Skip to main content
NCBI home page
Elsevier - PMC COVID-19 Collection logoLink to Elsevier - PMC COVID-19 Collection
. 2021 Nov 12;9(4):1943–1955. doi: 10.1016/j.cstp.2021.11.003

Assessment of COVID-19 induced travel pattern changes in Dhaka City

Tonmoy Paul 1,⁎, Abu Bakar Siddik Ornob 1, Rohit Chakraborty 1, Nafis Anwari 1
PMCID: PMC8588734  PMID: 34786335

Abstract

In response to Coronavirus 2019 (COVID-19) pandemic, Bangladesh enforced social distancing measures to mitigate the virus transmission rate through lockdowns. However, it is challenging for people to follow through and stay home in developing nations where socio-economic conditions are divergent from developed countries. This research aims to investigate COVID-19 induced travel pattern changes of residents and significant demographic factors affecting the trip generation in Dhaka City, the most densely populated Bangladeshi city. A questionnaire survey was used to extract information on demographic characteristics of respondents in Dhaka City and their travel patterns in the pre-pandemic era and during the pandemic. Analyses reveal striking differences in work trips except for workers and craftsmen. The use of telemedicine facilities is noticeable. Preference for public transport has decreased yet a decent percentage (9%) of people use buses during the pandemic. However, non-motorized modes are also very popular (19.93%) in the pandemic. The findings offer major implications for transportation planners and policymakers on how to dynamically plan for such crisis by combining a range of strategies so that safe and sustainable urban mobility and reduction of unnecessary travel demand can be ensured.

Keywords: COVID-19, Pandemic, Travel Behavior, Travel Characteristics, Ordinal Logistic Regression

1. Introduction

The Coronavirus 2019 (COVID-19) pandemic has killed more than 2 million while infecting an additional 90 million till January 2021. Bangladesh, a developing country, has been particularly affected by COVID-19 (Sakamoto et al., 2020) and is currently placed 32nd in the world (Worldometer, 2021). Dhaka, the ninth-largest and the sixth-most densely populated city in the world (World Urbanization Prospects, 2019), is a COVID-19 hotspot. (WHO, 2020). The main way to control virus transmission is by safeguarding people from COVID-19 exposure (Islam et al., 2020). Advanced nations like South Korea, Taiwan, and New Zealand have noticeably reduced virus transmission cases by following guidelines and strategically restricting travel (Shammi et al., 2021). Although Bangladeshi cities have encouraged social distancing measures, a city like Dhaka accommodating 32% population below the lower poverty line faces challenges in carrying out essential activities while staying at home (Shahadat, 2008). A population of over 18 million residents in Dhaka requires a robust transportation system to facilitate daily activities despite the pandemic situation. Incessant traffic congestion and delays increase physical contact and virus transmission chances in the city, especially among public transport captive riders (Ahmed et al., 2021, Hasnat et al., 2018, Quddus et al., 2019). Moreover, Information and Communication Technologies (ICTs) penetration is very low and skewed towards a few professions among the rich (Ahmed, 2009, Ray and Subramanin, 2013). Thus, people are not accustomed to online banking, education and shopping. Rather, they are habituated with performing activities physically, which is a huge safety concern against COVID-19 transmission.

2. Literature review

Numerous studies on COVID-19 outbreak related to the safety of healthcare, social, food and domestic safety have already appeared in many scientific literatures. The Socio-economic and psychological effect of COVID-19 is also explored (Haghani et al., 2020). Several studies proposed Guidelines to ensure the safe transport of suspected and confirmed COVID-19 patients (Liew et al., 2020). Transmission of COVID-19 has also been examined using travel data (Candido et al., 2020, Zhong et al., 2020, Lytras et al., 2020, Mouchtouri et al., 2020).

Travel has been identified as a prime factor in the global spread of infectious diseases (Gezairy, 2003, Lau et al., 2020, Zheng et al., 2020). Hence several researchers have paid attention to travel pattern change caused by COVID-19 and similar other outbreaks. Travel behavior can be defined as inert or habitual, which does not change very often (de Haas et al., 2018). Though particular key life events and job characteristics of a person have proved essential for decisions that change travel behavior (Lanzendorf, 2003, Oakil et al., 2011). This pandemic can be compared to a life-changing event that can significantly change the travel pattern. However, it affects society as a whole (de Haas et al., 2020).

For example, many people consider public transport as a major source for COVID-19 spread and have hence shifted to private modes in Turkey (Shakibaei et al., 2020), USA (Shamshiripour et al., 2020) and Sweden (Jenelius and Cebecauer, 2020). However, this can also increase traffic congestion and air pollution in the future (Shakibaei et al., 2020). Non-therapeutic measures such as lockdowns, quarantines and social distancing measures have reduced trip numbers in some cases, mainly arising from increased adoption of ICT in Spain (Aloi et al., 2020), the Netherlands (de Haas et al., 2020) and Australia (Beck and Hensher, 2020). On the other hand, essential trips such as food and shopping that cannot be substituted by ICT have remained unchanged in Australia (Beck and Hensher, 2020) and Japan (Parady et al., 2020).

Although most of these studies reported decreased physical travel, they were carried out in developed countries where socio-economic conditions are divergent from developing countries like Bangladesh. According to several COVID-19 travel impact investigations in Bangladesh, people continue to travel for work, and use bus because of its cheapness. (Anwari et al., 2021, Zafri et al., 2021). Increased reliance on urban areas, the concentration of commercial and industrial activities in a few areas, inefficient supply-chain structure, lack of self-sufficient regions, poverty and inequality were some of the main reasons for increased COVID-19 incidence rates in Bangladesh (Rahman et al., 2021). In contrast to previous studies focusing on Bangladesh generally, this study investigates Dhaka City, where the COVID-19 incidence is severe.

