Asian Development Bank Institute: ADBI Working Paper Series

ADBI Working Paper Series
COVID-19 IMPACT ON MICRO, SMALL,

AND MEDIUM-SIZED ENTERPRISES
UNDER THE LOCKDOWN: EVIDENCE
FROM A RAPID SURVEY IN
THE PHILIPPINES
Shigehiro Shinozaki and

Lakshman N. Rao
No. 1216
February 2021
Asian Development Bank Institute

Shigehiro Shinozaki is a senior economist and Lakshman N. Rao is a statistician,
both at the Economic Research and Regional Cooperation Department of the Asian
Development Bank.
The views expressed in this paper are the views of the author and do not necessarily reflect
the views or policies of ADBI, ADB, its Board of Directors, or the governments they
represent. ADBI does not guarantee the accuracy of the data included in this paper and
accepts no responsibility for any consequences of their use. Terminology used may not
necessarily be consistent with ADB official terms.
Working papers are subject to formal revision and correction before they are finalized and
considered published.
The Working Paper series is a continuation of the formerly named Discussion Paper series;
the numbering of the papers continued without interruption or change. ADBI’s working papers
reflect initial ideas on a topic and are posted online for discussion. Some working papers may
develop into other forms of publication.
Suggested citation:
Shinozaki, S. and L. N. Rao. 2021. COVID-19 Impact on Micro, Small, and Medium-Sized
Enterprises under the Lockdown: Evidence from a Rapid Survey in the Philippines. ADBI
Working Paper 1216. Tokyo: Asian Development Bank Institute. Available:
https://www.adb.org/publications/covid-19-impact-msme-under-lockdown-evidence-rapid-
survey-philippines
Please contact the authors for information about this paper.
Email: sshinozaki@adb.org, nagrajrao@adb.org
The authors would like to thank the Bureau of Small and Medium Enterprise Development
of the Department of Trade and Industry of the Republic of the Philippines and the
Philippine Chamber of Commerce and Industry, which were partners in conducting the
survey. They also thank Dave Pipon, Jude David Adarna Roque, and Josephine Penaflor
Ferre for their research assistance.
Asian Development Bank Institute

Kasumigaseki Building, 8th Floor
3-2-5 Kasumigaseki, Chiyoda-ku
Tokyo 100-6008, Japan
Tel: +81-3-3593-5500
Fax: +81-3-3593-5571
URL: www.adbi.org
E-mail: info@adbi.org
© 2021 Asian Development Bank Institute

ADBI Working Paper 1216 Shinozaki and Rao
Abstract
The novel coronavirus disease, COVID-19, has brought significant change to people’s
lives and business activities nationally, regionally, and globally. The Philippines took swift
action—including enhanced community quarantine (ECQ)—to contain the pandemic and
launched an emergency subsidy program with massive public spending to support disrupted
households and businesses. The strict lockdown ran from mid-March to the end of May 2020
in the national capital region and high-risk provinces, causing huge economic losses.
Six months after the March lockdown, the Philippine economy has moved to the recovery
stage, but micro, small, and medium-sized enterprises (MSMEs) are continuing to confront a
sharp drop in demand and revenue. This paper examines the initial impact on MSMEs of the
ECQ and lockdown measures using evidence obtained from a rapid nationwide survey
conducted from the end of March to mid-April 2020 and derives policy implications.
Keywords: COVID-19, economic crisis, economic impact, MSMEs, SME development,

access to finance, SME policy, Philippines
JEL Classification: D22, G20, L20, L50

Contents
1. INTRODUCTION............................................................................................................1
2. INITIAL NATIONAL RESPONSES TO THE COVID-19 PANDEMIC IN ASIA ............. 2
3. METHODOLOGY ...........................................................................................................6
3.1 Data Structure ....................................................................................................7

3.2 Weighting the Survey Data ................................................................................8
3.3 Analytical Approach .........................................................................................11
3.4 Linear Probability Model ..................................................................................12
4. PROFILE OF THE SURVEYED MSMEs..................................................................... 13
5. FINDINGS FROM THE DESCRIPTIVE STATISTICS AND ECONOMETRIC

ANALYSES ..................................................................................................................15
5.1 Impact on MSME Sales ...................................................................................21

5.2 Impact on MSME Revenue ..............................................................................22
5.3 Employment during the COVID-19 Pandemic................................................. 24
5.4 Wage Payments during the COVID-19 Pandemic .......................................... 26
5.5 Financial Condition during the COVID-19 Pandemic ...................................... 27
6. POLICY IMPLICATIONS .............................................................................................30
6.1 Identifying Focus Groups for Assistance ......................................................... 33

