minimum, maximum and average values were used for the analysis.
With the aim of studying if socioeconomical factors have an influence in the results, the analysis was done in terms of 'residential fires', 'formal settlement fires' and 'informal settlement fires'. The first category includes the second and third.
3. Results
This section presents a comparison between the number of fires during 2020 and the average from 2015 to 2019. The results were obtained for three cases: (a) residential fires, (b) formal residential fires and (c) informal residential fires (where the former is the summation of the latter two). Table 2 shows a summary for all three cases, presenting the average number of fires, the number of fire incidents in 2020 and the difference between them.
Summary of the Fire Statistics for Cape Town
Figure 2 depicts each case in detail. The bar graph shows the 2020’s fire incidents, where blue represents periods where alcohol sales were allowed and green showcase the periods where alcohol was banned. Additionally, the average number of fires between 2015 and 2019 are shown in the line chart, where this plot also presents the minimum and the maximum values. It can be seen that the average value (2015-2019 period), follows the same trend in the three cases. The peak fire season is summer (November-January) where it is hot and dry, and also parting is more common. After the festive season is over the number of fires reduces (February-May); however, as the days become colder the number of fires starts to slowly increase maintaining a relatively constant value between June and October. Figure 2 shows that 2020’s fire statistics did not follow that trend.
Alcohol ban impact on fires in Cape Town. (a) Residential fires. (b) Formal residential fires. (c) Informal residential fires.
Focusing on the total number of residential fires in Cape Town in Figure 2(a), there is an increase in the number of fires between January and March with respect to the previous years (around 29% rise). This was followed by a sharp decrease, a drop of 32%, in April 2020, almost immediately upon the implementation of the first ban. After that the number of fires increased, reaching values very similar to the 2015-2019 average in June, and remaining the same during the second alcohol ban. In September the number of fires increased, and a peak of 332 fires was observed in October (25% more than the average). Finally, the trends show a decrease in November and December, reaching values 9% less than the average. The fact that during the first quarter the number of fires was 29% higher than average, also highlights that even when the reduction in the number of fires during the second and third alcohol ban were small, there was still an influence. Furthermore, the average number of fires during the months following the first alcohol ban, where alcohol was not restricted, showed an increase of 10%.
The number of formal residential fires in Cape Town is shown in Figure 2(b). The number of formal fires during 2020 follows a similar trend to the one seen for residential fires. During the first quarter, there was an increase in the number of fires with respect to the 2015-2019 average (20% more). In April, the first month of the alcohol ban, there is a significant reduction, reaching a minimum of 73 fires, which is 37% less than the average. Between May and July, the numbers increased back to values slightly higher than the average until July. From August until December values similar to the ones seen in the first quarter were observed. However, November and December had less fires than the 2015-2019 average.
Finally, in Figure 2(c), the number of informal residential fires in Cape Town is shown. As per the previous cases, the number of informal fires during 2020’s first quarter increased by 38% with respect to the 2015-2019 average. In April and May, during the first of the alcohol bans, there is a reduction of 28% in the number of fires. Between June and August, the number of informal fires went up; however, the values were still below the average. From September to November the values increased, reaching a peak of 181 fires in October. In December a decreased of 9% in the number of fires was seen.
4. Discussion
This section discusses the results obtained in the previous section. The analysis is presented in terms of (a) impact of the alcohol ban on residential fires by also considering international trends, (b) differences between formal and informal settlements, and (c) limitations of the study.
4.1 Impact of the alcohol ban on residential fires
Figure 2 shows the alcohol ban had an influence on the number of 'residential fires', 'formal residential fires' and 'informal residential fires', especially during the first month the restriction was implemented. It makes sense the impact was higher during the first alcohol ban than for the second and third, as people were more "prepared" for it. After the first alcohol ban was lifted, people knew there could be another one and could buy alcohol in advance. After the first ban was lifted, liquor stores received many more buyers than usual, with some of them queuing from 4:30 AM[28]. Some liquor stores even had to limit the amount of alcohol a person could buy[28].
The fact that the number of fires went down during the alcohol ban periods confirms the perceptions of firefighters[10,21] and residents[10] with regards to alcohol consumption and fire risk. In[10] it was found that more than 50% of the residents of the Imizamo Yethu informal settlement mentioned they considered that 'irresponsible drinking behaviour' was the greatest fire risk in the settlement.
It has been acknowledged that during the analysis period, several measures besides the alcohol ban were in place; hence, it is not possible to determine which of them had more impact on the reduction in the number of fires. Furthermore, as mentioned above, the fact that there was an alcohol ban does not mean that there was no alcohol consumption. In order to understand the impact other measures had, in particular the stay-at-home restriction, the work done by Suzuki and Manzello[27] was used as benchmark. In[27] the influence of stay-at-home measures in the number of residential fires in New York City, London, San Francisco, and Tokyo were studied. All these cities experienced stay-at-home measures and all non-essential business were closed; however, this was mandatory in the first three cities, while for the latter it was voluntary. The comparison of the results obtained in this work for Cape Town with other cities around the world that did not experience an alcohol ban, allows the work to identify the effect each measure might have had in the number of residential fires.