3. Methodology

3.1. Data collection

A questionnaire survey was conducted to capture the details on demographic characteristics of the trip-makers and the information complemented to trip generation for “Before” and “During” COVID-19 crisis. Both web-based survey and face to face interview was performed from November 2020 to January 2021. A web-based survey can reach a large number of people within a short period of time and would be comfortable for the respondents to participate during the pandemic in the survey. On the other hand, face to face interview was conducted maintaining proper safety guidelines to extract data from people living in slums in Dhaka City as most of them neither have proper internet connection nor do use smartphones.

The questionnaire survey included 2 sections. The first section extracted details on demographic characteristics of the trip-makers – gender, age, net monthly income (NMI) and occupation. The second section of the questionnaire was designed to ask about the information related to purposes of trip generation and travel characteristics of the respondents for the “Before COVID-19” and “During COVID-19” situations. In this study, a two-way trip is accounted as a single trip for analysis and modelling. The trip purposes are classified as (i) Work/Academic (W/A) Trips, (ii) Shopping Trips, (iii) Grocery Trips, (iv) Recreational Trips, and (v) Medical Trips. To investigate the changes in travel characteristics due to COVID-19 data on mode of travel, the usual start time of the journey and average journey time is taken per trip were collected and investigated.

3.2. Data analysis

Initially, a pilot survey was conducted online on 50 respondents, based on which the final question structure was designed. The final survey collected a total of 823 responses from the trip-makers living in Dhaka City, which was reduced to 806 responses by data screening. The resultant data were processed and analyzed to specifically (i) assess the change in travel characteristics (ii) explore inter-relationships between demographic characteristics and different trip purposes, (iii) to ensure there is no multicolinearity between the chosen demographic factors by testing Spearman's rank Correlation between them, (iv) investigate the effect of those characteristics attributes on overall trip generation by developing home-based trip generation model by Ordinal Logistics Regression (v) discover whether the differences in trip frequency for each purpose between the before and during COVID-19 crisis are statistically significant or not using paired sample t-test. The main methods of analysis are briefly discussed below.

3.2.1. Spearman's rank Correlation

Before performing regression analysis, Spearman's Rank Correlation was used to test for multicollinearity among the demographic characteristics. The presence of ordinal variables necessitates the usage of Spearman’s correlation. In general, an absolute correlation coefficient of >0.7 among two or more predictors indicates the presence of multicollinearity (Akoglu, 2018).

3.2.2. Ordinal Logistics Regression (OLR)

Ordinal logistic regression falls under the class of generalized linear models. The cumulative logit model, also known as the proportional odds model (POM), has been used in this study as OLR. The OLR model usually estimates the set of independent variables using both numerical and categorical responses (Ananth and Kleinbaum, 1997). Since this study consists of categorical variables, OLR is a suitable model to assess demographic characteristics. The model can be defined as:

logit(P(Y≤j))=logP(Y≤j)1-P(Y≤j)=αj+Xβj=1,⋯j-1

where, P is the probability of dependent variable Y.

  • α = Regression Intercept

  • β = Co‐efficient of Independent Variable

  • X = Independent Variable

  • n = 1,2,3…

3.2.3. Paired sample t-test

The paired sample t-test is used to determine whether the mean difference between two sets of observations is significant (Kim et al., 2018).

4. Results and analysis

4.1. Demographic characteristics of respondents

Demographic attributes of survey respondents are summarized in Fig. 1 where most respondents are male (57.6%) and are aged 19–24 years (26.6%). Most respondents (41%) were low-income people (0–8000 BDT). In this study, the occupation worker is considered as day laborer, masonry worker, building painter. Craftsman is considered as mechanics, plumbers, electricians, artisan, smith, manufacturer etc. Unemployed is the person who has no job or searching for a job and “Others” includes persons with undefined occupations. In terms of occupation, students (30%) and private service holders (22%) are the most active participants. The response observation for the age group and occupation is due to the digital nature of the survey as young people (typically students) are more advanced in ICT facilities than adults are.

Fig. 1.

Fig. 1

Open in a new tab

Distribution of Gender, Age, Income and Occupation of respondents.

4.2. Travel pattern change reflected by travel characteristics

The specific modes presented as options in the questionnaire survey are based on consultation with transportation experts. A wide variety of modes run in the streets of Dhaka City, including rickshaw and leguna. Rickshaw is a three-wheeled, non-motorized vehicle that is a very popular and easily available non-motorized vehicle (NMV) in Bangladesh (Ismat Esrar, 1992) that plays a crucial role in Dhaka's transport system (Hossain, 2011). On the other hand, Leguna is a short-distance vehicle, also known as a local passenger van. They are recognized as a special transportation service in Dhaka City with high flexibility and availability in selected routes only (Rahman, 2016). The public transportation of Dhaka City consists of primarily buses and leguna.

Analysis of travel mode choice as shown in Fig. 2 reveals a drastic fall in bus and leguna use during the pandemic. Usually, in public transport, large number of people travel together within a limited space. COVID-19 pandemic induced social distancing measures reduced the preference for bus and leguna, while sharply increasing online activity. A large percentage of the respondents are students/young adults whose academic institutions and offices have shifted activity from physical tasks to online. The use of nearly all the modes of travel has decreased except for cycle and walking, where the preference of cycle increased slightly and that of walking escalated dramatically during COVID-19 period.

Fig. 2.

Fig. 2

Open in a new tab

The preference of overall travel mode travel Before and During COVID-19.

Inertia analyses of overall mode preferences using Sankey diagrams as shown in Fig. 3 assessed whether or not the respondents continue to rely on the same mode used before COVID-19. The overall mode preference before and during COVID-19 are displayed on the left and right sides of Fig. 3 respectively. Fig. 3 depicts that almost 50% bus users shifted to online while a smaller percentage shifted to private cars or ride sharing cars. The remaining people still using bus are at great risk of being affected by COVID-19. However, a small percentage of ride sharing car users shifted to the more available and cheaper rickshaw.

Fig. 3.

Fig. 3

Open in a new tab

Inertia of various mode of travel usage.