6.2 Differentiating Policy Measures by Firm Size.................................................. 33
7. CONCLUSION .............................................................................................................34
REFERENCES ........................................................................................................................35
1. INTRODUCTION
The novel coronavirus disease, COVID-19, has significantly altered people’s lives and
business activities at the national, regional, and global levels. The Philippine government
responded promptly, imposing enhanced community quarantine (ECQ) to contain the
spread of COVID-19 and implementing an emergency subsidy program of massive
public spending to support badly affected households and businesses. The lockdown—
or strict stay-at-home order—started on 16 March 2020, initially covering the national
capital region and high-risk provinces. The government extended it three times until the
end of May 2020. During the lockdown, the Philippine economy immediately experienced
a sharp decline in domestic and foreign demand, international trade, national production,
and consumer confidence. Together with funding constraints on businesses, it signaled
the start of huge economic losses (ADB 2020c, 2020d).
The Asian Development Bank (ADB 2020a) estimated that regional economic growth
in developing Asia would decline sharply from 5.1% in 2019 to ‒0.4% in 2020 due
to the pandemic’s effects. As the contraction did not emanate from economic or financial
turmoil, ADB forecast a 6.8% rebound in regional economic growth in 2021. This
assumed that the pandemic is contained by using expansionary fiscal and monetary
policies among ADB’s developing members. The estimation indicated that the Philippine
GDP would contract by 8.5% in 2020, with an expected strong recovery
to 6.5% growth in 2021, assuming that the restrictions ease and businesses
gradually reopen.
According to the Philippine Institute for Development Studies (Abrigo et al 2020), the
Philippines may suffer economic losses between P276.3 billion and P2.5 trillion due to
the COVID-19 pandemic. The most affected business sectors will be manufacturing, with
losses between P82.1 billion and P855.2 billion, wholesale and retail trade, with losses
between P93.2 billion and P724.8 billion, and transport/storage/communication, with
losses between P11.7 billion and P124.3 billion. Abrigo et al (2020) also estimated that,
if the ECQ continued to May 2020, it would potentially cost the Philippine economy at
least P150 billion given the decline in household consumption.
The COVID-19 crisis differs from the 1997–1998 Asian financial crisis and the
2008–2009 global financial crisis (GFC) as the primary cause was not regional or global
economic or financial turmoil, suggesting a sharp recovery in 2021 from the significant
contraction in 2020. The International Monetary Fund (IMF 2020a) estimated that the
global economy would drop sharply by ‒3% in 2020, a far worse fall than occurred during
the GFC. However, it will recover by 5.8% in 2021, assuming that countries control the
pandemic in the second half of 2020.
The global economy has experienced epidemics in the past—the Severe Acute
Respiratory Syndrome (SARS) and the Middle East Respiratory Syndrome (MERS). The
SARS outbreak in 2003 involved more than 8,000 cases affecting 26 countries, including
Hong Kong, China; the People’s Republic of China (PRC); Singapore; Taipei,China; and
Viet Nam. The MERS outbreak occurred in Saudi Arabia in 2012 and spread to several
Asian economies, including Malaysia, the Philippines, the Republic of Korea, and
Thailand. During the MERS outbreak, the number of micro, small, and medium-sized
enterprises (MSMEs) fell by 0.4% in the Philippines, with employment dropping by 3.3%
during 2012 and 2013. The numbers increased by 0.6% and 2.5%, respectively, after the
outbreak settled down in 2014.1
1 Authors’ calculation based on the Asia Small and Medium-Sized Enterprise Monitor 2020 database.
1
The estimates of the pandemic’s economic impact assume that countries will contain
COVID-19 in 2020. However, if the pandemic is prolonged, with second and third waves,
the economic damage will increase exponentially at the national, regional, and global
levels. Governments across the region must use their limited budgets effectively to
support the most affected groups, including MSMEs.
MSMEs are the backbone of the national economy but remain vulnerable to external
shocks, such as financial crises, disasters, and forced changes in the business
environment—like the COVID-19 pandemic response. In the Philippines, MSMEs drive
the national economy. They accounted for 99.5% of all enterprises and employed 63.2%
of the labor force as of the end of 2018 (ADB 2020b). Their ability to access finance faces
constraints even during non-crisis periods. The share of MSME loans in the total
outstanding bank loans was 6.1% in 2019. Bank loans to MSMEs accounted for only
3.2% of the GDP during the same period. This suggests that a very small number of
MSMEs have access to bank credit. A prolonged pandemic will make it more difficult for
MSMEs to raise funds from formal financial institutions and to survive the crisis and its
aftermath, which could contribute to more potential losses to the Philippine economy and
risk the projected economic rebound in 2021.
This paper examines the initial one-month impact on MSMEs of the ECQ lockdown
measures imposed by the Philippines, using a multivariate analytical model, and
presents policy implications with evidence obtained through a rapid nationwide survey
from the end of March to mid-April 2020. The next section reviews the initial policy
responses to support MSMEs affected by the COVID-19 in selected Asian economies.
The third section examines the methodology that the study used. The fourth and fifth
sections discuss the profile of the surveyed MSMEs and the first month impact of the
lockdown on MSMEs’ sales, revenue, employment, wages, and financial conditions,
followed by its policy implications and a concluding section.
2. INITIAL NATIONAL RESPONSES TO THE COVID-19

PANDEMIC IN ASIA
Many Asian countries acted quickly to contain the spread of COVID-19, not only taking
quick and responsive healthcare-related actions but also providing economic stimulus
packages involving substantial government spending. Table 1 summarizes the initial
policy actions to support private businesses, especially MSMEs, which selected Asian
countries implemented soon after the COVID-19 outbreak (World Bank 2020 and
IMF 2020b).
The countries consist of the 10 members of the Association of Southeast Asian Nations
(ASEAN) plus Japan, the PRC, and the Republic of Korea. Each country launched wide-
ranging policy packages not only to stop the spread of the virus but also to retain
economic activities a month after the World Health Organization (WHO) declared
COVID-19 to be a pandemic (on 11 March 2020). Debt finance was the most widely used
policy instrument to support MSMEs, followed by tax relief, employment support, and
support to retain businesses.
2
Table 1: Initial MSME Support Measures Responding to COVID-19

