Food Security 
https://doi.org/10.1007/s12571-023-01417-w
ORIGINAL PAPER
Using free Wi‑Fi to assess impact of COVID‑19 pandemic on traditional 
wet markets in Hanoi
Louis Reymondin1,4  · Thibaud Vantalon1  · Huong Thi Mai Pham1  · Hieu Trung Le3 · Tuyen Thi Thanh Huynh1 · 
Ricardo Hernandez1  · Brice Even1  · Thang Cong Nguyen3 · Trong Van Phan1  · Kien Tri Nguyen1  · 
Christophe Béné2 
Received: 1 February 2023 / Accepted: 31 October 2023 
© The Author(s) 2023
Abstract
Traditional wet markets are the main source of fresh food and the largest source of daily nutrient intake for citizens of Hanoi. 
Nevertheless, due to the lack of traceability and sales registration systems, food flows within these markets remain largely 
invisible. This makes it challenging to assess the impact of shocks, such as pandemics, on these markets. In this paper, we 
characterize the impact of COVID-19 by analyzing data from 25 Wi-Fi access points installed in five formally established 
wet markets. The study timeframe covers a pre-pandemic period from July 2019 to the end of the initial stage of the pandemic 
in November 2020. While providing free Internet access, data were continuously collected about devices in close vicinity 
to the access points. Based on this information, we tested five hypotheses about the number, frequency, time, and duration 
of visits to the markets as well as changes in inter-market activities. The results show that during the shock (February to 
mid-April 2020) and aftershock (mid-April to July 2020) periods, market actors significantly decreased the total number of 
market visits (-26% P < 0.001) and the frequency of market visits (up to -47% for very frequent market users, P < 0.001). The 
number of inter-market visits dropped sharply during the shock period (66% ± 17% of the baseline level, P < 0.001), and the 
peak time for market shopping shifted significantly by 90 min later in the day, P < 0.001. No change was observed in visit 
duration. Several factors identified in existing literature as affecting consumer behaviors provide possible explanations for 
the changes observed. We present a set of recommendations to limit the negative impact of the pandemic in terms of food 
security and livelihoods in Hanoi and to mitigate consumers’ negative perception of wet markets in terms of food safety.
Keywords Wet markets · COVID-19 · Cellphone metadata · Food system · Food safety · Vietnam
1 Introduction that had major positive impacts on the overall quality of 
life for Vietnamese citizens, including an improvement in 
Food demand patterns have changed dramatically in Viet- diet quality (Dien et al., 2004; Dollar et al., 2004; Thang & 
nam over the past 30 years. The “Đổi Mới” economic Popkin, 2004). This increase in quantity, quality, and diversi-
reforms of 1986 began a period of swift economic growth fication of food types relied heavily on formal and informal 
wet markets. Currently, these food outlets continue to play a 
*  Louis Reymondin key role in food retail in both urban and rural areas; regard-
 l.reymondin@cgiar.org less of a household’s income status and geographic loca-
1 International Center for Tropical Agriculture – Hanoi tion, traditional wet markets account for the largest share of 
Hub, Agricultural Genetics Institute, Pham Van Dong, household food expenditure (Raneri et al., 2019).
Bac Tu Liem, Hanoi, Vietnam Despite the development of modern retail outlets in urban 
2 International Center for Tropical Agriculture (CIAT), Km 17 and peri-urban areas (Vo & Smith, 2017), traditional wet 
Recta Cali-Palmira, Cali, Colombia markets remain the main source of fresh food (i.e., meat, 
3 Agriculture, Forestry, and Fishery Department – General fish and seafood, eggs, fruits and vegetables) in both rural 
Statistics Office of Vietnam, Nguyen Chi Thanh, Hanoi, and urban areas. They offer several advantages to consumers, 
Vietnam such as easy access, freshness of food, and social interac-
4 Bioversity International, Parc scientifique Agropolis II, 1990 tions associated with trust and food origins (Wertheim-Heck 
Bd de la Lironde, Montpellier, France
Vol.:(012 3456789)
 L. Reymondin et al.
et al., 2019). In Hanoi, wet markets still account for about smartphone is not always linked to a single person, as multi-
90% of vegetable sales, providing significant daily nutrient ple individuals may use the same device. This makes it dif-
intake, including 56% of energy, 70% of protein, 80% of vita- ficult to accurately attribute specific behaviors or patterns to 
mins A and C, and approximately 70% of calcium, iron and an individual (Arai et al., 2016). Finally, while mobile phone 
zinc (Raneri & Wertheim-Heck, 2019). In contrast, modern metadata can reveal valuable information about behavioral 
retail outlets, including supermarkets and chain convenience patterns, it falls short of providing insight into the underly-
stores, offer mostly non-food and ultra-processed food items ing reasons behind these patterns and how they change over 
(Harris et al., 2020). Hence these food outlets are not yet an time. Researchers must use complementary methods or rely 
important food source (Raneri et al., 2019), even for urban on the existing literature to understand the motivations and 
consumers. Therefore, despite recent development affecting drivers behind the observed behaviors.
the food retail sector, wet markets remain crucial for ensuring The full magnitude of the impact of the COVID-19 pan-
the food security of millions of consumers in Hanoi. demic on food systems and food environments is yet to be 
Given the importance of wet markets in Hanoi and in Viet- determined (Béné et al., 2021). During the first year of 
nam in general, an understanding of food flows and trades the pandemic, Vietnam’s very quick response was praised 
within wet markets is critical for guiding policy makers in tak- internationally. During 2020, only 1,456 cases were identi-
ing informed decisions to ensure food security for Vietnam- fied, and all were quickly quarantined. The country suffered 
ese citizens. Nonetheless, food flows in wet markets remain only 35 deaths due to the virus (World Health Organiza-
largely invisible in official statistics due to the absence of tion, 2020). Vietnam’s policy response to the COVID-19 
traceability systems and the use of outdated sales registra- pandemic was characterized by an immediate reaction, a 
tion systems. In fact, vendors and market management boards high level of prioritization, and highly coordinated political 
rely on manual records to track daily product sales, rather mobilization (Đỗ et al., 2020). Vietnamese authorities had 
than following a standardized system. This recording method already issued seven COVID-19 related official documents 
has demonstrated vulnerabilities, including errors in record- before January 23, 2020 (Le et al., 2021), when the first 
ing and the potential loss of records. This makes it challeng- two cases of COVID-19 were declared on Vietnamese ter-
ing to study the impact of economic shocks on such markets. ritory. By the end of July 2020, 959 policy documents had 
Consequently, alternative types of data are needed to better been issued by 33 different public agencies, illustrating the 
understand trends and consumer behavior in these markets. prompt and proactive response of the Vietnamese govern-
In recent years, different types of cellphone metadata, such ment (Le et al., 2021).
as location, date, and time a user makes a phone call, its dura- None of the policies issued within the first year of the 
tion, and its status (incoming, outgoing, or canceled), have pandemic specifically targeted wet markets. Even during 
been used to better understand human mobility and behavior the “Hanoi lock-down” from March 31 to April 23, 2020 
(Ghahramani et al., 2019; Williams et al., 2015). This meta- (Directive No. 16/CT-TTg), which consisted of stricter 
data has been explored to analyze the behavior of vulnerable social distancing measures and included the closure of 
populations for whom there is a lack of data, such as refugees “non-essential” businesses, food businesses (including 
and migrants (Hong et al., 2019; Kutscher & Kreß, 2018; supermarkets, convenience stores and wet markets) were 
Pastor-Escuredo et al., 2019). Traditionally, analysis of the exempt from closure (Directive No. 05/CT-UBND). On the 
behavior of such populations is heavily reliant on census and other hand, informal markets and street vending in general 
survey datasets and on interviews (Hong et al., 2019). Col- were banned throughout the city following a decision by 
lection of such datasets is expensive and time consuming. Hanoi People’s Committee on March 13, 2020 (Official 
Therefore, analysts consider cellphone data as efficient data- Letter 871/UBND-KT). Such restrictions together with 
sets, complementing traditional methods (Pastor-Escuredo reduced urban-rural mobility have had a strong negative 
et al., 2019). This interest in cellphone metadata analysis has impact on street vendor livelihoods (Turner et al., 2021).
exploded with the surge of the COVID-19 pandemic (Kishore To better characterize the behavioral changes of market 
et al., 2020). For instance, Google has made smartphone actors and understand the impact of COVID-19 on Hanoi’s 
geolocation data, aggregated at global and national levels, wet markets during the first stage of the pandemic, we ana-
publicly available to help policy makers to better understand lyzed and interpreted mobile device tracking data. From 
the impact of the pandemic (Google LLC, 2022a). July 2019 to November 2020, 25 Wi-Fi access points were 
Using mobile phone metadata for research has certain formally established in five wet markets under the direct 
limitations. First, sampling bias is a significant concern. supervision of the district authorities in Hanoi. The access 
Analysis of mobile phone metadata is limited to individu- points provided free internet access to market users while 
als who own smartphones, which inherently introduces bias collecting data on the number of Wi-Fi-enabled devices pre-
towards the population that can afford such devices (Lazer & sent in the market at any given time. The data allowed us to 
Radford, 2017). Second, ownership bias poses a challenge. A understand the flows of food and people by characterizing 
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
the behavior of market actors and the frequency of their 2: 200–399 stores and class 3: < 200 stores. As shown in 
visits. We hypothesized that market actors changed their Table 1 and Fig. 1, five traditional markets, including whole-
food purchasing behavior during the pandemic. Behavio- sale and retail, were selected. To ensure representativeness 
ral changes could be due to: (a) the swift response of the of market types, two retail markets were selected in each 
Vietnamese government, which resulted in restrictions that district, one under Class 2 and the other under Class 3. As 
sharply reduced citizen mobility (Le et al., 2021); (b) fear shown in Table 1, TO market is considered as a wholesale 
of social interaction that could lead to contagion (Munster market at night and a retail market during the day. As there 
et al., 2018; Spiehler & Fischer, 2021); (c) fear about safety is no wholesale market in the urban district (Cầu Giấy), a 
in wet markets triggered by reports linking the emergence wholesale market in a nearby urban district (Bắc Từ Liêm) 
of COVID-19 to a wet market in Wuhan, Hubei Province, was selected for the study, as it is known to provide food 
China (Cohen, 2020; Li et al., 2020; Mizumoto et al., 2020; products to most of the markets in the surrounding area. The 
Spiehler & Fischer, 2021); and (d) reduced income (Kang five selected markets are all under direct management of the 
et al., 2021). We explored consumer behavior by testing the district authorities.