Fig. 4 reveals the variation in average journey duration at pre-COVID and COVID times. Although the journey time has decreased from a majority of 31–45 min before COVID-19 to 16–30 min during COVID-19 (because of lower vehicle density on roads), the reduction is small. This is because a considerable number of people continue to make trips because of the physical nature of their jobs. Consequently, the proportion of people travelling during morning rush hour has not changed much during COVID-19, as shown in Fig. 5 . Thus virus transmission may be high.

Fig. 4.

Fig. 4

Open in a new tab

The average journey duration Before and During COVID-19.

Fig. 5.

Fig. 5

Open in a new tab

Usual journey start time Before and During COVID-19.

4.3. Analyses of trip purposes and overall trip generation

Fig. 6 represents the percentage of respondents making trips of various frequencies for work and academic purposes per week. Fig. 6 reveals a large drop in (5–8 times per week) W/A trips, which have been replaced by online activities. Most of these trips would be the trips to educational institutes. However, 37.7% respondents still travel 5–8 times per week while 10.4% travel 1–4 times per week during COVID-19, indicating these people still go to work physically. So, compared to students, the job-holders and earners of the family are at greater risk of getting infection. In this study, shopping trips are considered as trips to shopping centers and malls to purchase cosmetics, electronic products and other luxurious products. In the figure FUTMST stands for “Feels Unnecessary To Make Shopping Trips”. As shown in Fig. 7 , both shopping and grocery trips have decreased. People are either buying the bare necessities or buying more things in the same trip. Interestingly, shopping trips have decreased more than grocery trips have, possibly because of greater online prevalence for shopping trips.

Fig. 6.

Fig. 6

Open in a new tab

Trip frequency responses for different trip purposes.

Fig. 7.

Fig. 7

Open in a new tab

Overall Number of Trips (ONT) Before and During COVID-19.

The proportion of respondents disinterested in recreational trips more than doubled to 78.7% during COVID-19. This is because people feel recreation trips are considered as non-essential trips and thus avoid these trips to protect themselves from virus contagion. On the other hand, medical trips have decreased less than other types of trips. Although, the trips for other purposes declined, people still have to make trips to hospitals, medical test centers and to pharmacies. The situation has been exacerbated because of transportation of COVID-19 patients. Hence, a scope still remains for the spreading of virus through trip generation. There has been a slight increase in the telemedicine activity, which matches with the report from National Telehealth Centre (DGHS, 2021a). This should be promoted more in the future.

Overall, respondents have shifted from high frequency trips to lower frequencies as presented in Fig. 7. Still, a considerable amount of trips is generated during the pandemic as offices, markets and emergency services are still open except for academic institutes.

4.4. Examining the difference in trip frequency for each purposes

Five separate t-tests have been analyzed for five trip purposes. Like many statistical procedures, the null hypothesis asserts that there is no significant difference in the true mean between the paired samples. On the other hand, the alternate hypothesis assumes that there is a significant difference between the means. Table 1 reveals a significant change in trip frequency in all trip purposes since the significant value (Sig. value) is less than 0.05 at 95% confidence level for all cases. Based on the t-value, shopping trips hold the greatest difference while medical trips hold the least differences due to the COVID-19 pandemic.

Table 1.

Paired sample t-test of trip purposes Before COVID-19 (BC-19) and During COVID-19 (DC-19).

Paired Differences
 t df Sig. (2-tailed)
Mean Std. Deviation Std. Error Mean 95% Confidence Interval of the Difference
Lower Upper
Pair 1 BC19 W/A Trips - DC19 W/A Trips 0.715 1.099 0.039 0.639 0.791 18.460 805 0.000
Pair 2 BC19 Shopping Trips - DC19 Shopping Trips 0.553 0.589 0.021 0.513 0.594 26.675 805 0.000
Pair 3 BC19 Grocery Trips - DC19 Grocery Trips 0.583 0.860 0.030 0.524 0.643 19.241 805 0.000
Pair 4 BC19 Recreation Trips - DC19 Recreation Trips 0.687 0.826 0.029 0.630 0.744 23.625 805 0.000
Pair 5 BC19 Medical Trips - DC19 Medical Trips 0.089 0.696 0.025 0.041 0.137 3.646 805 0.000
Open in a new tab

4.5. Association between trip purposes and demographic characteristics

Fig. 8 temporally compares trip frequency for various trip purposes related to the characteristics using stacked graphs. It can be seen that, W/A trips reduced during COVID-19 as expected, particularly the reduction by females was more than of males. This indicates that males are more vulnerable to COVID-19 infection. A large percentage of people from age groups 25–30, 31–40 and 41–60 still work physically per week from office to ensure smooth operation. This age group refers adults, who are often the main income source of the family, and are at high risk of COVID-19 contagion (DGHS, 2021a). There is negligible change in trips for self-employed people, workers and craftsmen because they have little knowledge in ICT, and their jobs cannot be performed online. Thus blue collar workers are at more risk than white collar workers are in contracting COVID-19.

Fig. 8.

Fig. 8

Fig. 8

Fig. 8

Open in a new tab

(a): Temporal comparison of trip frequency for Work trip purpose, (b). Temporal comparison of trip frequency for Shopping trip purpose, (c). Temporal comparison of trip frequency for Grocery trip purpose, (d): Temporal comparison of trip frequency for Recreation trip purpose, (e): Temporal comparison of trip frequency for medical trip purpose.

Fig. 8 reveals that females generated less shopping trips during the pandemic compared to males, highlighting the greater adaptation of females to the usage of online shopping facilities (Dasha, 2019). Children and the elderly reduced shopping trips significantly during COVID-19 whereas the younger and middle-aged people still make noticeable shopping trips. Worker, craftsman and people from “Others” occupation stopped making shopping trips and feel unnecessary to make shopping trips during the crisis.