in Selected Asian Countries
Debt Finance
Liquidity Relaxed New
Support for Debt Lending Lending to
Financial Restructuring/ Conditions/ MSMEs/
Institutions/ Deferral of Interest Emergency Special
Capital Debt Rate Refinancing Credit Regulatory
Item Injection Payments Reduction Facility Guarantees Forbearance
Lower-middle- CAM √ √ √ √ √
income LAO √ √ √ √ √
economies
MYA √ √ √ √
PHI √ √ √ √ √ √
VIE √ √ √ √ √
Upper-middle- INO √ √ √ √ √ √
income MAL √ √ √ √ √ √
economies
PRC √ √ √ √ √ √
THA √ √ √ √ √ √
High-income BRU √ √
economies JPN √ √
ROK √ √ √ √ √
SIN √ √ √ √
Tax Relief
Payroll, Social
Corporate Tax Expedited Tax Incentives for Security, VAT, and
Item Reduction Refunds Investors Land Taxes
Lower-middle- CAM √
income LAO
economies
MYA √
PHI
VIE √ √ √
Upper-middle- INO √ √
income MAL √ √ √
economies
PRC √ √ √
THA √ √
High-income BRU
economies JPN √
ROK √
SIN √ √
Employment Support
Support for
Informal/Self-
Wage employed Unemployment New Working Labor Training
Item Subsidies Workers Benefits Arrangements Subsidies
Lower-middle- CAM √
income LAO
economies
MYA
PHI √ √ √
VIE √ √
Upper-middle- INO
income MAL √ √ √
economies
PRC √ √ √
THA √ √
High-income BRU
economies JPN √ √
ROK √ √ √ √
SIN √ √ √
continued on next page
3
Table 1 continued
Business Support
Reduction/ Focus Vouchers
Reduced Waiver of Reduced Group for Remote
Utility Reduced Government Import Expenditure Business
Item Payments Rent/Leasing Fees Restrictions Programs Services
Lower-middle- CAM √ √
income LAO √
economies
MYA
PHI √
VIE √ √
Upper-middle- INO √ √
income MAL √ √ √ √
economies
PRC √ √ √ √ √ √
THA √ √
High-income BRU
economies JPN √
ROK √ √ √ √
SIN √ √ √ √
BRU = Brunei Darussalam; CAM = Cambodia; INO = Indonesia; JPN = Japan; LAO = Lao People’s Democratic Republic
(Lao PDR); MAL = Malaysia; MSME = micro, small, and medium-sized enterprise; MYA = Myanmar;
PHI = Philippines; PRC = People’s Republic of China; ROK = Republic of Korea; SIN = Singapore; THA = Thailand;
VIE = Viet Nam.
Sources: Recomposed from the World Bank’s “Map of SME-Support Measures in Response to COVID-19.” 14 April 2020.
https://www.worldbank.org/en/data/interactive/2020/04/14/map-of-sme-support-measures-in-response-to-covid-19;
International Monetary Fund. “COVID-19 Policy Tracker.” https://www.imf.org/en/Topics/imf-and-covid19/Policy-
Responses-to-COVID-19. Accessed on 24 June 2020.
Central banks used several liquidity support measures to facilitate banks’ lending to
MSMEs and those sectors that COVID-19 and the quarantine measures most affected,
such as large-scale capital injections for commercial and policy banks (for example,
Cambodia, Indonesia, and the PRC), reduced base rates for lending (Cambodia, the
Philippines, the Republic of Korea, and Viet Nam), relaxed capital requirements for banks
(the Philippines), and related regulatory easing to stimulate MSME finance. Most
countries allowed the deferral of loan repayments and loan restructuring for MSMEs.
Malaysia granted a 6-month moratorium on loan repayments, and the Philippines set a
30-day grace period for loan repayments.
In parallel, many countries launched emergency concessional loan schemes, special
funds, and refinancing facilities to encourage MSMEs to access new loans during the
crisis. Cambodia established a new public bank for MSMEs. Malaysia established a
COVID-19 special relief facility for working capital financing for MSMEs. Myanmar
created a COVID-19 fund to finance affected MSMEs and sectors such as tourism with
concessional interest rates of 1%. Thailand and Viet Nam launched soft loan packages
for MSMEs with lower interest rates. Japan provided largely zero interest rate loans and
full credit guarantees to MSMEs facing a sharp decrease in sales. Malaysia and the
Republic of Korea also offered special credit guarantees to affected MSMEs.
Tax relief is a key component of economic stimulus packages in several countries, where
corporate tax reductions and exemptions and deferred payments are the major support
that MSMEs, such as those in manufacturing and tourism, could tap. Indonesia gradually
reduced its corporate income tax to 22% for 2020 and 2021 and to 20% for 2022, mainly
targeting manufacturing. Malaysia, Myanmar, and Singapore accepted deferred (for
three months) corporate income tax payments for MSMEs. Singapore also offered
corporate tax rebates. The PRC accepted 8-year loss carry-overs for sectors such as
transport, catering, hotels, and tourism.
4
Many Asian countries reduced the social security contribution or exempted disrupted
MSMEs and industries from paying it (for example, Cambodia, Japan, Malaysia, the
PRC, Thailand, and Viet Nam). The payment of value-added tax (VAT) by companies
was also reduced or exempted in many countries (Indonesia, Malaysia, the PRC,
Singapore, and Viet Nam). They provided various tax holidays/breaks for businesses
and some sectors, especially small business owners and the self-employed. In
Indonesia, hotels and restaurants in major tourist destinations, such as Bali, enjoyed
temporary suspension of tax payments (six months), while the central government
compensated local governments for the resultant tax revenue losses.
Several countries created various subsidy schemes for employers to pay wages and
cash transfer arrangements for displaced workers, and made them available to qualified
MSMEs and priority sectors. Cambodia paid 60% of the minimum wage
for furloughed workers in the garment sector. Malaysia launched an enhanced
wage support scheme to support MSME employees. The Philippines provided
one-off financial assistance for affected workers in businesses adopting flexible
work arrangements or temporary closures (P5,000 [or $100]) under its COVID-19
Adjustment Measures Program (CAMP). Singapore’s Jobs Support Scheme helped
employers to pay wages (25% of the designated first monthly wage for 9 months). Viet
Nam provided cash handouts to employees and facilitated zero-interest rate loans for