following hypotheses:
2.2  Data collection
1. The overall number of visits to wet markets dropped 
significantly during the first phase of the pandemic. To monitor the behavior of different market actors, we set-
2. The frequency with which consumers purchased food up Wi-Fi routers and signal amplifiers to register and keep 
from wet markets decreased during the first phase of the track of each “Active Scanning” request received. Active 
pandemic. scanning is a probe request regularly sent by any device 
3. Consumers spent less time at wet markets. with Wi-Fi connection enabled to detect available Wi-Fi 
4. Consumers changed the time of day at which they visited hotspots. Devices must share their Media Access Control 
wet markets (MAC) address, a unique device identifier, when sending 
5. Retailers reduced their visits to wholesale markets, lead- an active scanning request. This allows us to register all 
ing to a significant decrease in inter-market visits. the devices present in the market, whether or not they are 
connected to the Wi-Fi network, and their movement in 
Given the timeline of the initial project, we could deter- space and time. For instance, a simple metric that can be 
mine a pre-shock baseline (based on data from the second extracted from this system is to count the average num-
half of 2019). Therefore, our results focus on the first six ber of individual phones passing near the Wi-Fi access 
months of the outbreak, covering only the “first wave” of points. Indeed, by counting the number of individual MAC 
the COVID-19 pandemic in Vietnam addresses within a certain range of the access points, we 
can get an indicator of the number of people present in the 
market at 2-min intervals. To ensure anonymity of all the 
2  Materials and methods devices and their respective owners, all the collected MAC 
addresses were anonymized through a Secure Hash Algo-
2.1  Study area rithms (SHA-3 256) hashing function (Dworkin, 2015) 
together with a randomly generated 4096-bit salt value. 
Two Hanoi districts, Cầu Giấy and Đông Anh, one urban The hashed MAC addresses ensure anonymity, while the 
and one peri-urban (Haan et al., 2020; Huynh et al., 2021) hashing is consistent across all markets and during the 
were selected for this study. The traditional market system whole period of the analysis. We can therefore ensure that 
in Vietnam is usually divided into three categories based a given MAC address found in two different markets or 
on the number of food stores: class 1: > 400 stores, class periods will correspond to the same device. Combining 
Table 1  Characteristics of wet District ID Market name Type Number of Number of 
markets included in this study stores food stores
Cầu Giấy DX Đồng Xa Retail 440 165
NT Nghĩa Tân Retail 186 85
Đông Anh TT Trung Tâm Retail 121 72
TO Tó Retail during the day, 280 52
wholesale at night
Bắc Từ Liêm MK Minh Khai Wholesale 900 900
1 3
 L. Reymondin et al.
Fig. 1  Market locations in Hanoi, with background from Open Street Map (OpenStreetMap contributors, 2021), showing the five markets in the 
study: Đồng Xa (DX), Nghĩa Tân (NT), Trung Tâm (TT), Tó (TO) and Minh Khai (MK)
the original data with a large salt value is a common step aggregated among four periods with the exact same num-
in the application of a one-way cryptographic hash func- ber of days, each starting on a Monday and ending on a 
tion, such as SHA-3 (Gilchrist, 2003). In this case, it is Sunday. Each period is exactly 11 weeks. As described in 
used as a secret key without which it is nearly impossible Table 2, each of these periods represents a key step in the 
to retrieve the anonymized MAC address corresponding impact analysis.
to a given original MAC address, even if both the database The official number of daily COVID-19 cases was 
and the original MAC address are known. accessed and compiled for the same four periods from the 
The goal of this study was to assess the behavior World Health Organization (2020). Finally, a set of key 
changes of market actors before and after the first wave measures implemented by the Vietnamese government to 
of the COVID-19 outbreak in Vietnam. Data available mitigate the propagation of the virus was compiled from 
between July 2019 and July 2020 were compiled and multiple online sources and peer-reviewed literature.
Table 2  Definition of the Period Start date End date Description
periods studied for the impact 
analysis Baseline 2019/07/29 2019/10/13 reference data representing business-as-usual behavior
Pre-shock 2019/10/14 2019/12/29 2019/12/29: the state of the markets immediately before the first 
cases in the COVID-19 outbreak in Vietnam
Shock 2020/02/03 2020/04/19 2020/02/03 – 2020/04/19: the state of the markets during the first 
wave of the COVID-19 outbreak
Aftershock 2020/04/20 2020/07/05 the state of the markets immediately after the first COVID-19 wave
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
2.3  Data filtering 2.4  Data aggregation
To ensure comparability and to reduce noise in the data, a 2.4.1  Dataset 1: Aggregation at device level
series of steps were implemented to filter the list of device 
observations that were collected. First, the period just For each device remaining after data filtering, the following 
before, during and immediately after Tết (lunar new year metrics were computed:
celebration, January 2020) was removed from the analysis, 
as it is a period with a markedly different pattern to the rest • An anonymized MAC address
of the year, as seen in Fig. 2. Second, all randomized MAC • The name of the market where the phone was seen
addresses were removed from the database, to ensure high • The median time at which the user is first seen in the 
data quality. To increase security, modern devices generate markets (in minutes starting at 0 from midnight)
a random MAC address for each access point with which • The median time at which the user is last seen in the 
they communicate (Apple Inc., 2022; Cisco Systems Inc., markets (in minutes starting at 0 from midnight)
2021; Google LLC, 2022b). The removal of randomized • The average number of visits made to the market daily. A 
MAC addresses is implemented directly within the Wi-Fi period of at least 2 h must elapse between two consecu-
routers, and we are therefore not able to report the number tive observations of the same user in the market for the 
of observations that were discarded based on this criterion. observations to be counted as different visits.
The third step was to remove all observations with a signal • The average time spent in the market per visit
strength (RSSI) of less than 30. Based on tests in situ, this • The average number of visits per week
threshold includes only devices that are up to about 5 m • The total number of days that a user was seen in the market
away from the access points and reduces the risk of regis- • Total time spent in the market
tering devices outside of the markets. Finally, all devices 
that were seen only once and for less than 10 min during 2.4.2  Dataset 2: Aggregation in time
the whole study period were removed from the datasets. 
This was to remove passersby who were only very weakly Based on the data aggregated at the user level, two additional 
related to the markets. levels of aggregation were implemented to assess the impact 
Fig. 2  Time series showing the number of devices seen aggregated ing a nationwide isolation of 15 days from April 1 to 15, 2020; 4) No 
daily and across all markets during the study period. Key events and new community-transmitted cases were registered in the country from 
directives are highlighted as follows: 1) Prime Minister issued a first this moment until the end of the study, and all imported cases were 
directive against COVID-19 (Official Letter No. 121/CD-TTg dated kept under strict quarantine on arrival; and 5) Directive No. 16/CT-
January 23, 2020); 2) Prime Minister established a national steering TTg was issued on April 23, 2020, ending national social-distancing 
committee for prevention and control of COVID-19 (Decision 170/ measures
QD-TTg); 3) Prime Minister issued directive No. 16/CT-TTg order-
1 3
 L. Reymondin et al.
of COVID-19 on market actors’ behavior. The first level is level chosen for this study to reject the null hypothesis, and 
designed to assess whether an impact on the number of visits thus validate the original hypothesis, was set to P < 0.01.
to the markets can be observed. We compiled visits into a 
daily time series within each market for the whole duration Hypothesis 1: The overall number of visits to wet markets 
of the study. dropped significantly during the first phase of the pandemic.
2.4.3  Dataset 3: Aggregation by market user type To explore the existence of a drop in wet market visits 
during the shock period, we tested the null hypothesis stipu-
The third level of aggregation is designed to better under- lating that no change was observed amongst the four periods 
stand which categories of market users were most impacted studied. We computed the daily average number of visits, 
by the COVID-19 outbreak and the actions taken to mitigate based on the time series from dataset 2, over the four periods 
propagation. This was achieved in two steps. First, the devices of the study and compared each period against the baseline. 
dataset was further filtered to only include devices that were A Student’s t-test was used to verify the significance level 
seen in at least two different periods to increase comparability of differences in frequency of visits.
amongst periods. Second, each device was categorized into 
three types of behavior given the following rules: Hypothesis 2: The frequency with which consumers went 
to purchase food at wet markets decreased during the first 
Passerby d evices seen between 2 and 11 times over the phase of the pandemic.
11 weeks (strictly less than once a week).
Frequent d evices seen between 11 and 44 times over To better understand the impact of the first phase of the 
the 11 weeks (once to strictly less than four pandemic on the frequency of visits, we normalized data-
times per week). set 4 with respect to the baseline for the pre-shock, shock, 
Very frequent d evices seen between 44 and 77 times over and aftershock data. We then computed descriptive statistics 
the 11 weeks (between four and seven times metrics for each period to explore the differences within 
per week). each category of market user with respect to the baseline.