The reason for less reduction of grocery trips could be the regular obligatory demand of grocery items. In fact, people aged between (31–40) category and (41–60) are going to markets physically to buy groceries, exposing themselves to virus infection. Lack of online activity in Fig. 8 for purchasing groceries in “Craftsman” category describes that not all people have access or knowledge to use internet.

All categories of people drastically reduced their recreational trips. All high frequency trip generations declined, while most people became disinterested in making recreational trips as mentioned earlier in Section 4.4.

There has been a noticeable increase in the usage of Telemedicine service. People have become more interested during the pandemic to satisfy their medical purposes using fewer trips.

4.6. Correlation matrix

It can be seen from Fig. 9 that none of the value of rs is above 0.7; meaning that factors are distinct and independent of each other referring there is no multicollinearity between the factors. Hence, it would be possible to examine the individual explanation of all the independent variables on the dependent variable by statistical modelling which is explained in the following section.

Fig. 9.

Fig. 9

Open in a new tab

Correlation matrix for demographic attributes of trip-makers.

4.7. Home-based trip generation model

This section presents the estimated parameters and Odd Ratio (OR) of OLR models to assess the effect of trip-makers demographic characteristics attributes on the overall trip generation. OR indicates the multiplicative change in the odds of being in a higher category on the Overall Number of Trips (ONT) generated before COVID-19 for every one-unit increase on the demographic factors, holding the remaining factors constant. Two models have been analyzed, one focusing on trip numbers before COVID-19 pandemic, and another one during the pandemic. The independent variables are chosen after ensuring there was no multicollinearity among the independent variables, as shown in Fig. 9.

  • Model 1

Model 1 (M1) investigates the effect of demographic factors on the overall number of trips (ONT) generated before COVID-19, where the considered factors are Gender (Y1), Age (Y2), Net monthly income (NMI) (Y3) and Occupation (Y4). From Table 2 , it can be seen that Gender, Age and NMI have significant impact on ONT. The significance value (Sig.) is less than 0.05, which indicates that the variables are statistically significant. Estimates (β) for Y2 is negative, meaning older people made fewer trips before COVID-19. The estimate parameter for Y3 is positive, referring before COVID-19 high income group generate more trips than the remaining people. Since, the significance level of Y4 greater than 0.05, generation of (ONT) before COVID-19 is independent of occupation. OR for Age (0.780) is below 1, indicates moving from one category to another (downward), the odd of increasing ONT generated before COVID-19 is 0.780 by one ordinal scale.

Table 2.

Overall number of trips generated before COVID-19 model estimation (M1)

Estimate (αj) Std. Error Wald df Sig. Odd Ratio (OR) 95% Confidence Interval
Lower Bound Upper Bound
Threshold [ONT before COVID-19 = 0] −1.437 0.187 59.301 1 0.000 −1.803 −1.072
[ONT before COVID-19 = 1] 0.596 0.178 11.259 1 0.001 0.248 0.944
[ONT before COVID-19 = 2] 2.668 0.205 169.877 1 0.000 2.267 3.069
[ONT before COVID-19 = 3] 3.923 0.254 237.647 1 0.000 3.424 4.421



Estimate (β)
Demographic Factors Y1: Gender 0.952 0.139 46.754 1 0.000 2.591 0.679 1.225
Y2: Age −0.248 0.056 19.812 1 0.000 0.780 −0.358 −0.139
Y3: Income 0.378 0.053 51.005 1 0.000 1.459 0.274 0.481
Y4: Occupation 0.041 0.029 1.948 1 0.163 1.049 −0.016 0.098
Open in a new tab

Based on Table 2 the equation can be expressed as

logPY≤j1-PY≤j=αj+0.952Y1-0.248Y2+0.378Y3+0.041Y4where,j=1,⋯.J-1

  • Model 2

Model 2 (M2) investigates the significance of demographic factors on the overall number of trips (ONT) generated during COVID-19 pandemic, where the considered factors are Gender (Y1), Age (Y2), Net monthly income (NMI) (Y3) and Occupation (Y4). During COVID-19, the travel behavior has changed. From Table 3 , it can be seen that Gender, NMI and Occupation have significant impact on ONT during COVID-19 period. Since, Y2 has significance level greater than 0.05, generation of (ONT) during COVID-19 is independent of Age. The estimate (β) for NMI is still positive. This indicates, even in the pandemic people with high income generate more trips.

Table 3.

Overall number of trips generated during COVID-19 model estimation (M2)

Estimate (αj) Std. Error Wald df Sig. Odd Ratio (OR) 95% Confidence Interval
Lower Bound Upper Bound
Threshold [ONT during COVID-19 = 0] 1.948 0.216 81.494 1 0.000 1.525 2.371
[ONT during COVID-19 = 1] 3.772 0.247 232.855 1 0.000 3.287 4.256
[ONT during COVID-19 = 2] 6.367 0.339 352.969 1 0.000 5.703 7.031
[ONT during COVID-19 = 3] 7.203 0.423 290.086 1 0.000 6.374 8.032



Estimate (β)
Demographic Factors Y1: Gender 1.083 0.155 48.934 1 0.000 2.955 0.780 1.387
Y2: Age −0.092 0.060 2.347 1 0.126 0.912 −0.210 0.026
Y3: Income 0.568 0.058 94.657 1 0.000 1.764 0.453 0.682
Y4: Occupation 0.304 0.032 87.581 1 0.000 1.355 0.240 0.368
Open in a new tab

Based on Table 3 the equation can be expressed as

logPY≤j1-PY≤j=αj+1.083Y1-0.092Y2+0.564Y3+0.304Y4where,j=1,⋯.J-1

4.7.1. Model fitting information

The significance value (Sig.) of the chi-square test for both ordered regression models are less than 0.05. Therefore, we reject the null hypothesis where null hypothesis states there is no significant difference between Baseline Model to Final Model. As a result, there is a significant difference between Baseline Model to Final Model, hence the Final Model is well established. For both the models, the (Sig.) of the Pearson statistics and deviation statistics are less than 0.05 for the fitting degree test, which indicates that the fitting degree of both the models are good (Zhang et al., 2019). Suitabilities are assessed and confirmed for M1 and M2 respectively using the following goodness of fit indicators: log Likelihood (1080.840, 1002.441), Pearson goodness of fit statistic (1104.490, 1391.792), Deviance statistic (897.611, 878.733) (Cramer, 1987, Vrieze, 2012).