employers to pay salaries. Japan, the PRC, and the Republic of Korea also offered wage
subsidies to affected firms, especially MSMEs. Japan, Malaysia, the Philippines, the
Republic of Korea, Singapore, Thailand, and Viet Nam used cash transfers to cover
informal sector workers, the self-employed, and displaced workers.
Some Asian countries promoted new working environments by revising terms of
employment—including pay cuts and unpaid leave options (Malaysia), expediting
overtime work for COVID-19-related businesses, such as producers of face masks and
disinfection products (the Republic of Korea), and promoting electronic labor contracts
(the PRC). The PRC, the Republic of Korea, and Singapore provided labor/vocational
training subsidies for the self-employed and laid-off workers either in person or online.
The PRC set a maximum layoff rate of 20% for firms with fewer than 30 people.
Several countries discounted or waived utility payments, rental/leasing fees, and
government fees and charges. The Lao People’s Democratic Republic (Lao PDR)
revised its electricity tariff. Malaysia offered a 15% discount on monthly electricity bills
for affected businesses, such as hotels, travel agencies, shopping malls, and theme
parks. Thailand alleviated water and electricity bill payments. The PRC and Viet Nam
temporarily cut electricity prices. Malaysia reduced or waived office rental fees for MSME
retailers. The PRC, the Republic of Korea, and Singapore also waived commercial rental
fees. Singapore froze all government fees and charges for 1 year, and the PRC provided
similar waivers on administrative fees for MSMEs.
MSMEs that rely on imported goods for production benefitted from lenient customer
procedures (Cambodia and the Republic of Korea) and deferred import tax payments
with relaxed regulations (Indonesia). Some Asian countries encouraged MSMEs
to adopt remote business services. Malaysia supported agri-based MSMEs in selling
their products through e-commerce platforms. Singapore promoted digital solutions for
MSMEs to retain business operations through its enhanced Go Digital program. The
PRC and the Republic of Korea encouraged the digital transformation of MSMEs’
business, and Japan established special help desks for businesses (consultation
services).
5
To stimulate consumption during the pandemic, several Asian countries used special
expenditure programs targeting disrupted industries. Cambodia launched campaigns to
promote domestic tourism (such as the Angkor Wat complex). Indonesia implemented
stimulus packages to promote tourism (Bali) and financed social media infrastructure to
promote tourist destinations. Malaysia offered travel discount vouchers and special
income tax relief to promote domestic tourism. The Philippines and Thailand expanded
their national budgets to stimulate tourism spending.
The aim of these early policies was to help maintain people’s daily lives and business
operations as much as possible at the pre-COVID-19 levels. The large government
support packages commenced within 1 month of the start of the pandemic. However,
they also risked further bloating national budgets and deteriorating banking sector
balance sheets in the long term. Given the uncertainty about the length of the pandemic,
governments need to focus the budget allocations on those groups that are most
devastated by or vulnerable to COVID-19, including MSMEs. It is thus helpful for
governments to have a better understanding of the evolving demand-side conditions to
design the most feasible and effective policy measures using a phased approach.
This paper assesses the real conditions that MSMEs faced at the start of the
COVID-19 pandemic or during lockdowns. The rapid survey aimed to verify the evidence
for policy design to support MSMEs in the Philippines.
3. METHODOLOGY
The ADB conducted a rapid survey from 30 March to 16 April 2020 to assess the initial
1-month impact on MSMEs of the COVID-19-associated quarantine and lockdown
measures. The survey also explored possible government policy options to support
MSMEs’ needs. It was carried out online via social media (Facebook) and networks of
the Bureau of Small and Medium Enterprise Development under the Department of
Trade and Industry and the Philippine Chamber of Commerce and Industry.
The survey questionnaire consisted of four components: (i) a company profile that
identified companies’ primary business, location, operating period, employment, wage
per employee, total assets, internet penetration, and exposure to global business as
of the end of 2019; (ii) business conditions after the COVID-19 outbreak, including
changes in the business environment, sales volume, revenue, employment, wage
payments, and fiscal and funding conditions; (iii) business concerns and obstacles that
MSMEs are facing after the outbreak and related government quarantine measures; and
(iv) policy interventions that MSMEs would like to receive from the government to
maintain or restart their business. The survey set 15 March 2020 as the base date
for the COVID-19 pandemic. The MSME classification refers to the employment criterion
of the Philippine Statistics Authority (PSA): (i) a microenterprise is a firm with one to nine
employees; (ii) a small enterprise is a firm with 10–99 employees; and
(iii) a medium-sized enterprise is a firm with 100–199 employees.2
This study provides both descriptive and regression analyses to estimate the
COVID-19 impact on MSMEs at the initial stage of the pandemic, addressing the impact
by firm size and industry.
2 The Philippines uses two different criteria to classify firm size: (i) employment levels, which the Philippine
Statistics Authority (PSA) set for statistical purposes; and (ii) total assets (excluding land), which the Small
and Medium Enterprise Development Council Resolution No. 01 Series of 2003 defined. When comparing
the ADB and PSA survey data, the definition of firm size corresponds to the employment criterion set by
the PSA.
6
3.1 Data Structure