Additionally, for each period studied, users were assigned Hypothesis 3: Consumers spent less time in wet markets.
to the following categories: “new” if the device had not been 
seen during previous periods, “out” for devices that were not To identify a statistically significant difference in the 
seen again in the following period, and “break” for devices duration of time that consumers spent in the market, we 
that skipped a period but were seen again during the study. tested the null hypothesis that there was no significant dif-
These data were produced for each period to better under- ference amongst the periods studied. To do so, we used the 
stand the dynamics of change across time for each category. average duration spent in the market per visit from dataset 
1. The first step of the analysis was to define consumers as 
2.4.4  Dataset 4: Aggregation at inter‑market level owners of devices that were seen in retail markets for strictly 
less than one hour per visit. We then compared the distri-
To study the connectivity of the different markets, all the bution of durations in each period with the baseline using 
devices that were detected in pairs of markets were identified a two-sample Kolmogorov-Smirnov test. As an alternative 
and compiled into a weighted network graph. Each vertex hypothesis for the Kolmogorov-Smirnov test, we defined 
of the graph represents a market, while the weighted edges the visits that happened during the periods following the 
represent the strength of the connection between each pair baseline as being shorter.
of markets, measured by the total number of devices seen 
in both markets during each of the four key periods studied. Hypothesis 4: Consumers changed the time of day at which 
These graphs are particularly useful to visualize the level of they visited wet markets.
connectivity between each pair of markets.
To identify a shift in consumer behavior in terms of 
2.5  Hypothesis testing the time of day at which markets were visited, we used 
the median time at which devices were first seen in the 
All the hypotheses presented hereafter were tested following markets from dataset 1. As previously, we first defined 
the standard method of assessing and rejecting the alterna- consumers as owners of devices that were seen in retail 
tive null hypothesis under a certain significance level. The markets for strictly less than one hour per visit. We then 
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
computed the distribution of times at which devices were Figure 2 shows the time series of the number of devices 
observed in the markets separately for the morning and seen daily, together with the official number of COVID-
afternoon and performed a peak analysis to identify the 19 cases published by the Vietnamese authorities. The 
two busiest times of the day. Finally, we tested the null first COVID-19 cases were registered during Tết festivi-
hypothesis that no change was observed with respect to ties. A series of measures was taken very quickly, and, for 
the baseline for each subsequent period using a two-sided instance, schools were asked to remain closed after the 
Kolmogorov–Smirnov test. Tết break for a duration of three months. As cases were 
increasing in the country, the Hanoian authorities man-
Hypothesis 5: Retailers reduced their visits to wholesale dated a partial lock-down between April 5 and 23, 2020. 
markets, leading to a significant decrease in inter-market Finally, after the lock-down and most other restrictions 
visits. were lifted, no additional community-transmitted cases 
were registered in the city; subsequent cases were those 
To assess if there was a significant decrease in inter-mar- imported from foreign countries and were kept under strict 
ket visits, we tested the null hypothesis that no difference can quarantine.
be observed between the baseline and subsequent periods. As shown in Fig. 2, the number of devices seen daily in 
To do so, we used dataset 4. First, we normalized the con- the markets was on average lower following Tết celebrations 
nectivity strengths of the pre-shock, shock, and aftershock and for the whole remaining period analyzed. The impact 
periods with respect to the baseline. We then computed the analysis shows that the average number of devices seen in 
average normalized weights and performed a Student’s t-test the markets dropped from 1, 128 ± 164 during the baseline 
to assess if the null hypothesis could be rejected. to 836 ± 128 during the shock period and 787 ± 176 dur-
ing the aftershock period (Table 4). On the other hand, the 
pre-shock period is relatively similar to the baseline, with 
3  Results devices seen daily on average. For both the shock and after-
shock periods, the Student’s t-test found that the null hypoth-
3.1  Data filtering esis, stipulating that there are no differences between these 
periods and the baseline, should be rejected with P < 0.001 
Scanning the markets every two minutes over all the time in both cases. There is therefore a statistically significant 
periods studied yielded a total of 173,668,702 observa- decrease in the number of visits to the wet markets between 
tions, where each observation is a specific device being the baseline and the shock and aftershock periods. On the 
seen at a specific date and time (Table 3). After filtering other hand, P = 0.6 > 0.01 was observed when comparing 
out all observations with signal strengths too weak to be the pre-shock period to the baseline, resulting in the con-
included, 13,238,809 usable observations remained. After clusion that there is no statistically significant difference 
aggregating these observations, a total of 656,789 differ- between those two periods. As shown in Table 4, a statisti-
ent devices were observed in the markets between July 29, cally significant difference between the baseline and pre-
2019 and July 5, 2020. shock periods was identified only in TO market, which reg-
istered a 16% decrease with respect to the baseline during the 
pre-shock period. Overall, NT market was the most sharply 
3.2  Hypothesis testing impacted with a drop of 48% observed during the shock 
period and 49% during the aftershock period. TT, DX and 
Hypothesis 1: The overall number of visits to wet markets TO markets saw a decrease in frequency of visits of between 
dropped significantly during the first phase of the pandemic. 12% and 32% during the shock period and a slight increase 
Table 3  Total observations Period Total observations After signal strength Number of unique 
before and after data filtering filtering devices after 
aggregation
Baseline 51,419,597 3,930,381 237,886
Pre-shock 46,571,340 3,297,988 230,785
Shock 33,156,224 2,343,634 192,985
Aftershock 42,521,541 3,666,806 176,294
Total 173,668,702 13,238,809 656,789
1 3
 L. Reymondin et al.
Table 4  Number of devices seen in each market and relative differ- In summary, we conclude that the overall number of visits 
ence with respect to the baseline to wet markets dropped significantly during the first phase 
MARKET PERIOD MEAN  𝚫[%] P VALUE of the pandemic, confirming our first hypothesis.
NT Baseline 219 56 0 -
Hypothesis 2: The frequency with which consumers went 
Pre-shock 212 28 -3 
0.35 to purchase food at wet markets decreased during the first 
Shock 113 57 -48 
< 0.001 phase of the pandemic.
Aftershock 111 24 -49 
< 0.001
TT Baseline 144 44 0 - Figure 3 and Table 5 show the changes in actor behav-
Pre-shock 150 16 4 
0.27 ior during each of the periods studied. Across all the periods 
Shock 128 24 -12 
0.002 studied, passersby represent 89.8% ± 1.2% of the total popula-
Aftershock 133 35 -8 
0.03 tion, frequent visitors 7.6% ± 0.7% and very frequent visitors 
DX Baseline 212 64 0 -
2.5% ± 0.5% . Although very frequent visitors make up the 
Pre-shock 197 24 -7 
0.52 smallest share of the whole population, it is this category that 
Shock 145 33 -32 
< 0.001 registered the sharpest changes during the shock and after-
Aftershock 178 32 -16 
< .001 shock periods. With respect to the baseline, very frequent visi-
TO Baseline 258 60 0 - tors decreased by 37% during the shock period and 47% during 
Pre-shock 218 54 -16 
< 0.001 the aftershock period. On the other hand, passersby decreased 
Shock 187 36 -27 
< 0.001 by 17% and 25% over the same periods. Finally, frequent visi-
Aftershock 208 59 -20 
< 0.001 tors decreased by 33% during the shock period and remained 
MK Baseline 294 81 0 - at this level during the aftershock period.
Pre-shock 305 114 4 
0.48 As illustrated in Fig. 3, of the 47% reduction in very fre-
Shock 263 127 -11 
0.04 quent actors seen during the shock and aftershock period 
Aftershock 157 160 -47 
< 0.001 compared to the baseline, 58.6% of actors reduced the fre-
ALL Baseline 1,128 164 0 - quency of their visits from 4–7 times per week to 1–4 
Pre-shock 1,083 132 -4 
0.6 visits per week. Another 7.4% reduced their visits to less 
Shock 836 128 -26 
< 0.001 than once per week. Finally, 34% were not seen again dur-
Aftershock 787 176 -30 
< 0.001 ing the study. This represents 1955 devices that were seen 
s denotes results from a one-tailed Student’s t-test, used when a at least four times per week during the baseline and pre-
decrease was expected (shock, aftershock periods) shock periods that were not seen in subsequent periods. 
d denotes results from a two-tailed Student’s t-test, used when there Indeed, as shown in Fig. 3, across all categories and for 
was no a priori assumption about the direction of change (or no the whole duration of the study, a share of devices left the 
change was expected, e.g., pre-shock) markets, never to be seen again. The proportion of devices 
categorized as very frequent that left the study area during 
the shock and aftershock periods more than doubled when 
in frequency of visits during the aftershock period, with val- compared to the baseline and pre-shock periods.
ues ranging between 8% and 20% below that of the baseline In summary, the significant drop in overall number of 
period. Finally, the wholesale market (MK) registered one visits to wet markets is characterized as a sharp reduction in 
of the smallest decreases during the shock period (1 1% ) but market visit frequency, confirming our second hypothesis.
this was followed by the sharpest decrease recorded during 
the aftershock period of 47% compared to the baseline. This Hypothesis 3: Consumers spent less time in wet markets.
finding should be interpreted in light of the many power 
outages that occurred in MK market during the aftershock As shown in Fig. 4, most visits registered with the system 
period. Although measures were taken to mitigate the impact in place lasted between 2 and 20 min. The average duration 
of missing data, as explained in the Materials and Meth- observed during the baseline was 10.8 min with a standard 
ods section above, a lack of data values occurred very fre- deviation of 8.6 min (Table 6).
quently in the aftershock period in this market and thus the It is important to note that the filtering used to select 
frequency of visits rate may be underestimated. observations to be included in this study was very restrictive. 