5. Discussion and policy implications

These findings display an unconventional change in the urban mobility landscape. As the world is still fighting the pandemic, this change can transcend into a new regular way of living. Hence, priority must be given to appropriate transport planning and policy measures reflecting on sustainable urban mobility objectives and accomplishing the greater community goals under such crucial circumstances.

Since public transport has been associated with an increased risk of viral transmission, (Anwari et al., 2021) the use of bus and leguna has drastically declined during COVID-19. On the other hand, private car usage has not decreased much during the pandemic period, as people feel safer travelling in private cars. Consequently, traffic congestion and air quality in Dhaka City continue to get worse (Rahman et al., 2019, Sakib, 2021). Interestingly, the rich prefer to use private cars, while the poor and middle class rely on public transport during pandemic. 8.91% of respondents continue to use the bus during pandemic (Fig. 2) as the travel cost is far less than that of private transport, especially on long journeys. Thus, transport planners and operators need to focus on improvements of public transport by upgrading both safety and level of service of buses and leguna to ensure a sustainable transportation system for Dhaka City. The internal design of buses can be modified and rearranged while using dividers to facilitate physical distancing among passengers as they board and descend from them (Bhaduri et al., 2020, Politis et al., 2021). To prevent virus transmission through banknote payments in buses and leguna, contactless payment like mobile financial services (MFS) can be implemented.

Another mode which the Bangladesh government should focus on is the rickshaw. Fig. 2 reveals a negligible change in rickshaw usage during the pandemic. Rickshaw occupies a special role in being both affordable to the poor and offering greater social seclusion compared to public transport. However, the rickshaw is usually suitable for short-distance travel. The current governmental limitation imposed on bus usage (to 50% seats) has forced some users to switch to rickshaws, which may explain the relatively unchanged travel preference for rickshaws during the pandemic (Anwari et al., 2021). However, the problem with rickshaws is that rickshaw pullers are either capricious regarding sanitizing the passenger seat or unable to afford financially the disinfectants and safety masks. Thus, the viral transmission increases through high-volume movements of rickshaws. Recommending minimal use of rickshaws can be an easy way to reduce the rate of transmission, but this decision can cut the income source of the rickshaw pullers who hail from low-income households. In addition, low and middle-income groups are more likely to be socially excluded if both the bus and rickshaw are not available anymore. Hence, transport policymakers need to emphasize creating smart and sustainable solutions to ensure the safe travelling of the people hailing from poor families. Policies like the provision of face masks, facilities for sanitization to the rickshaw pullers may help to resolve the concerns of the NMV users.

Although Dhaka’s poverty rate has dropped in recent years, still many inhabitants belong to the low-income group (Population Stat, 2021). These people are still making a substantial number of trips and that in public transport that leads them at high risk of getting infected. Bangladesh government released guidelines to prevent COVID-19 infection and transmission (DGHS, 2021b), yet it is increasing at an alarming rate from March 2021 (MIS, 2020). As a result, country-wide lockdown has been enforced again anticipating the second wave of COVID-19 (Business Standard). Consequently, the income of the indigent rickshaw pullers and C.N.G. auto-rickshaw drivers have drastically decreased. According to the World Bank Report, Bangladesh reached a lower-middle-income status in 2015, but the COVID-19 pandemic impacted Bangladesh immensely (World Bank, 2021). Policymakers from the transportation sector should prioritize safe transportation of workers, craftsmen during this pandemic through dynamic planning. In this regard, the use of cycles and rickshaws with proper sanitizing facilities can be recommended; so that all these people hailing from the low-income groups can do safe travels and at the same time maintain their income.

Active transport (bicycle and walking) can be promoted during the pandemic to ensure social distancing while retaining mental and physical fitness (De Vos, 2020). Despite the potential of bicycles to reduce air pollution and help maintain a healthy lifestyle, its usage has not increased much during the pandemic (Fig. 2) because of insufficient facilities for bicycle riders in the roadway (Chiran, 2021). Transport infrastructure operators should thus give emphasis on building and maintaining safe and bicycle lanes to promote cycling amongst the youth and to attract the private vehicle riders to shift to bicycles.

Completely opposite to the trend observed in other modes, pedestrian activity has jumped in the COVID-19 pandemic (Fig. 2). Studies in other countries also revealed similar results on walking (Aloi et al., 2020, de Haas et al., 2020, Mogaji, 2020). However, previous studies have revealed the lack of safe walking conditions in Dhaka City (Debnath et al., 2021). Thus, the Government needs to improve walkway facilities in Dhaka City.

According to Fig. 8, male trip-makers travel more compared to females even during the pandemic. Hence, they are more susceptible to COVID-19 infection. As a result, maximum COVID-19 patients are males in Bangladesh (DGHS, 2021a). Numerous studies have revealed similar behavior on the risk of male travelers (Abdullah et al., 2020, Bhaduri et al., 2020, Anwari et al., 2021, Politis et al., 2021). This is because in developing countries like Bangladesh, mostly males have to work outside to earn for the family, whereas females look after the home (Ahmed et al., 2021). Hence, local authorities need to create awareness among the men and take necessary action so that males take extra safety practices during their pandemic trips. On the other hand, adults aged between 31 and 60 years have generated a considerable amount of trips during the pandemic for working and buying groceries physically, as shown in Fig. 8. Thus, national authorities and policymakers should focus on these specific age groups of people to minimize exposure to the virus.