In total, the survey received 2,329 responses from companies across the Philippines, of
which 1,804 were from MSMEs that completed the survey—77.5% of the total
respondents. This paper used only completed responses for analysis.
As the respondents were not selected randomly, it is critical to compare the distribution
of the unweighted sample with an existing framework to understand the extent of bias
and identify ways to minimize its impact on the survey estimates. We thus compared the
survey data with the List of Establishments that the PSA surveyed in 2018.
Table 2A presents a breakdown of the enterprises by firm size and aggregated industrial
sector. The unweighted sample is under-represented by nearly 8 percentage points
when compared with the PSA list for microenterprises, while small enterprises are over-
represented by nearly 7 percentage points and medium-sized enterprises
by 1 percentage point. The industry data was aggregated into three broad
classifications—agriculture, manufacturing (including construction), and services. The
unweighted agriculture sample is over-represented by nearly 5 percentage points,
and manufacturing by 23 percentage points; while services are under-represented by
27 percentage points.
Table 2A: Comparison between ADB and PSA Surveys

by Firm Size and Aggregated Industrial Sector
ADB Rapid MSME Survey
on COVID-19 Impact PSA List of Difference
(Unweighted) Establishments, 2018 between
ADB and
Employment Grouping Employment Grouping PSA
Item Total Share (%) Total Share (%) Surveys (%)
By firm size, total 1,804 100.0 998,342 100.0
Micro 1,461 81.0 887,272 88.9 (7.9)
Small 318 17.6 106,175 10.6 7.0
Medium 25 1.4 4,895 0.5 0.9
By aggregated industrial sector, total 1,804 100.0 998,342 100.0
Agriculture 99 5.5 8,506 0.9 4.6
Manufacturing 629 34.9 120,580 12.1 22.8
Services 1,076 59.6 869,256 87.1 (27.4)
ADB = Asian Development Bank; MSME = micro, small, and medium-sized enterprise; PSA = Philippine Statistics
Authority.
Note: The firm size classification of the ADB survey data is based on the number of employees for comparison with the
PSA data.
Source: Asian Development Bank, Rapid Survey for the COVID-19 Impact on Micro, Small, and Medium-Sized Enterprises
in the Philippines.
A breakdown of the ADB survey data and PSA sampling frame by industry shows the
differences in a more granular manner (Table 2B). By industry, the difference in each
sector’s share of the total respondents between the ADB and PSA data distribution was
less than 5 percentage points (the vast majority had less than 1%), except for
manufacturing (20 percentage points above the PSA distribution) and wholesale and
retail trade (21 percentage points below the PSA distribution).
The geographic distribution of firms responding to the ADB survey reveals another layer
of bias (Table 2B). Of the 17 regions, 10 were under-represented in the ADB samples
relative to the PSA list. The difference in each region’s share of the total respondents
between the ADB and PSA data distribution was less than 5 percentage points, except
7
for Region IV-A (Calabarzon) (10.3 percentage points above the PSA distribution),
Region X (Northern Mindanao) (5.6 percentage points above the PSA distribution), and
the National Capital Region (NCR) (5 percentage points below the PSA distribution).
3.2 Weighting the Survey Data