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
Fig. 3  Change in behavior of market actors in terms of the frequency period, and “break” for devices that skipped a period but were seen 
of market visits during each study period. The categories of market again during the study. Node values represent the number of devices 
actors are: “new” if the device had not been seen during previous seen for each period and each category (rounded to three significant 
periods, “out” for devices that were not seen again in the following figures)
Only observations with very high signal strength were con- in term of visit duration cannot be rejected for any of the 
sidered. This was to ensure the selection of devices that were periods analyzed, given the significance level chosen for this 
seen within the food area of the wet markets and not in the study. Hypothesis 3 is therefore not confirmed based on the 
surrounding shops nor on the streets around the markets. data gathered for this study.
The durations presented here are therefore not representative 
of the overall time spent in the markets but rather the time Hypothesis 4: Consumers changed the time of day at which 
spent within close range of the Wi-Fi access points scattered they visited wet markets.
around the food area of the wet markets.
As shown in Table 6, the average visit duration was very As shown in Fig. 5, there are two peak times during the 
consistent during the whole study. Indeed, the two-sample day corresponding to rush hours when people are traveling 
Kolmogorov-Smirnov test results show that the null hypoth- to work in the morning and returning home in the after-
esis stipulating that consumers did not change their habits noon. To compare the shock period to the baseline, the day 
was split into two periods: the morning, 04:00–13:30; and 
the afternoon, 13:30–20:00.
Table 5  Relative change in market visit frequency by different actor As shown in Fig. 5 and Table 7, a clear shift in the 
categories with respect to the baseline period peak time for visits shorter than 60 min can be observed 
Passerby Frequent Very frequent during the shock period when compared to the baseline. 
Indeed, this effect was particularly strong during the morn-
Pre-shock -3% -2% 5% ing period, as the peak time was.
Shock -17% -33% -37% The two-sample Kolmogorov-Smirnov test results show 
Aftershock -25% -33% -47% that the null hypothesis stipulating that both the morning 
1 3
 L. Reymondin et al.
Fig. 4  Distribution of visit 
duration in minutes for all retail 
markets during each period 
studied. Density is the percent-
age of total observations at a 
given duration
and afternoon samples come from the same distribution as As shown in Fig. 6, the average level of inter-market 
the baseline can be rejected with P < 0.001. This confirms visits during the pre-shock period was about 106% ± 20% 
hypothesis 4, showing that consumers significantly changed that observed during the baseline. This level dropped 
the time of day at which they visited wet markets during the to 66% ± 17% during the shock period and increased to 
shock period. However, this effect was not observed during 75% ± 33% during the aftershock period. Although the 
the aftershock period. aftershock period shows a slight increase in comparison 
with the shock period, it is also the period with the great-
Hypothesis 5: Retailers reduced their visits to wholesale est heterogeneity. As shown in Fig. 6, all market pairs saw 
markets, leading to a significant decrease in inter-market a decrease in inter-market visits during the shock period. 
visits. Some pairs of markets saw levels increase during the after-
shock or return to values close to that of the baseline (e.g., 
As seen in Fig. 6, although all market pairs registered MK-DX, MK-TT). Some market pairs retained the same 
inter-market visits, there are two clear clusters of markets level in the aftershock period as that observed during the 
strongly linked to each other. The first cluster encompasses shock period (e.g., TT-TO) and others registered a further 
both peri-urban markets (TT, TO) while the second cluster decrease in inter-market visits during the aftershock period 
includes both urban markets together with the wholesale (e.g., NG-TO, NG-DX).
market (DX, NG, MK). As presented in Table 8, the result of the one-tailed Stu-
dent’s t-test shows that the null hypothesis, stipulating that 
similar numbers of inter-market visits were registered dur-
ing the shock and pre-shock periods, can be rejected with 
Table 6  Average visit duration and standard deviation, in minutes, 
for all retail markets studied. P values are the results of one-sided P < 0.001 . One can thus conclude that during the shock 
Kolmogorov-Smirnov tests to either confirm or reject the null hypoth- period, the strength of inter-market connectivity was sig-
esis that no change was observed between visit durations during the nificantly lower than during the pre-shock period. Finally, 
baseline period and the periods studied the null hypothesis of the aftershock being similar to the 
Period Average visit  Standard deviation P value pre-shock period cannot be rejected with a statistically sig-
duration (minutes) (minutes) nificant level ( P = 0.023 > 0.01) . In summary, although the 
connectivity levels observed during the aftershock period are 
Baseline 10.8 8.6 on average 75% lower than the baseline, the high variance 
Pre-shock 10.7 8.3 0.04 observed in the data does not permit the identification of 
Shock 10.0 7.5 0.98 a statistically significant difference between pre-shock and 
Aftershock 10.1 8.3 0.99 after-shock levels.
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
Fig. 5  Intra-day distribution of visits shorter than 60 min for each afternoon peak times observed during the baseline, and solid vertical 
period studied. Density is the percentage of total observations at lines show the morning and afternoon peak times observed during the 
a given duration. Dashed vertical lines represent the morning and shock period
4  Discussion pre-shock periods to 09:30 during the shock period. This 
effect was only observed during the shock period and disap-
In summary, our results show that a sharp decrease was peared during the aftershock. A shift was also observed dur-
observed in the number of people visiting traditional markets ing the afternoon peak time, but it was of 30 min only. Such 
after the first COVID-19 cases were recorded (-26%) as well a shift confirms hypothesis 4. On the other hand, we could 
as a statistically significant drop in the frequency with which not conclude that consumers changed the duration of time 
people visited the markets, confirming hypotheses 1 and 2. they spent in the market during the shock and aftershock 
The category of devices for which the strongest decrease periods based on the data collected for this study, thus reject-
was observed were those that were previously seen on more ing hypothesis 3. Finally, the connectivity between markets 
than four days per week (-37%). Some of these devices were was also shown to drop significantly during the same period 
seen with less frequency but there is also a large proportion ( 66% ± 17% of the level observed during the baseline), con-
that were not observed again during the study. Addition- firming hypothesis 5.
ally, we observed that the daily peak time of visits to the These findings are in line with similar sharp impacts 
retail markets shifted from 08:00 during the baseline and observed in Vietnam and other South-East Asian countries. 
Indeed, the COVID-19 pandemic and its related restric-
tions had devastating effects on food markets in Vietnam; 
food supply chains were compromised, informal wet mar-
Table 7  Morning and afternoon peak times of visits shorter than 60 kets were closed, and poverty and food insecurity worsened 
minutes for each period studied (Kang et al., 2021; Nguyen et al., 2021; van Melik et al., 
Period Morning peak time Afternoon peak time 2021). The pandemic severely affected food environments, 
access to and availability of fresh food, and reduced house-
Baseline 08:03 17:13 hold incomes (Nguyen et al., 2021). In urban areas, the effect 
Pre-shock 08:00 17:04 was magnified, as there is evidence of higher job losses and 
Shock 09:32 17:41 reduced income (76.5%), and a larger reduction (31%) in 
Aftershock 08:15 17:14 food expenditures (Kang et al., 2021).
1 3
 L. Reymondin et al.
Fig. 6  Network of inter-market 
links, showing the number of 
visits seen in each pair of mar-
kets as registered during each 
period studied. On each link, 
values show the total number 
of devices seen in both markets 
during the baseline together 
with a chart showing how this 
figure changed during the fol-
lowing periods with respect to 
the baseline. The width of the 
line between any two markets 
represents the volume of obser-
vations where devices were 
recorded at both locations
Although Hanoi People’s Committee requested district by: (i) a shift from wet markets to other outlets and online 
and ward authorities to eliminate informal wet markets and shopping as a reaction to the perceived risks regarding wet 
street vending activities (Official Letter 871/UBND-KT), markets and to mobility restriction measures, (ii) reduced 
none of the policies issued within the period of this study visit frequency combined with purchases of bigger quantities 
specifically targeted formal wet markets. Indeed, even dur- of fresh food as a reaction to mobility restriction measures 
ing the period with the strictest restrictions on mobility, from and public health messaging, (iii) a shift from fresh food 
March 31 to April 23, 2020 (Directive No. 16/CT-TTg), food purchases to purchases of products with longer shelf-life 
businesses were exempt from closure (Directive No. 05/CT- as a component of food storing strategies to cope with food 
UBND). Furthermore, during the period studied, a very low scarcity events or imposed home quarantines, (iv) loss of 
number of COVID-19 cases were observed in the country; income among vulnerable populations, and (v) a reduction 
as of August 31, 2020, only 509 cases were reported (with 0 in informal activities in the vicinity of the studied markets. 
deaths) (World Health Organization, 2020). Therefore, the The confirmation of hypothesis 4 shows a significant shift 
strong impact of the pandemic observed in this study cannot in the main peak time for shopping, which can be explained 
be explained by a single policy nor by the outbreak of the by (vi) strategies to cope with restrictions such as strict work 
virus itself. from home policies. Finally, the confirmation of hypothesis 
There are several factors affecting consumer behavior that 5 shows a significant decrease in inter-market visits. This 
are possible explanations for the changes observed through can be explained by the restrictions in mobility already men-
the confirmation of our initial hypotheses. The confirma- tioned as well as by (vii) an overall decrease in sales result-
tion of hypotheses 1 and 2 shows a sharp decrease in fre- ing in decreased exchanges between wholesale and retail 
quency of visits to wet markets, which can be explained markets. These topics are discussed in more depth below.