Although the average travel time decreases during COVID-19, the traffic congestion and travel delay still occurred during the peak period of the day, mostly during 5:00 AM-9: 00 AM (Fig. 5). This may still create crowding situations and stimulate virus transmission. Policymakers should dynamically plan by combining a range of strategies involving regulation of flexible operating hours of various managerial departments and essential businesses implementation of the role of ICT in behavioral changes. Developing countries like Bangladesh where a large proportion of people do not have access or have limited usage knowledge about ICT tools should focus on ensuring how people can still participate in activities that have largely shifted to ICT solutions to reduce unnecessary travel demand (Anwari et al., 2021). Helplines and psychological assistance services can be widened for all age groups and occupations to use the internet to meet a variety of social needs. All meetings, conferences and desk jobs should be encouraged to be performed online.

6. Conclusion

The main concern of this study was to investigate the change in travel behavior and trip characteristics of residents of Dhaka City due to the outbreak of COVID-19. A questionnaire survey was conducted online along with a face-to-face interview to assess the travel behavior of 806 Dhaka City residents. Through various descriptive analyses, appropriate statistical tests, and home-based trip-generation models, this study has compared the fundamental mobility patterns at during-COVID-19 and pre-COVID-19 levels.

Bus and leguna usage have declined during COVID-19, yet 9% of respondents still use those modes during the pandemic and hence are at increased risk of disease contraction. On the other hand, there is a negligible decrease in the usage of cars and rickshaws. Moreover, cycle usage has increased slightly while walking has increased dramatically, indicating people’s increased preference for active transport. Fig. 4 shows that the trip duration of most trips has decreased from 30 to 45 min (before pandemic) to 16–30 min during the pandemic. However, respondents’ preference for making trips during the morning peak periods remain considerably high during the pandemic (Fig. 5). Paired sample t-tests revealed significant differences in trips, particularly in shopping trips. Stacked graphs display that during COVID-19, females use more online facilities to do academic classes, perform their works and shopping compare to males. Online education/work tendency decreases among people aged over 24 years. Blue-collar workers still have to commute physically. Although people reduced their shopping trip frequencies, adults aged 31–60 still visit grocery markets physically 1–2 times per week. OLR modelling reveals that gender, age and NMI had significant impact on overall travel before COVID-19, while gender, NMI and occupation had a significant impact on overall travel during COVID-19.

This study suggests numerous future research insights. First, there is a need for continued estimations considering more exclusive factors that will enable researchers to measure how travel experiences and behavior change in a broader way over time till the COVID-19 pandemic continues. Second, future research can estimate changes in activities and travel behavior post-pandemic.

CRediT authorship contribution statement

Tonmoy Paul: Methodology, Software, Validation, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Visualization. Abu Bakar Siddik Ornob: Methodology, Software, Validation, Formal analysis, Investigation, Writing – original draft, Visualization. Rohit Chakraborty: Methodology, Software, Validation, Writing – original draft, Visualization. Nafis Anwari: Conceptualization, Supervision, Writing – review & editing.