For field surveys, a sample is typically drawn from an existing sampling frame that
uses a well-defined sampling strategy and includes details on the distribution of the
population. To increase the efficiency of the sample design for the survey, it is necessary
to divide the sample into strata that are as homogeneous as possible. The number of
geographic domains of analysis, the industrial classification, and enterprise type (firm
size) are the three main determinants of the sample size and allocation as a minimum
level of precision is necessary for each region. This also affects the sampling efficiency
because the lower number of firms per primary statistical unit (PSU), especially by region,
implies the need to enumerate more sample PSUs. The sample should be designed to
provide nationally representative estimates and, if desired,
high-precision estimates at the regional level for the key indicators. Ideally, the sample
will be representative with respect to all the variables to estimate from the survey data.
However, there are several challenges to utilizing an optimal sampling design in an
online survey due to self-selection bias. This typically occurs when respondents have a
choice of filling out or not filling out a survey, which is often the case in online surveys.
Furthermore, some respondents may not have access to the Internet, which could lead
to over- or under-representation of certain subgroups within a sample. Given these
problems, it is difficult to draw unbiased inferences from the observed survey data, and
the only way to proceed is to make corrections for the lack of representativeness.
A popular correction technique to minimize the bias is weighting adjustments.
Researchers assign larger weights to under-represented groups and smaller weights to
over-represented groups, making it possible to derive weighted estimates of key
indicators.
For this study, the PSA made available key statistics from its 2018 List of Establishments
(LE), which includes both registered and unregistered enterprises and forms the basis of
the sampling frame for the ADB survey data. The statistical unit of the LE is the
establishment. It defines an establishment as an economic unit, which engages, under
single ownership or control (under a single legal entity), in one or predominantly one kind
of economic activity at a single fixed physical location. The operational definition of an
establishment is a unit that engages in the production of
the most homogeneous group of goods and services, usually at one location but
sometimes over a wider area, for which separate records are available that provide data
concerning the production of the goods and services and the materials, labor, and
physical resources that it uses in production.
For weighting adjustments, an important requirement is to have auxiliary variables, which
are measured in the survey but also available for the total population (for
the sampling frame). The ADB survey data contain information on the number of
establishments by a combination of variables (municipality, industry, and firm size) for
the NCR and a combination of variables (province, industry, and firm size) for non-NCR
firms. For each possible combination, the total number of establishments in the LE and
the number of establishments surveyed are observed.
8
Table 2B: Comparison between ADB and PSA Surveys—By Industrial Sector
and Region
ADB Rapid MSME Survey on COVID-19
Impact
Employment Grouping
Item Micro Small Medium Total Share (%)
By industrial sector, total 1,461 318 25 1,804 100.0
A - Agriculture, forestry, and fishing 79 18 2 99 5.5
B - Mining and quarrying – – – – –
C - Manufacturing 458 111 6 575 31.9
D - Electricity, gas, steam, and air conditioning supply 11 1 1 13 0.7
E - Water supply; sewerage, waste management, and remediation activities – – – – –
F - Construction 36 15 3 54 3.0
G - Wholesale and retail trade; repair of motor vehicles and motorcycles 393 55 1 449 24.9
H - Transport and storage 21 6 – 27 1.5
I - Accommodation and food service activities 208 41 4 253 14.0
J - Information and communication 29 13 1 43 2.4
K - Financial and insurance activities 6 12 1 19 1.1
L - Real estate activities 11 2 – 13 0.7
M - Professional, scientific, and technical activities 34 9 1 44 2.4
N - Administrative and support service activities 63 15 1 79 4.4
O - Public administration and defense; compulsory social security – – – – –
P - Education 8 6 3 17 0.9
Q - Human health and social work activities 7 1 – 8 0.4
R - Arts, entertainment, and recreation – – – – –
S - Other service activities 97 13 1 111 6.2
By region, total 1,461 318 25 1,804 100.0
National Capital Region (NCR) 210 63 5 278 15.4
Cordillera Administrative Region (CAR) 11 5 – 16 0.9
Region I (Ilocos Region) 144 21 1 166 9.2
Region II (Cagayan Valley) 58 10 – 68 3.8
Region III (Central Luzon) 128 25 1 154 8.5
Region IV-A (CALABARZON) 394 57 3 454 25.2
MIMAROPA Region 41 11 – 52 2.9
Region V (Bicol Region) 103 16 4 123 6.8
Region VI (Western Visayas) 33 12 3 48 2.7
Region VII (Central Visayas) 88 41 4 133 7.4
Region VIII (Eastern Visayas) 26 3 1 30 1.7
Region IX (Zamboanga Peninsula) 25 3 1 29 1.6
Region X (Northern Mindanao) 133 34 1 168 9.3
Region XI (Davao Region) 21 8 1 30 1.7
Region XII (SOCCSKSARGEN) 23 6 – 29 1.6
Region XIII (Caraga) 22 3 – 25 1.4
Autonomous Region in Muslim Mindanao (ARMM) 1 – – 1 0.1
9
Table 2B continued
Difference
PSA List of Establishments, 2018 between
Employment Grouping ADB and
PSA
Surveys
Item Micro Small Medium Total Share (%) (%)
By industrial sector, total 887,272 106,175 4,895 998,342 100.0
A - Agriculture, forestry, and fishing 5,837 2,512 157 8,506 0.9 4.6
B - Mining and quarrying 492 302 21 815 0.1 (0.1)
C - Manufacturing 103,590 11,678 1,067 116,335 11.7 20.2
D - Electricity, gas, steam, and air conditioning 478 633 98 1,209 0.1 0.6
supply
E - Water supply; sewerage, waste management, 677 711 49 1,437 0.1 (0.1)
and remediation activities
F - Construction 2,304 1,715 226 4,245 0.4 2.6
G - Wholesale and retail trade; repair of motor 427,101 33,577 1,087 461,765 46.3 (21.4)
vehicles and motorcycles
H - Transport and storage 7,264 3,511 231 11,006 1.1 0.4
I - Accommodation and food service activities 125,396 18,802 337 144,535 14.5 (0.5)
J - Information and communication 27,421 1,973 153 29,547 3.0 (0.6)
K - Financial and insurance activities 37,813 8,053 167 46,033 4.6 (3.6)
L - Real estate activities 9,478 1,975 79 11,532 1.2 (0.4)
M - Professional, scientific, and technical activities 13,617 2,164 104 15,885 1.6 0.8
N - Administrative and support service activities 14,073 3,022 474 17,569 1.8 2.6
O - Public administration and defense; compulsory – – – – – –
social security
P - Education 9,105 8,312 391 17,808 1.8 (0.8)
Q - Human health and social work activities 26,076 2,325 200 28,601 2.9 (2.4)
R - Arts, entertainment, and recreation 13,755 1,563 34 15,352 1.5 (1.5)
S - Other service activities 62,795 3,347 20 66,162 6.6 (0.5)
By region, total 887,272 106,175 4,895 998,342 100.0
National Capital Region (NCR) 166,921 34,523 1,868 203,312 20.4 (5.0)
Cordillera Administrative Region (CAR) 18,783 1,587 47 20,417 2.0 (1.2)
Region I (Ilocos Region) 46,708 3,977 122 50,807 5.1 4.1
Region II (Cagayan Valley) 28,547 2,119 52 30,718 3.1 0.7
Region III (Central Luzon) 104,875 10,754 444 116,073 11.6 (3.1)
Region IV-A (CALABARZON) 133,640 13,778 778 148,196 14.8 10.3
MIMAROPA Region 21,948 1,914 57 23,919 2.4 0.5
Region V (Bicol Region) 37,111 3,215 118 40,444 4.1 2.8
Region VI (Western Visayas) 55,482 5,894 214 61,590 6.2 (3.5)
Region VII (Central Visayas) 61,176 8,775 444 70,395 7.1 0.3
Region VIII (Eastern Visayas) 28,324 2,355 70 30,749 3.1 (1.4)
Region IX (Zamboanga Peninsula) 30,888 2,216 73 33,177 3.3 (1.7)
Region X (Northern Mindanao) 33,040 4,079 155 37,274 3.7 5.6
Region XI (Davao Region) 52,449 5,758 252 58,459 5.9 (4.2)
Region XII (SOCCSKSARGEN) 41,581 3,121 120 44,822 4.5 (2.9)
Region XIII (Caraga) 18,069 1,687 67 19,823 2.0 (0.6)
Autonomous Region in Muslim Mindanao (ARMM) 7,730 423 14 8,167 0.8 (0.8)
ADB = Asian Development Bank; MSME = micro, small, and medium-sized enterprise; PSA = Philippine Statistics
Authority.
Notes: The firm size classification of the ADB survey data is based on the number of employees for comparison with the
PSA data. The ADB survey data are unweighted.
Source: Asian Development Bank, Rapid Survey for the COVID-19 Impact on Micro, Small, and Medium-Sized Enterprises
in the Philippines.
10
A simple correction of the existing bias induced by self-selection and non-response