Table 8  Relative number Period Average inter-market visits as a Standard P value from Student t-test for 
of people traveling between proportion of the baseline deviation comparison with pre-shock 
markets with respect to the period
baseline period
Pre-shock 1.06 0.2 -
Shock 0.66 0.17  < 0.001
Aftershock 0.74 0.33 0.023
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
Perceived safety risk: The confirmation of hypotheses 1 measures resulting in mobility restrictions might have led to 
and 2 shows that at the very early stage of the pandemic, a shift in consumers’ practices.
market visits dropped significantly in terms of both the 
overall number of visits and the frequency of visits. In Viet- Diet shift toward more processed food: Although the phe-
nam, food safety has been a major concern of consumers nomenon is not consistent globally and across all popula-
for many decades but only influences food choices to some tions, it is clear that the COVID-19 crisis induced a diet shift. 
extent (Nguyen-Viet et al., 2017; Vandevijvere et al., 2019). Both favorable and unfavorable changes were reported, with 
Although consumers tend to see traditional wet markets as one trend being the increased consumption of unhealthy food, 
more prone to food safety issues, the transition from wet including processed and ultra-processed food (Buckland 
markets to modern food outlets such as supermarkets has et al., 2021; Deschasaux-Tanguy et al., 2021). For some pop-
been very slow in Vietnam compared to countries with a ulation categories, lock-down events and the related emo-
similar level of economic development (Wertheim-Heck tional distress caused by such mobility restrictions seem to 
et al., 2014). Supermarkets are perceived as inconvenient be associated with increased snacking, increased consump-
and time-consuming, and the safe foods they offer are tion of sweets, and decreased consumption of fresh food, 
considered more expensive and less fresh. Supermarkets especially fish and fruits (Deschasaux-Tanguy et al., 2021). 
mainly contribute to the consumption of ultra-processed In addition, we can hypothesize that consumers stockpiling 
foods (Wertheim-Heck et al., 2019). We suspect that the and hoarding food tend to choose longer shelf-life products 
COVID-19 outbreak and the related “fear for safety” gener- and less fresh food, for obvious perishability reasons. In 
ated among consumers might have accelerated the transi- Vietnam, food hoarding was reported in the news preceding 
tion process from “dirty” wet markets perceived as unsafe announcement of the lock-down (Cao Mai Phuong, 2021). 
to “clean” modern outlets perceived as safer (Ha et al., 2020; As already mentioned above, processed products and those 
The University of Adelaide, 2018). Furthermore, although with long shelf-life are commonly found in modern outlets 
Vietnamese consumers tend to distrust online platforms for such as supermarkets and online platforms. This process has 
food products (Kim Dang et al., 2018), a very sharp increase potentially strengthened the trend of consumers reducing 
in the use of such platforms has been observed since the their visits to traditional markets in favor of more corporate 
pandemic started, especially in urban areas (Nguyen et al., retail options. This factor might thus explain to some extent 
2021). Altogether, negative narratives around wet markets the drop in market visits observed with the confirmation of 
might have exacerbated a slow ongoing process in urban hypotheses 1 and 2.
areas of Vietnam that unveils a profound change in consumer 
perceptions of wet markets in favor of more modern outlets Income loss among vulnerable populations: In Vietnam and 
such as supermarkets and online platforms. This process other Southeast Asian countries, the COVID-19 pandemic 
might have led some consumers to avoid purchasing food and its related restrictions had a strong negative impact on 
products in wet markets and may account for a share of the poor and vulnerable populations (Boughton et al., 2021; 
sharp decrease in devices seen in the markets since the pan- Kang et al., 2021; Parks et al., 2020; Turner & Binh, 2021). 
demic started. Notably, the pandemic led to a substantial reduction in 
both income levels and household food expenditures, with 
Less frequent visits for larger purchases: The confirmation a disproportionately adverse impact on poor populations, 
of hypothesis 2 shows that there is a clear drop in the fre- particularly those living in urban areas (Kang et al., 2021; 
quency of visits to markets. Indeed, as shown in Table 5, Nguyen et al., 2021). It is important to highlight that poor 
the number of devices seen very frequently in the mar- urban communities predominantly depend on wet markets as 
kets dropped by 37% with respect to the baseline during their primary source for procuring food (Raneri et al., 2019). 
the shock period and 47% during the aftershock period. Consequently, it is reasonable to suggest that the observed 
Despite consumers’ negative perception and low level of decline in household food expenditures (Kang et al., 2021; 
trust regarding overall food safety of wet markets, they are Nguyen et al., 2021) has adversely affected the functioning 
still largely used for purchasing fresh food (Ha et al., 2020). of wet markets, as evidenced by the decrease in the number 
Vietnamese consumers tend to shop daily (Wertheim-Heck of visits made to the wet markets included in our study.
et al., 2014), purchasing small quantities to get the freshest 
food possible. This is particularly important for meat (Unger Ban on informal activities: While the markets selected for 
et al., 2019), fish and seafood (purchased alive) and leafy this study were all formal markets under direct management 
vegetables whose shelf-life is very short. An increased per- of the district authorities, it is common for informal traders 
ceived risk of frequenting wet markets during the pandemic to operate on the outskirts of formal markets (Maruyama & 
together with the implementation of strict social distancing Trung, 2010). Enforcing a decision from March 13, 2020 
1 3
 L. Reymondin et al.
(Official Letter 871/UBND-KT), local authorities severely Tracking the number and movement of Wi-Fi enabled 
limited street vending activities across the city. Given the devices has some limitations. Firstly, although the pro-
geographic proximity of informal and formal vendors, the portion of people owning a smartphone in urban areas is 
reduction in the number of informal traders, as well as their high and continues to increase (Newzoo, 2021), the most 
usual customers, probably influenced the overall reduction marginalized populations do not yet have access to high-
in traffic observed by our system within formal markets. end technologies. A large proportion of such populations 
is therefore invisible to the system used in this study to 
Strategies to cope with mobility restrictions: We can assume quantify the number of visits to wet markets. The popu-
that during the baseline and pre-shock period, and thus in larization of using such an approach to study social pro-
normal circumstances, many of the consumers visiting cesses might create a bias by excluding these populations 
the market during the morning and afternoon peak times and further marginalizing vulnerable communities. This 
are commuting (mostly by motorbike) and stopping at the study only provides a first glimpse of COVID-19 impacts 
markets (staying on their bikes) before or after going to on Vietnamese wet markets, requiring further study with 
work. Devices observed during this time therefore probably additional micro analysis. Secondly, the observed drop in 
belong to active people, with likely more resources. On the the number of visits and in visit frequency does not neces-
other hand, people going before or after rush hour are more sarily involve reductions in income, as consumers might 
likely to be older (retired, and on foot, thus living in close buy larger quantities during each visit to the markets. The 
proximity to the markets), or unemployed, thus with fewer results presented in this study are however corroborated by 
time constraints and the ability to avoid peak times. Dur- other studies that have found a strong negative impact on 
ing the shock, the city of Hanoi was placed under a partial entire food supply chains (Kang et al., 2021). Thirdly, this 
lockdown, with non-essential businesses closed and office study follows a small market sample during a relatively 
workers required to work from home (Directive No.16/CT- short period of time, furthermore, frequent power outages 
TTG). As shown by the validation of hypothesis 4, it is likely in the Minh Khai wholesale market during the aftershock 
that consumers had to change their daily routine to adapt to period resulted in increased variability in the data. These 
these restrictions, which resulted in a shift in the morning limitations were nevertheless mitigated during analysis by 
and afternoon peak times. However, it is key to note that removing this market for computation of metrics related 
although the shopping time shifted, the overall shape of the to consumer behaviors. Finally, this study was conducted 
distribution itself, with a main peak during the morning and with high standards for anonymization of data. Even if the 
another peak in the late afternoon, remained the same dur- aggregated version of the database is made publicly avail-
ing the shock period. There was therefore little impact of the able, a known MAC address from a given device cannot 
mobility restrictions in terms of increasing social distancing be linked to its anonymized record in the database. This 
within the wet markets. was, however, developed after data collection and it is not 
a feature built into the access points installed in the mar-
Decreased exchange between wholesale and retail mar‑ kets. There is a need for the development of strong frame-
kets A decrease in visit frequency is not necessarily linked works for the ethical use of phone metadata, as proposed 
to a drop in income as consumers might buy more goods by Kishore et al. (2020).
per visit to the markets. Nonetheless, the confirmation of Our results suggest that wet markets have been affected 
hypothesis 5 shows a decline in inter-market visits during the since the early stages of the COVID-19 outbreak in Vietnam. 
shock period. As shown in Fig. 6, the number of inter-market Although we do not have data for the period after July 2020, 
visits between the wholesale market Minh Khai and the the observed phenomena during the first semester of 2020 
retail markets Đồng Xa and Nghĩa Tân dropped by 39% and were likely emphasized in 2021, due to (i) a stronger ampli-
47% respectively. Similarly, inter-market visits between Tó tude of the outbreak, and (ii) further restrictions to access to 
(wholesale at night) and Trung Tâm (retail) markets dropped markets, with for instance the implementation of a ticket sys-
by 25% during the shock period. This drop in inter-market tem to regulate the days and times at which Hanoi citizens 
visits probably relates to a decrease in business-to-business were allowed to go grocery shopping (official letter 1304/
activities, i.e., retailers purchasing fewer products from UBND-KT by Tây Hồ district authorities on July 26, 2021).
wholesale markets. This decrease could be explained by an 
overall reduction in demand at retail markets, generating a 4.1  Policy recommendations
loss of income for retailers. This is in line with the observa-
tion of Kang et al. (2021) who found a 31% decrease in food Given the importance of wet markets (Raneri & Wertheim-
expenditures during the first year of the pandemic. Heck, 2019; Wertheim-Heck et al., 2019), the sharp decrease 
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
observed in the number of people visiting traditional markets Acknowledgements We thank Harri Washington, consultant to Alli-
since the very beginning of the pandemic raises concerns ance of Bioversity International and CIAT Science Writing Service, 
regarding food security and livelihoods of the most vulner- for language editing of this paper. This work is funded by the CGIAR 
Big Data Platform for Agriculture Inspire Challenge. Additionally, this 
able urban poor populations, as many rely on wet markets for work was supported by a grant from the CGIAR Program on Policies, 
both food purchases and livelihoods. We provide a series of Institutions, and Markets (PIM) aimed at studying COVID-19 impact.