Figure 3 presents a summary of the measures implemented in each city for the first semester of 2020. In orange is the period where the alcohol ban was imposed and in yellow the period where stay-at-home measures were implemented. Furthermore, Figure 3 includes the difference between number of residential fires in 2020 and the 2015-2019 averages. It is possible to see that while in Cape Town there is a clear reduction in the number of fires, the cities analysed in[27] did not present any appreciable difference in the number of residential fires and no clear trend was found. The number of residential fires were very similar to the average, with the exception of San Francisco and Tokyo which had an increase in March[27] Although the findings from[27] are inconclusive, these could support the hypothesis that the reduction seen in Cape Town in March could be attributed to the alcohol ban, as this was a primary difference between these cities.
Summary of COVID-19 measures and residential fire trends in different cities for the first semester of 2020. Alcohol ban is shown in orange, stay-at-home measure is shown in yellow.
For the second semester of 2020, Table 2 shows an increase in the number of fires, continuing with the trend seen during the first quarter. The exception was December, where the number of incidents was 10% less than the average. That can be attributed to the additional measures that were implemented to prepare for the festive season[19]. Curfews and gathering restrictions were reinstated, while parks and beaches were closed. Furthermore, a third alcohol ban was put in place on 28 December as shown in Figure 2.
4.2 Difference between formal and informal settlements
When comparing the results obtained for formal and informal settlements, differences arise. While the number of informal fires went down throughout both alcohol bans and the festive season, in formal settlements the decrease was only seen in April and December. This could also support the impact of the alcohol ban. As mentioned before, it is very likely that people prepared for upcoming bans, especially formal settlement residents who probably had more budget to spend on alcohol. The authors believe that this could be the main reason why the number of fires did not decrease in formal settlements during the second alcohol ban. Additionally, as time passed, they could have run out of alcohol. This would explain the results seen in July and August, where for the first month the fires decreased by 9% and for the second month by 17%. However, even annual fluctuations in the number of incidents make decisive statements difficult.
4.3 Limitations
The results obtained in this work suggest that alcohol consumption, or at least alcohol availability, increases the probability of a fire event in residential occupancies. However, the findings of this study are limited by (a) the uncertainties associated with the work include that no data regarding alcohol consumption in fire reports is available, (b) the fact that the work has focused on Cape Town only, and (c) the lack of consideration of the impact of geographical and cultural conditions. Typically, these limitations could be improved in future work. However, it is extremely unlikely that this study can be replicated in other areas, or even again in South Africa. That is why, despite the limitations of the study, the work provides novel and difficult to capture information.
5. Conclusions
This work studied the influence of the alcohol ban imposed in South Africa during the COVID pandemic on the number of residential fires. This was a unique opportunity to understand the relationship, or minimally a correlation, between alcohol consumption and the number of residential fires. The 2020 fire statistics from Cape Town were compared with the average obtained for the 2015-2019 period. Three categories were considered 'residential fires', 'formal residential fires', and 'informal residential fires'. Results showed that despite the rise in the number of fires during the first quarter of 2020, once the first alcohol ban was introduced there was a clear reduction in the number of fires, with a drop of around 30% during the first month across all three categories. Furthermore, this followed a trend where the number of fires prior to the ban was higher than in previous years. On the other hand, the second alcohol ban showed a decrease in the number of informal residential fires. This is attributed to the potential preparation the residents might have had prior to the reinstitution of the restriction.
The work suggests that alcohol consumption, or at least alcohol availability, increases the probability of a fire event in residential occupancies for all three categories. However, the results were more consistent for informal settlements, where a reduction in the number of fires was seen throughout the different alcohol bans. This could be due to the greater capacity of formal settlements residents to stockpile alcohol. The study presents several limitations, one of the most important being the fact that it is extremely unlikely that this work can be replicated again. Despite that the work provides novel and difficult to capture information.
Notes
Author Contributions
Conceptualization, N. Flores Quiroz and R. Walls; methodology, N. Flores Quiroz; analysis, N. Flores Quiroz; data curation, N. Flores Quiroz; writing-original draft preparation, N. Flores Quiroz; writing-review and editing, R. Walls; supervision, R. Walls; funding acquisition, R. Walls. All authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
The authors declare no conflict of interest.
Acknowledgments
The authors would like to acknowledge the financial support of (a) the Lloyd’s Register Foundation under the “Fire Engineering Education for Africa” project (Grant GA 100093), (b) the Royal Academy of Engineering/ Lloyd’s Register Foundation under the “Engineering Skills Where They are Most Needed” grant (Grant ESMN 192-1-141), and (c) the SFPE Educational & Scientific Foundation’s Student Research Grant. Furthermore, the assistance of Head Command and Control of the City of Cape Town Fire Rescue Service is gratefully acknowledged. The provision of readily available data by the City of Cape Town greatly contributes to academic research and the understanding of such challenges as possible. Sayaka Suzuki is thanked for initially suggesting this study, albeit it was supposed to be part of a bigger study, which led to the findings of this work being identified.