involves dividing the population percentage by the response percentage for each
combination to achieve representativity by location, industry, and firm size. It should
be noted that the level of aggregation differs between NCR and non-NCR firms
simply because more disaggregated auxiliary data are available for NCR firms (at the
municipality level) than non-NCR firms (at the provincial level).
This weighting adjustment may not eliminate all the biases in the estimated parameters
but helps to minimize self-selection and non-response bias to some extent. For instance,
because data for all combinations of location–industry–firm size are not available in the
sample, some biases will persist, although the extent may not be statistically significant
as they are likely to be less prevalent in the LE anyway.3 If the unweighted and weighted
results are comparable, then self-selection and non-response bias may not be serious in
the survey data.
3.3 Analytical Approach

The argument on the choice of econometric modeling in the presence of a binary
dependent variable has been explored in detail in economics literature. Generally, it has
compared two approaches: (i) the linear probability model (LPM) and (ii) probit and
logistic models. The LPM allows the fitting of the data using a simple linear regression
following the least squares approach. In contrast, probit and logistic regressions are
drawn from the standard normal cumulative distribution function or the cumulative
distribution function drawn from a logistic random variable.
The probit and logistic regressions were frequently recommended due to some
shortcomings of the LPM: (i) the possibility of obtaining estimated values less than
zero or greater than one, which is inconsistent within the context of a limited dependent
variable, which in this case falls between zero and one; and (ii) the LPM yields
inconsistent estimates (Amemiya 1977 4). In fact, Horrace and Oaxaca (2006) elucidated
the second point by stating that using ordinary least squares (OLS) on the LPM is
theoretically possible to yield an unbiased estimation but requires fortuitous
circumstances. They added that consistency seems to be exceedingly rare, as one would
have to accept extraordinary restrictions on the joint distribution of the regressors,
concluding that OLS is frequently a biased estimator and an inconsistent estimator of the
LPM. They continued by suggesting that the only situations in which the LPM might make
sense is when there are endogenous dummies or when using panel data. Within the
context of measurement error associated with a binary dependent variable,
misclassification may arise with some values coded as zero when they are in reality
equal to one or vice versa. Hausman, Abrevaya, and Scott-Morton (1998) investigated
the implications of such an error for the logit and probit models as well as the LPM and
found the implications to be more severe for the LPM than for their non-linear
counterparts.
3 If the PSA had made the microdata for the LE available, we could have pursued further attrition correction
in addition to the probability weighting that we have already implemented. We would have applied two
steps: (i) a propensity score adjustment, which uses the available characteristics of the firms from the LE
(age, gender of firm owner, location, sector, type, etc.) to account for unit non-response, and the inverse
of this probability; and (ii) a post-stratification adjustment drawn from the LE inflated for industrial growth.
4 Amemiya (1985) showed that there are some correspondences among LPM, probit, and logit estimators,
though.
11
In contrast, Angrist and Pischke (2008) made several arguments in favor of the LPM.
The first is that, while probit and logistic regressions may fit the limited dependent
variable bounds, it is the marginal effects that analysts typically care about in these non-
linear settings. Through an empirical example, they showed that these marginal effects
are close to the LPM’s estimated coefficients irrespective of whether the conditional
expectation function is linear or non-linear. The second argument is that non-linear
models are less computationally tractable, messier to interpret, and less transparent,
particularly in the light of weighting, instrumental variables, and panel data. The finding
of Angrist and Pischke (2008) is also consistent with that of Wooldridge (2002, 455),
which states that “If the main purpose is to estimate the partial effect of the independent
variable on the response probability, averaged across the distribution of [the independent
variable], then the fact that some predicted values are outside the unit interval may not
be very important.”
Taking account of all the above, this study chose the LPM for the following reasons. First,
probit and logistic models rely on several strong assumptions with respect to error terms,
which may not always hold. In addition, probit and logistic models are difficult to interpret
and issues arise when justifying the results. Meanwhile, the LPM is convenient and
easier to interpret, computationally less intensive, and reveals similar marginal effects to
its non-linear counterparts, as Angrist and Pischke (2008) showed.
3.4 Linear Probability Model

To assess the first 1-month impact on MSME operations, employment, wage payments,
and fiscal and funding conditions after the COVID-19 pandemic and associated
government containment measures, we estimated eight LPMs with six independent
variables: industrial sector, business location, operating period, business ownership
(gender), global business exposure, and firm size (employment grouping). The survey
data were weighted in accordance with the 2018 PSA List of Establishments.
Yi = α + β Indi + δ Regi + ϕ Opsi + ψ Womi + η GVCi + τ MSMEi + ϵ (1)
In this model, Y includes five areas with eight dimensions (models) that measure the
level of a firm’s resilience to the COVID-19 pandemic and the associated government
measures (Table 3). 𝑌𝑌𝑖𝑖 in each model is a separate binary dependent variable for each
observed firm i; 𝐼𝐼𝐼𝐼𝑑𝑑𝑖𝑖 is the vector of categories for industry classification with
“agriculture” as the base; 𝑅𝑅𝑅𝑅𝑔𝑔𝑖𝑖 is the vector of categories for business location with
“NCR” as the base; 𝑂𝑂𝑂𝑂𝑠𝑠𝑖𝑖 is the vector of categories for years of operation with “0–5 years”
as the base; 𝑊𝑊𝑊𝑊𝑚𝑚𝑖𝑖 is a binary variable that takes the value one if the owner of the
establishment is a “woman” and zero if the owner is a “man”; 𝐺𝐺𝐺𝐺𝐶𝐶𝑖𝑖 is a binary variable
that takes the value one if the establishment is involved in a global supply chain or
export/import business and zero otherwise; 𝑀𝑀𝑀𝑀𝑀𝑀𝐸𝐸𝑖𝑖 is the vector of categories for
enterprise classification with “medium-sized firm” as the base; and ϵ is a residual.
12
Table 3: Model Specification