recommendations to increase the resilience of wet markets 
and guarantee food security for the poor during such a crisis. Funding Consortium of International Agricultural Research 
Centers,Inpire award 2019, Louis Reymondin, Policies, Institutions, 
First, it is critical to mitigate consumers’ negative percep- Markets—COVID-19 grant, Louis Reymondin
tion of wet markets in terms of food safety. Consequently, 
this effort can reduce the shift from wet markets to corporate Data repository The aggregated anonymized observations will be 
retail outlets, ensuring accessibility to and affordability of made publicly available on https:// doi. org/ 10. 5281/ zenodo. 57073 12 
upon publication of this study.
fresh food. To do so, wet markets and related infrastructure 
should be upgraded with the aim of enabling the adoption Declarations 
of improved practices regarding hygiene and food safety 
conditions across all market actors. This recommendation Ethical approval This study was performed in line with the principles 
is in line with the plan issued in 2021 by the Hanoi People’s of the Declaration of Helsinki. The Ethics Committee of International 
Center for Tropical Agriculture granted approval to conduct the study 
Committee for the development and management of Hanoi (Date: April 21, 2019).
wet markets in 2021–2025, during which 169 wet markets 
will be upgraded and 141 will be newly constructed (Plan Conflict of interest All authors declare that they have no conflicts of 
228/KH-UBND dated October 12, 2021). Additionally, interest.
traceability systems need to be implemented and standard-
ized in wet markets, especially for fresh food products. Open Access This article is licensed under a Creative Commons Attri-
Second, in the interest of public health, policy should bution 4.0 International License, which permits use, sharing, adapta-
reverse the shift from fresh and nutritious food to less tion, distribution and reproduction in any medium or format, as long 
as you give appropriate credit to the original author(s) and the source, 
healthy processed food (Fiolet et al., 2018; Hall et al., 2019; provide a link to the Creative Commons licence, and indicate if changes 
Srour et al., 2019; Vandevijvere et al., 2019). We therefore were made. The images or other third party material in this article are 
recommend the strengthening of public health messaging included in the article’s Creative Commons licence, unless indicated 
related to nutrition and health, highlighting the importance otherwise in a credit line to the material. If material is not included in 
the article’s Creative Commons licence and your intended use is not 
of eating fresh and nutritious food during pandemics to permitted by statutory regulation or exceeds the permitted use, you will 
maintain healthy immune systems and lower the risks of need to obtain permission directly from the copyright holder. To view a 
disease (Aman & Masood, 2020). copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Finally, it is crucial to emphasize that the COVID-19 
pandemic has disproportionately affected poor communi-
ties that are already disadvantaged and vulnerable, leading to 
a substantial decline in their income and food expenditures References
(Adams-Prassl et al., 2020; Kang et al., 2021), and affect-
ing the wet markets on which they rely for procuring food. Adams-Prassl, A., Boneva, T., Golin, M., & Rauh, C. (2020). Inequality 
Therefore, to mitigate the negative consequences faced by in the impact of the coronavirus shock: Evidence from real time 
these populations, authorities should take proactive meas- surveys. Journal of Public Economics, 189, 104245. https:// doi. 
ures, such as implementing robust safety nets and financial org/ 10. 1016/j. jpube co. 2020. 104245
Aman, F., & Masood, S. (2020). How nutrition can help to fight against 
support mechanisms. Additionally, efforts should be made COVID-19 Pandemic. Pakistan Journal of Medical Sciences, 36. 
to increase the recognition and inclusion of informal vendors https:// doi. org/ 10. 12669/ pjms. 36. COVID 19- S4. 2776
operating in and around the wet markets, as vital contribu- Apple Inc. (2022). Use private Wi-Fi addresses on iPhone, iPad, iPod 
tors to the local economy. Authorities and researchers should touch, and Apple Watch. Retrieved March 07, 2023, from https:// 
suppo rt. apple. com/ en- us/ HT211 227
also conduct retailer and consumer surveys to complement Arai, A., Fan, Z., Matekenya, D., & Shibasaki, R. (2016). Compara-
cell phone metadata with analysis that sheds more light on tive perspective of human behavior patterns to uncover ownership 
behavior. By adopting such measures, policymakers can fos- bias among mobile phone users. ISPRS International Journal of 
ter a more equitable and resilient socio-economic environ- Geo-Information., 5(6), 85. https:// doi. org/ 10. 3390/ ijgi5 060085
Béné, C., Bakker, D., Rodriguez, M. C., Even, B., Melo, J., & Sonneveld, 
ment, thereby alleviating the challenges faced by the mar- A. (2021). Impacts of COVID-19 on people’s food security: Foun-
ginalized and vulnerable segments of society. dations for a more resilient food system. In CGIAR COVID-19 
1 3
 L. Reymondin et al.
Hub Discussion Paper. https:// doi. org/ 10. 2499/ p1573 8coll2. Google LLC. (2022b). Implementing MAC Randomization. Retrieved 
134295 03/07/2023 from https:// source. andro id. com/ docs/ core/ conne ct/ 
Boughton, D., Goeb, J., Lambrecht, I., Headey, D., Takeshima, H., wifi- mac- rando mizat ion
Mahrt, K., Masias, I., Goudet, S., Ragasa, C., Maredia, M. K., Ha, T. M., Shakur, S., & Pham Do, K. H. (2020). Risk perception and 
Minten, B., & Diao, X. (2021). Impacts of COVID-19 on agricul- its impact on vegetable consumption: A case study from Hanoi. 
tural production and food systems in late transforming Southeast Vietnam. Journal of Cleaner Production, 271, 122793. https://d oi. 
Asia: The case of Myanmar. Agricultural Systems, 188, 103026. org/ 10. 1016/j. jclep ro. 2020. 122793
https:// doi. org/ 10. 1016/j. agsy. 2020. 103026 Haan, S. de, Huynh, T., Duong, T. T., & Rubin, J. (2020). Defining the 
Buckland, N. J., Swinnerton, L. F., Ng, K., Price, M., Wilkinson, L. benchmark research sites (rural to urban transect) in Vietnam. 
L., Myers, A., & Dalton, M. (2021). Susceptibility to increased https:// hdl. handle. net/ 10568/ 113150
high energy dense sweet and savoury food intake in response to Hall, K. D., Ayuketah, A., Brychta, R., Cai, H., Cassimatis, T., Chen, 
the COVID-19 lockdown: The role of craving control and accept- K. Y., Chung, S. T., Costa, E., Courville, A., Darcey, V., Fletcher, 
ance coping strategies. Appetite, 158, 105017. https:// doi. org/ 10. L. A., Forde, C. G., Gharib, A. M., Guo, J., Howard, R., Joseph, P., 
1016/j. appet. 2020. 105017 & v, McGehee, S., Ouwerkerk, R., Raisinger, K., Zhou, M. (2019). 
Cao Mai Phuong, L. (2021). Food and beverage stocks responding to Ultra-processed diets cause excess calorie intake and weight gain: 
COVID-19. Investment Management and Financial Innovations, An inpatient randomized controlled trial of ad libitum food intake. 
18(3), 359–371. https:// doi. org/ 10. 21511/ imfi. 18(3). 2021. 30 Cell Metabolism, 30(1), 67-77.e3. https:// doi. org/ 10. 1016/j. cmet. 
Cisco Systems Inc. (2021). Meraki and iOS 14 MAC Address Rand- 2019. 05. 008
omization. Retrieved March 07, 2023, from https://d ocume ntati on. Harris, J., Nguyen, P. H., Tran, L. M., & Huynh, P. N. (2020). 
meraki. com/ Gener al_ Admin istra tion/ Cross- Platf orm_ Conte nt/ Nutrition transition in Vietnam: Changing food supply, food 
Meraki_ and_ MAC_ Addre ss_ Rando mizat ion prices, household expenditure, diet and nutrition outcomes. 
Cohen, J. (2020). Mining coronavirus genomes for clues to the out- Food Security, 12(5), 1141–1155. https:// doi. org/ 10. 1007/ 
break’s origins. Retrieved March 07, 2023, from https:// www. s12571- 020- 01096-x
scien ce. org/ conte nt/ artic le/ mining- coron avirus- genom es- clues- Hong, L., Wu, J., Frias-Martinez, E., Villarreal, A., & Frias-Martinez, 
outbr eak-s- origi ns V. (2019). Characterization of internal migrant behavior in the 
Deschasaux-Tanguy, M., Druesne-Pecollo, N., Esseddik, Y., de Edelenyi, F. immediate post-migration period using cell phone traces. Pro-
S., Allès, B., Andreeva, V. A., Baudry, J., Charreire, H., Deschamps, ceedings of the Tenth International Conference on Information 
V., Egnell, M., Fezeu, L. K., Galan, P., Julia, C., Kesse-Guyot, E., and Communication Technologies and Development. https:// doi. 
Latino-Martel, P., Oppert, J.-M., Péneau, S., Verdot, C., Hercberg, org/ 10. 1145/ 32870 98. 32871 19
S., & Touvier, M. (2021). Diet and physical activity during the Huynh, T. T. T., Pham, H. T. M., Trinh, H. T., Duong, T. T., Nguyen, T. 
coronavirus disease 2019 (COVID-19) lockdown (March–May 2020): M., Hernández, R., Lundy, M., Nguyen, K. T., Nguyen, L. L. T., 
Results from the French NutriNet-Santé cohort study. The American Vuong, V. T., Nguyen, H. T., Truong, M. T., Do, H. P. T., Raneri, 
Journal of Clinical Nutrition, 113(4), 924–938. https:// doi. org/ 10. J., HoangKy, T., & Haan, S. de. (2021). Partial food systems base-
1093/ ajcn/ nqaa3 36 line assessment at the Vietnam benchmark sites. https://h dl.h andle. 