References

  1. Abdullah M., Dias C., Muley D., Shahin M.d. Exploring the impacts of COVID-19 on travel behavior and mode preferences. Transp. Res. Interdiscip. Perspect. 2020;8:100255. doi: 10.1016/j.trip.2020.100255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ahmed, M.T., Hadiuzzaman, M., Islam, M.R., Anwari, N., Islam, M.A.A., 2021. Exploring Factors Affecting Commuters’ Perceptions on Intercity Rail Service Quality using Ordinal Logistic Regression. TRB 100th Annu. Meet.
  3. Ahmed S. Bangladesh’s economy: surrounded by deadly threats. Int. J. Soc. Econ. 2009;36(1/2):138–181. doi: 10.1108/03068290910921244. [DOI] [Google Scholar]
  4. Akoglu H. User’s guide to correlation coefficients. Turkish J. Emerg. Med. 2018;18(3):91–93. doi: 10.1016/j.tjem.2018.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Aloi A., Alonso B., Benavente J., Cordera R., Echaniz E., Gonzalez F., Ladisa C., Lezama-Romanelli R., Lopez-Parra A., Mazzei V., Perucci L., Prieto-Quintana D., Rodriguez A., Sanudo R. Effects of the COVID-19 lockdown on urban mobility: empirical evidence from the city of Santander (Spain) Sustainability. 2020;12:1–18. [Google Scholar]
  6. Ananth C.V., Kleinbaum D.G. Regression models for ordinal responses: a review of methods and applications. Int. J. Epidemiol. 1997;26:1323–1333. doi: 10.1093/ije/26.6.1323. [DOI] [PubMed] [Google Scholar]
  7. Anwari N., Tawkir Ahmed M.d., Rakibul Islam M.d., Hadiuzzaman M.d., Amin S. Exploring the travel behavior changes caused by the COVID-19 crisis: a case study for a developing country. Transp. Res. Interdiscip. Perspect. 2021;9:100334. doi: 10.1016/j.trip.2021.100334. [DOI] [Google Scholar]
  8. Bangladesh Population, Population Stat [WWW Document], 2021. Popul. world Stat. data. URL <https://populationstat.com/bangladesh/>.
  9. Beck M.J., Hensher D.A. Insights into the impact of COVID-19 on household travel and activities in Australia – the early days under restrictions. Transp. Policy. 2020;96:76–93. doi: 10.1016/j.tranpol.2020.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bhaduri E., Manoj B.S., Wadud Z., Goswami A.K., Choudhury C.F. Transportation Research Interdisciplinary Perspectives Modelling the effects of COVID-19 on travel mode choice behaviour in India. Transp. Res. Interdiscip. Perspect. 2020;8:100273. doi: 10.1016/j.trip.2020.100273. [DOI] [Google Scholar]
  11. Business-standard, https://www.business-standard.com/article/current-affairs/bangladesh-govt-contemplates-full-lockdown-for-a-week-from-april-14-121041000206_1.html. (Accessed 10 April 2021).
  12. Chiran, M., 2021. The bike lanes to nowhere [WWW Document]. Dly. Star. URL <https://www.thedailystar.net/city/news/the-bike-lanes-nowhere-2038397>.
  13. Cramer J.S. Mean and variance of R2 in small and moderate samples. J. Econom. 1987;35:253–266. doi: 10.1016/0304-4076(87)90027-3. [DOI] [Google Scholar]
  14. Candido Darlan Da S., Watts Alexander, Abade Leandro, Kraemer Moritz U.G., Pybus Oliver G., Croda Julio, de Oliveira Wanderson, Khan Kamran, Ester C Sabino N.R.F. Routes for COVID-19 importation in Brazil. J. Travel Med. 2020;27 doi: 10.1093/jtm/taaa042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dasha M, 2019. Men vs. Women. Who Runs the Online Shopping World? [WWW Document]. Belvg. URL <https://belvg.com/blog/men-vs-women-who-runs-the-online-shopping-world.html>.
  16. de Haas M., Faber R., Hamersma M. How COVID-19 and the Dutch ‘intelligent lockdown’ change activities, work and travel behaviour: evidence from longitudinal data in the Netherlands. Transp. Res. Interdiscip. Perspect. 2020;6:100150. doi: 10.1016/j.trip.2020.100150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. de Haas M.C., Scheepers C.E., Harms L.W.J., Kroesen M. Travel pattern transitions: applying latent transition analysis within the mobility biographies framework. Transp. Res. Part A Policy Pract. 2018;107:140–151. doi: 10.1016/j.tra.2017.11.007. [DOI] [Google Scholar]
  18. De Vos J. The effect of COVID-19 and subsequent social distancing on travel behavior. Transp. Res. Interdiscip. Perspect. 2020;5:100121. doi: 10.1016/j.trip.2020.100121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Debnath M., Hasanat-E-Rabbi S., Hamim O.F., Hoque M.S., McIlroy R.C., Plant K.L., Stanton N.A. An investigation of urban pedestrian behaviour in Bangladesh using the Perceptual Cycle Model. Saf. Sci. 2021;138:105214. doi: 10.1016/j.ssci.2021.105214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. DGHS, 2021a. National Telehealth Center.
  21. DGHS, 2021b. Resistance and Remedy Guidelines.
  22. Gezairy H.A. Travel epidemiology: WHO perspective. Int. J. Antimicrob. Agents. 2003;21(2):86–88. doi: 10.1016/S0924-8579(02)00365-5. [DOI] [PubMed] [Google Scholar]
  23. Haghani M., Bliemer M.C.J., Goerlandt F., Li J. The scientific literature on Coronaviruses, COVID-19 and its associated safety-related research dimensions: a scientometric analysis and scoping review. Saf. Sci. 2020;129:104806. doi: 10.1016/j.ssci.2020.104806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hasnat M.M., Islam M.R., Hadiuzzaman M. Emergency response during disastrous situation in densely populated urban areas: a GIS based approach. Geogr. Tech. 2018;13:74–88. doi: 10.21163/GT. [DOI] [Google Scholar]
  25. Hossain M., Susilo Y.O. Rickshaw Use and Social Impacts in Dhaka, Bangladesh. Transportation Research Record. 2011;2239(1) doi: 10.3141/2239-09. [DOI] [Google Scholar]
  26. Islam S., Emran G.I., Rahman E., Banik R., Sikder T., Smith L., Hossain S. Knowledge, attitudes and practices associated with the COVID-19 among slum dwellers resided in Dhaka City: a Bangladeshi interview-based survey. J. Public Health (Bangkok) 2020;43(1):13–25. doi: 10.1093/pubmed/fdaa182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ismat Esrar Md. Home-based trip generation modelling for Dhaka city. Buet; 1992. http://lib.buet.ac.bd:8080/xmlui/handle/123456789/2902 [Google Scholar]
  28. Jenelius E., Cebecauer M. Impacts of COVID-19 on public transport ridership in Sweden: analysis of ticket validations, sales and passenger counts. Transp. Res. Interdiscip. Perspect. 2020;8:100242. doi: 10.1016/j.trip.2020.100242. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kim, H., Park, C., Wang, M., 2018. Paired t–test based on robustified statistics.
  30. Lanzendorf, M., 2003. Mobility biographies. A new perspective for understanding travel behaviour. In: 10th Int. Conf. Travel Behav. Res. 20.