Dien, l. e. N., Thang, N. M., & Bentley, M. E. (2004). Food consump- net/ 10568/ 113122
tion patterns in the economic transition in Vietnam. Asia Pacific Kang, Y., Baidya, A., Aaron, A., Wang, J., Chan, C., & Wetzler, E. 
Journal of Clinical Nutrition, 13(1), 40–47. https://p ubmed.n cbi. (2021). Differences in the early impact of COVID-19 on food 
nlm. nih. gov/ 15003 913/ security and livelihoods in rural and urban areas in the Asia 
Đỗ, T. K., Hansen, A., & Wertheim-Heck, S. (2020). Governing Covid- Pacific Region. Global Food Security, 31, 100580. https:// doi. 
19 in Vietnam: the Politics of Pandemic Control. https:// halshs. org/ 10. 1016/j. gfs. 2021. 100580
archi ves- ouver tes. fr/ halshs- 03151 049 Kim Dang, A., Xuan Tran, B., Tat Nguyen, C., Le Thi, H., Thi Do, H., 
Dollar, D., Glewwe, P., & Agrawal, N. (2004). Economic Growth, Duc Nguyen, H., Hoang Nguyen, L., Huu Nguyen, T., Thi Mai, 
Poverty, and Household Welfare in Vietnam (N. Agrawal, P. H., Dinh Tran, T., Ngo, C., Thi Minh, Vu., & T., Latkin, C. A., 
Glewwe, & D. Dollar, Eds.). The World Bank. https://d oi.o rg/1 0. Zhang, M. W. B., & Ho, R. C. M. (2018). Consumer preference 
1596/0- 8213- 5543-0 and attitude regarding online food products in Hanoi, Vietnam. 
Dworkin, M. (2015). SHA-3 standard: permutation-based hash and International Journal of Environmental Research and Public 
extendable-output functions. federal information processing stand- Health, 15(5), 981. https:// doi. org/ 10. 3390/ ijerp h1505 0981
ards. (NIST FIPS), National Institute of Standards and Technol- Kishore, N., Kiang, M., & v, Engø-Monsen, K., Vembar, N., Schroeder, 
ogy, Gaithersburg, MD. https:// doi. org/ 10. 6028/ NIST. FIPS. 202 A., Balsari, S., & Buckee, C. O. (2020). Measuring mobility to 
Fiolet, T., Srour, B., Sellem, L., Kesse-Guyot, E., Allès, B., Méjean, monitor travel and physical distancing interventions: A common 
C., Deschasaux, M., Fassier, P., Latino-Martel, P., Beslay, M., framework for mobile phone data analysis. The Lancet Digi-
Hercberg, S., Lavalette, C., Monteiro, C. A., Julia, C., & Touvier, tal Health, 2(11), e622–e628. https:// doi. org/ 10. 1016/ S2589- 
M. (2018). Consumption of ultra-processed foods and cancer risk: 7500(20) 30193-X
Results from NutriNet-Santé prospective cohort. British medical Kutscher, N., & Kreß, L.-M. (2018). The ambivalent potentials of 
journal, 360. https:// doi. org/ 10. 1136/ bmj. k322 social media use by unaccompanied minor refugees. Social Media 
Ghahramani, M., Zhou, M., & Hon, C. T. (2019). Mobile phone data + Society, 4(1), 2056305118764438. https:// doi. org/ 10. 1177/ 
analysis: A spatial exploration toward hotspot detection. IEEE 20563 05118 764438
Transactions on Automation Science and Engineering, 16(1), Lazer, D., & Radford, J. (2017). Data ex Machina: Introduction to Big 
351–362. https:// doi. org/ 10. 1109/ TASE. 2018. 27952 41 Data. Annual Review of Sociology, 43, 19–39. https:// doi. org/ 10. 
Gilchrist, J. (2003). Encryption. Encyclopedia of Information Systems, 1146/ annur ev- soc- 060116- 053457
87–100. https:// doi. org/ 10. 1016/ B0- 12- 227240- 4/ 00054-X Le, T.-A.T., Vodden, K., Wu, J., & Atiwesh, G. (2021). Policy 
Google LLC. (2022a). Google COVID-19 Community Mobility responses to the COVID-19 pandemic in Vietnam. International 
Reports. Retrieved 03/07/2023 from https:// www. google. com/ Journal of Environmental Research and Public Health, 18(2), 
covid 19/ mobil ity/ 559. https:// doi. org/ 10. 3390/ ijerp h1802 0559
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
Li, Q., Guan, X., Wu, P., Wang, X., Zhou, L., Tong, Y., Ren, R., Leung, Healthier Diets and Sustainable Food Systems in Viet Nam. 
K., Lau, E., Wong, J. Y., Xing, X., Xiang, N., Wu, Y., Li, C., https:// doi. org/ 10. 2499/ P1573 8COLL2. 133433
Chen, Q., Li, D., Liu, T., Zhao, J., Li, M., & Feng, Z. (2020). Spiehler, A., & Fischer, B. (2021). Animal agriculture, wet markets, 
Early transmission dynamics in Wuhan, China, of novel coronavi- and COVID-19: A case study in indirect activism. Food Ethics, 
rus–infected pneumonia. New England Journal of Medicine, 382. 6(2), 10. https:// doi. org/ 10. 1007/ s41055- 021- 00090-z
https:// doi. org/ 10. 1056/ NEJMo a2001 316 Srour, B., Fezeu, L. K., Kesse-Guyot, E., Allès, B., Méjean, C., 
Maruyama, M., & Trung, L. V. (2010). The nature of informal food Andrianasolo, R. M., Chazelas, E., Deschasaux, M., Hercberg, 
bazaars: Empirical results for urban Hanoi. Vietnam. Journal of S., Galan, P., Monteiro, C. A., Julia, C., & Touvier, M. (2019). 
Retailing and Consumer Services, 17(1), 1–9. https:// doi. org/ 10. Ultra-processed food intake and risk of cardiovascular disease: 
1016/j. jretc onser. 2009. 08. 006 Prospective cohort study (NutriNet-Santé). BMJ, 365. https://d oi. 
Mizumoto, K., Kagaya, K., & Chowell, G. (2020). Effect of a wet mar- org/ 10. 1136/ bmj. l1451
ket on coronavirus disease (COVID-19) transmission dynamics in Thang, N. M., & Popkin, B. M. (2004). Patterns of food consumption 
China, 2019–2020. International Journal of Infectious Diseases, in Vietnam: Effects on socioeconomic groups during an era of 
97, 96–101. https:// doi. org/ 10. 1016/j. ijid. 2020. 05. 091 economic growth. European Journal of Clinical Nutrition, 58(1), 
Munster, V. J., Bausch, D. G., de Wit, E., Fischer, R., Kobinger, G., 145–153. https:// doi. org/ 10. 1038/ sj. ejcn. 16017 61
Muñoz-Fontela, C., Olson, S. H., Seifert, S. N., Sprecher, A., The University of Adelaide. (2018). The Vietnam urban food consumption 
Ntoumi, F., Massaquoi, M., & Mombouli, J.-V. (2018). Outbreaks & expenditure study. Retrieved March 07, 2023, from https:// www. 
in a rapidly changing Central Africa — Lessons from Ebola. New adela ide. edu. au/ global- food/ resea rch/ vietn am- consu mer-survey
England Journal of Medicine, 379(13), 1198–1201. https:// doi. Turner, S., & Binh, N. N. (2021). Street Vendor Struggles: Maintain-
org/ 10. 1056/ NEJMP 18076 91/ SUPPL_ FILE/ NEJMP 18076 91_ ing a Livelihood Through the COVID-19 Lockdown in Hanoi, 
DISCL OSURES. PDF Vietnam. In B. Doucet, R. van Melik, & P. Filion (Eds.), Vol-
Newzoo. (2021). Global Mobile Market Report. Retrieved 03/07/2023 ume 1: Community and Society (07 ed., Vol. 1, pp. 21–30). ILRI 
from https://n ewzoo.c om/r esour ces/t rend-r eport s/n ewzoo-g lobal- Research Brief 91. Nairobi, Kenya: ILRI. https:// hdl. handle. net/ 
mobile- market- report- 2021- free- versi on 10568/ 102172
Nguyen, C., Tran, D., Nguyen Tu, A., & Nguyen, N. (2021). The effects Turner, S., Langill, J. C., & Nguyen, B. N. (2021). The utterly unfore-
of perceived risks on food purchase intention: The case study of seen livelihood shock: COVID-19 and street vendor coping 
online shopping channels during COVID-19 pandemic in Viet- mechanisms in Hanoi, Chiang Mai and Luang Prabang. Singapore 
nam. Journal of Distribution Science, 19, 19–27. https:// doi. org/ Journal of Tropical Geography, 42(3), 484–504. https:// doi. org/ 
10. 15722/ jds. 19.9. 202109. 19 10. 1111/ SJTG. 12396
Nguyen, P. H., Kachwaha, S., Pant, A., Tran, L. M., Ghosh, S., Sharma, Unger, F., Nguyen, T. T., Pham, V. H., Le, T. T. H., Hung, N.-V., Sinh, 
P. K., Shastri, V. D., Escobar-Alegria, J., Avula, R., & Menon, P. D.-X., Nguyen, T. D. N., Nguyen, T. L., Nguyen, T. T. H., Tran, 
(2021). Impact of COVID-19 on household food insecurity and T. B. N., Pham, D. P., Grace, D., & Nguyen, T. Q. C. (2019). 
interlinkages with child feeding practices and coping strategies Overview of typical pork value chains in Vietnam.
in Uttar Pradesh, India: A longitudinal community-based study. van Melik, R., Filion, P., & Doucet, B. (2021). Volume 3: Public Space 
BMJ Open, 11(4). https://d oi.o rg/1 0.1 136/b mjope n-2 021-0 48738 and Mobility (1st ed., p. 155). Bristol University Press. http:// 
Nguyen-Viet, H., Tran, T.-H., Unger, F., Xuan Sinh, D., & Grace, D. www. jstor. org/ stable/ j. ctv1t 4m1m6
(2017). Food safety in Vietnam: Where we are at and what we Vandevijvere, S., Jaacks, L. M., Monteiro, C. A., Moubarac, J.-C., 
can learn from international experiences. Infectious Diseases of Girling-Butcher, M., Lee, A. C., Pan, A., Bentham, J., & Swinburn, 
Poverty, 6. https:// doi. org/ 10. 1186/ s40249- 017- 0249-7 B. (2019). Global trends in ultraprocessed food and drink product 
OpenStreetMap contributors. (2021). Planet dump retrieved from sales and their association with adult body mass index trajec-
https:// planet. osm. org.https:// www. opens treet map. org tories. Obesity Reviews, 20(S2), 10–19. https:// doi. org/ 10. 1111/ 
Parks, T., Chatsuwan, M., & Pillai, S. (2020). Enduring the Pandemic: obr. 12860
Surveys of the Impact of COVID-19 on the Livelihoods of Thai Peo- Vo, K., & Smith, G. (2017). Vietnam Retail Foods. Sector Report 2016. 
ple. Bangkok, Thailand: The Asia Foundation. https://a siafo undat ion. In Global Agricultural Information Network (Vol. VM6081).
org/w p-c onten t/u pload s/2 020/0 9/E nduri ng-t he-P andem ic-C ovid-1 9- Wertheim-Heck, S., Raneri, J. E., & Oosterveer, P. (2019). Food 
Impact- on- Thail and- Livli hoods- Sept- 2020. pdf safety and nutrition for low-income urbanites: Exploring a 
Pastor-Escuredo, D., Imai, A., Luengo-Oroz, M., & Macguire, D. social justice dilemma in consumption policy. Environment 
(2019). Call Detail Records to Obtain Estimates of Forcibly and Urbanization, 31(2), 397–420. https:// doi. org/ 10. 1177/ 
Displaced Populations. In A. A. Salah, A. Pentland, B. Lepri, E.  09562 47819 858019
Letouzé (Ed.), Guide to Mobile Data Analytics in Refugee Sce- Wertheim-Heck, S., Vellema, S., & Spaargaren, G. (2014). Constrained 
narios: The “Data for Refugees Challenge” Study (pp. 29–52). consumer practices and food safety concerns in Hanoi. Interna-
Springer International Publishing. https:// doi. org/ 10. 1007/ tional Journal of Consumer Studies, 38(4), 326–336. https:// doi. 
978-3- 030- 12554-7_2 org/ 10. 1111/ ijcs. 12093
Raneri, J. E., & Wertheim-Heck, S. (2019). Retail diversity for die- Williams, N. E., Thomas, T. A., Dunbar, M., Eagle, N., & Dobra, A. 
tary diversity: Resolving food-safety versus nutrition priorities (2015). Measures of human mobility using mobile phone records 
in Hanoi. UNSCN Nutrition, 44, 61–69. https:// hdl. handle. net/ enhanced with GIS data. PLoS ONE, 10(7), e0133630. https://d oi. 
10568/ 103240 org/ 10. 1371/ JOURN AL. PONE. 01336 30
Raneri, J. E., Kennedy, G., Nguyen, T., Wertheim-Heck, S., Do, H., World Health Organization. (2020). WHO COVID-19 Dashboard. 
& Nguyen, P. H. (2019). Determining Key Research Areas for Retrieved March 07, 2023, from https:// covid 19. who. int/
1 3
 L. Reymondin et al.
Louis Reymondin Louis Hieu Trung Le Hieu Trung Le is 
Reymondin is a senior scientist at the Deputy Director General of the 
the Alliance of Bioversity General Statistics Office of Vietnam. 
International and CIAT where he He oversees and directs the activities 
leads the Data Driven of the following units: Department of 
Sustainability research program. National Accounts System; Depart-
Louis is an expert in developing ment of Industry and Construction 
solutions that harness the power Statistics; Price Statistics Depart-
of Machine Learning to make ment; Department of Legal Affairs 
sense of Big Earth Data. His and Statistical Inspection; Institute of 
background is in machine Statistical Science; Journal of Num-
learning, software development bers and Facts. Furthermore, he is the 
and remote sensing. Louis’ head of the Department for the 
research focuses on the development and implementation of large-scale Advancement  of  Women in 
monitoring systems for near real time human impact assessment Statistics.
combining multi sources of data.
Thibaud Vantalon Thibaud Vantalon Tuyen Thi Thanh Huynh Mrs 
is a post-doctoral fellow at the Alli- Tuyen Thi Thanh Huynh is a 
ance of Bioversity International and research team leader for the 
CIAT. Thibaud holds a PhD in Theo- Food Environment and Con-
retical Physics from University sumer Behavior (FECB) lever in 
Autonoma de Barcelona. During his Asia at the Alliance of Biover-
PhD in particle physics, Thibaud sity and CIAT. She works as a 
developed robust research protocols Country Coordinator of OneC-
to analyse the results of many simu- GIAR Sustainable Healthy Diets 
lations of particles interactions at through Food Systems Transformation (SHiFT) initiative in Vietnam, 
high energy levels and analyzed theo- where she leads the coordination unit in Vietnam and its support of the 
retical implication of experimental national government and international agencies and movements in pro-
results. At the Alliance of Bioversity moting delivery and use of sustainable healthy food that ultimately 
International and CIAT, Thibaud provides economic, social and nutritional benefits to all consumers.
uses his strong background in 
research, programming and mathe- Ricardo Hernandez Dr. Ricardo 
matic to develop deep-learning methods to extract key insights from Hernandez is an Agri-food Econo-
large amount of remote sensing data in different tropical landscapes. mist for the Alliance of Bioversity 
and CIAT (ABC) -Asia, and the 
Huong Thi Mai Pham Huong Thi country representative for ABC in 
Mai Pham is a Senior Research Vietnam. His work is related to The 
Associate for the Food Environ- Alliance’s initiative on Food Envi-
ment and Consumer Behavior ronment & Consumer Behavior, 
(FECB) lever in Asia at the Alli- especially on market assessments 
ance of Bioversity and CIAT. for nutritious fresh fruits and vege-
Shehas 15 years of research tables, consumer behavior of urban, 
experience, where she has been peri-urban, and rural households, 
working with smallholder farm- and assessment of nutrition inter-
ers’ organizations and enter- ventions related to Nutrition-Sensitive Value Chain framework. Previously, 
prises and interacting with dif- he worked as a Research Coordinator for IFPRI’s PRSSP in Bangladesh 
ferent partners in Vietnam. where he led different value chain and household level analysis. Previous to 
Huong’s research focuses on that he worked as an Assistant Professor at the Department of Agricultural, 
value chain development, food systems, and food environment. Previ- Food and Resource Economics at Michigan State University (MSU). He has 
ously she coordinated the adaptation of Farmer Business Schools to a PhD in Agricultural Economics from MSU and his professional interests 
local context for indigenous vegetable and vegetable seed value chains are related to food system transformations, agricultural and supply value 
in Northern Vietnam. chains development and food security in development countries.
1 3
Free Wi-Fi to assess impact of COVID-19 on Hanoi wet markets
Kien Tri Nguyen Kien Tri Nguyen 
is a Statistical Modeler at the 
Brice Even Brice Even is a Sus- Alliance of Bioversity and CIAT. 
tainable Food System specialist Prior to that, he worked as a Data 
at the Alliance of Bioversity Manager where he helped the 
International and CIAT, based in Alliance’s research teams in Asia 
Hanoi since 2012. Brice has develop household survey data 
nearly 15 years of experience collection protocols, design 
working on market access, value smartphone data collection 
chain development, food systems applications and process and 
and food environment. He has store the data for further analy-
been involved in 25+ projects across Asia-Pacific and Africa. Brice sis. Currently, he is a part of the 
holds one MSc in agricultural and rural development from ISTOM and Digital Inclusion research lever to work on modelling of crop-climate 
one MSc in development studies from IEDES – Sorbonne University. interactions.
Thang Cong Nguyen Thang Cong Christophe Béné Christophe 
Nguyen is a senior statistician. He Béné is Principal Scientist and 
has been working at General Statis- Senior Policy Advisor currently 
tics Office of Vietnam for 16 years. affiliated to the CGIAR Systems 
His background is in statistics. He Transformation Science Group 
is a deputy head of Sampling and seconded to Wageningen 
Design and Survey Division under University. He has 20+ years of 
Statistical Data Collection and IT experience conducting inter-
Application Department. He disciplinary research focusing on 
updates methodologies and sample poverty alleviation and food 
designs for surveys in the field of security in low and middle-
agricultural statistics and others income countries. In his career, 
conducted by GSO every year. he worked on a wide range of topics, including natural resource man-
agement, science-policy interface, resilience (measurement), and more 
Trong Van Phan Trong Van Phan recently food system (sustainability).
is a GIS and Remote Sensing Spe-
cialist based in Alliance Biover-
sity-CIAT Asia Regional Office 
in Hanoi, Vietnam. He has years 
experiences in the field of GIS 
and remote sensing especially 
working with Satellite data, geo-
database systems and spatial data 
analysis application. He contrib-
utes to the development and main-
tenance of a Terra-i GIS database 
in South East Asia. His research 
interests are the applications of 
GIS and remote sensing data, modeling in the fields such as land use, land 
cover change under the impacts of Social economic activities and climate 
change, forest monitoring, and crops mapping using deep learning.
1 3