  31. Lau H., Khosrawipour V., Kocbach P., Mikolajczyk A., Schubert J., Bania J., Khosrawipour T. The positive impact of lockdown in Wuhan on containing the COVID-19 outbreak in China. J. Travel Med. 2020;27:1–7. doi: 10.1093/jtm/taaa037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Liew Mei Fong, Siow Wen Ting, Yau Y.W., See K.C. Safe patient transport for COVID-19. Crit. Care. 2020;24 doi: 10.1186/s13054-020-2828-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. MIS, D., 2020. Coronavirus COVID-19 Dashboard [WWW Document]. URL <http://103.247.238.92/webportal/pages/covid19.php> (accessed 4.17.21).
  34. Mogaji E. Impact of COVID-19 on transportation in Lagos, Nigeria. Transp. Res. Interdiscip. Perspect. 2020;6 doi: 10.1016/j.trip.2020.100154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nations, U., 2019. World Urbanization Prospects.
  36. Oakil A.T.M., Ettema D., Arentze T.A., Timmermans H.J.P. 16th Int. Conf. Hong Kong Soc. Transp. Stud. 2011. A longitudinal analysis of the dependence of the commute mode switching decision on mobility decisions and life cycle events; pp. 1–8. [Google Scholar]
  37. Parady G., Taniguchi A., Takami K. Travel behavior changes during the COVID-19 pandemic in Japan: analyzing the effects of risk perception and social influence on going-out self-restriction. Transp. Res. Interdiscip. Perspect. 2020;7:100181. doi: 10.1016/j.trip.2020.100181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zhong Ping, Guo Songxue, Chen T. Correlation between travellers departing from Wuhan before the Spring Festival and subsequent spread of COVID-19 to all provinces in China. J. Travel Med. 2020;27(3) doi: 10.1093/jtm/taaa036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Politis I., Georgiadis G., Papadopoulos E., Fyrogenis I., Nikolaidou A., Kopsacheilis A., Sdoukopoulos A., Verani E. Transportation research interdisciplinary perspectives COVID-19 lockdown measures and travel behavior: the case of Thessaloniki. Greece. Transp. Res. Interdiscip. Perspect. 2021;10 doi: 10.1016/j.trip.2021.100345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Quddus M., Rahman F., Monsuur F., de Ona J., Enoch M. Analyzing bus passengers’ satisfaction in Dhaka using discrete choice models. Transp. Res. Rec. 2019;2673(2):758–768. doi: 10.1177/0361198119825846. [DOI] [Google Scholar]
  41. Rahman M.H., Zafri N.M., Ashik F.R., Waliullah M.d., Khan A. Identification of risk factors contributing to COVID-19 incidence rates in Bangladesh: a GIS-based spatial modeling approach. Heliyon. 2021;7(2) doi: 10.1016/j.heliyon.2021.e06260. e06260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rahman S M Mostafizur. Service quality assessment of paratransit mode by structure equation model (SEM) in Dhaka city. Military Institute of Science and Technology; 2016. [Google Scholar]
  43. Rahman M.M., Mahamud S., Thurston G.D. Recent spatial gradients and time trends in Dhaka, Bangladesh, air pollution and their human health implications. J. Air Waste Manag. Assoc. 2019;69(4):478–501. doi: 10.1080/10962247.2018.1548388. [DOI] [PubMed] [Google Scholar]
  44. Ray D., Subramanin S. India’s lockdown: an interim report. J. Chem. Inf. Model. 2013;53:1689–1699. [Google Scholar]
  45. Zheng Ruizhi, Yu Xu, Wang Weiqing, Ning Guang, Bi Y. Spatial transmission of COVID-19 via public and private transportation in China. Travel Med. Infect. Dis. 2020;34:101626. doi: 10.1016/j.tmaid.2020.101626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sakamoto M., Begum S., Ahmed T. Vulnerabilities to COVID-19 in Bangladesh and a reconsideration of sustainable development goals. Sustainability. 2020;12:1–15. doi: 10.3390/su12135296. [DOI] [Google Scholar]
  47. Sakib, S.N., 2021. Bangladesh: Air pollution engulfs lives, environment [WWW Document]. Anadolu Agency. URL <https://www.aa.com.tr/en/environment/bangladesh-air-pollution-engulfs-lives-environment/2190506>.
  48. Shahadat H. Rapid Urban Growth and Poverty in Dhaka City. Urban Stud. 2008;5:1–24. [Google Scholar]
  49. Shakibaei S., de Jong G.C., Alpkökin P., Rashidi T.H. Impact of the COVID-19 pandemic on travel behavior in Istanbul: a panel data analysis. Sustainable Cities Soc. 2020;65:102619. doi: 10.1016/j.scs.2020.102619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Shammi M., Bodrud-Doza M.d., Islam A.R.M.T., Rahman M.M. Strategic assessment of COVID-19 pandemic in Bangladesh: comparative lockdown scenario analysis, public perception, and management for sustainability. Environ. Dev. Sustainbility. 2021;23(4):6148–6191. doi: 10.1007/s10668-020-00867-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Shamshiripour A., Rahimi E., Shabanpour R., Mohammadian A., (Kouros) How is COVID-19 reshaping activity-travel behavior? Evidence from a comprehensive survey in Chicago. Transp. Res. Interdiscip. Perspect. 2020;7 doi: 10.1016/j.trip.2020.100216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. The World Bank In Bangladesh [WWW Document], 2021. World Bank. URL <https://www.worldbank.org/en/country/bangladesh/overview#1>.
  53. Lytras Theodore, Dellis George, Flountzi Anastasia, Hatzianastasiou Sophia, Nikolopoulou Georgia, Tsekou Katerina, Diamantis Zafiris, Stathopoulou Grigoria, Togka Marianthi, Gerolymatos Gerasimos, Rigakos George, Spiridon S., Tsiodras Sapounas. High prevalence of SARS-CoV-2 infection in repatriation flights to Greece from three European countries. J. Travel Med. 2020;27 doi: 10.1093/jtm/taaa054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Mouchtouri Varvara A, Dirksen-Fischer Martin, Hadjichristodoulou Christos. Health measures to travellers and cruise ships in response to COVID-19. J. Travel Med. 2020;27(3) doi: 10.1093/jtm/taaa043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Vrieze S.I. Model selection and psychological theory: a discussion of the differences between the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) Psychol. Methods. 2012;17:228–243. doi: 10.1037/a0027127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. WHO, 2020. Last days., WHO Bangladesh COVID-19 Morbidity and Mortality Weekly Update.
  57. Worldometer, 2021. Reported Cases and Deaths by Country or Territory [WWW Document]. URL <https://www.worldometers.info/coronavirus/#countries>.
  58. Zafri N.M., Khan A., Jamal S., Alam B.M. Impacts of the COVID-19 Pandemic on Active Travel Mode Choice in Bangladesh: a Study from the Perspective of Sustainability and New Normal Situation. Sustainability. 2021;13:6975. doi: 10.3390/su13126975. [DOI] [Google Scholar]
  59. Zhang Yuzhao, Wang Jianqiang, Cai Wenjuan. Passengers’ demand characteristics experimental analysis of EMU trains with sleeping cars in Northwest China. Sustainability. 2019;11(19):5338. doi: 10.3390/su11195338. [DOI] [Google Scholar]

Articles from Case Studies on Transport Policy are provided here courtesy of Elsevier

RESOURCES