Area (5) Dimension (8) Definition
1. Monthly sales Sales 1 Firm’s sales condition 1. Totally no sales in March 2020
(after the imposition of ECQ) or not.
Sales 2 Firm’s sales condition 2. A sales decrease in March as
compared with February 2020 (before the imposition of
ECQ) or not.
2. Monthly revenue Revenue 1 Firm’s income/revenue condition 1. Totally no
income/revenue in March 2020 or not.
Revenue 2 Firm’s income/revenue condition 2. An income/revenue
decrease in March 2020 from the previous month or not.
3. Employment Employment Firm’s employment condition assessed by a decrease or
increase in employees (including no change) in March 2020
from the previous month or not.
4. Wage payments Wage 1 Firm’s wage/salary payment condition to employees 1.
Totally no wage payments to employees in March 2020 or
not.
Wage 2 Firm’s wage/salary payment condition to employees 2. A
decrease in the total wage payments in March 2020 from the
previous month or not.
5. Financial condition Finance Firm’s financial condition assessed as already having no
cash/savings or running out of cash/funds in a month at the
time of the survey.
ECQ = enhanced community quarantine.
Notes: The survey set 15 March 2020 as the base date of the COVID-19 pandemic. The ECQ began on 16 March and
ended on 31 May 2020.
Source: Authors.
4. PROFILE OF THE SURVEYED MSMES

By firm size, 81.0% of the MSME respondents (1,804 firms) or 1,461 firms were owners
of microenterprises, followed by small enterprises (17.6% or 318 firms) and medium-
sized enterprises (1.4% or 25 firms). 5 By region, the top response rates were largely
concentrated in Luzon, with Region IV-A (Calabarzon) (25.2%), the NCR (15.4%),
Region I (Ilocos) (9.2%), and Region III (Central Luzon) (8.5%), accounting for nearly
58% of the responses. For Visayas, Region VII (Central Visayas) accounted for 7.4% of
the responses, while in Mindanao, Region X (Northern Mindanao) contributed the most,
with 9.3% of the total sample.
Based on the three broad industry classifications (Table 4), 59.6% of the surveyed
MSMEs fell into the service category, followed by 34.9% in manufacturing. Agriculture
accounted for 5.5% of the respondents. Wholesale and retail trade (24.9%) and
accommodation and food services (14.0%) were the second- and third-largest sectors
among the respondents.
5 We caution readers of this article to interpret the statistics on medium-sized enterprises carefully given
the small sample size.
13
Table 4: Industry Classification

Board Industry
Classification PSA Industry Classification
Agriculture Agriculture, forestry, and fishing
Manufacture Manufacturing
Construction
Services Mining and quarrying
Electricity, gas, steam, and air conditioning supply
Water supply; sewerage, waste management, and remediation activities
Wholesale and retail trade; repair of motor vehicles and motorcycles
Transport and storage
Accommodation and food service activities
Information and communication
Financial and insurance activities
Real estate activities
Professional, scientific, and technical activities
Administrative and support service activities
Public administration and defense; compulsory social security
Education
Human health and social work activities
Arts, entertainment, and recreation
Other service activities
PSA = Philippine Statistics Authority.
Source: Authors.
We found that 59.6% of the surveyed MSMEs had been in operation for between 0 and
5 years (young start-ups mostly belonging to microenterprises and in services), followed
by those operating for 6–10 years (18.8%), 11–15 years (8.5%), 16–30 years (9.6%),
and over 31 years (3.4%).
More than half (56.1%) of the surveyed MSMEs had a female head. By firm size, women-
led MSMEs accounted for 58.0% of microenterprises, 50.3% of small enterprises, and
16.0% of medium-sized enterprises. By sector, they accounted for 58.1% of
manufacturing-related enterprises, followed by service-oriented firms (56.0%). The
percentage of female ownership was lower for agriculture, with women owning 44.4% of
agricultural enterprises.
As the enterprise size increases, so does the percentage of women employed. Based
on the responses, 28.0% of microenterprises had female employees of at least 51% of
the firm’s workforce. This share grew when it came to small enterprises (34.3%) and
medium-sized enterprises (40.0%). In contrast, 54.5% of microenterprises, 23.0%
of small enterprises, and 4.0% of medium-sized enterprises reported that female
comprised less than 10% of the workforce. By industry, nearly half the sample for all
three sectors (agriculture [48.5%], manufacturing [49.7%], and services [47.4%]) had
less than 10% female employees. The percentage of firms with more than 50% female
employees was 30.5% for services, 28.7% for manufacturing, and 20.2% for agriculture.
By wage structure, more than half (60.4%) of firms reported average monthly wages
of not more than P9,000. This was followed by 35.1% with average monthly wages
ranging between P9,001 and P18,000. Firms with average monthly wages greater
than P18,001 accounted for less than 5% of the sample. Notably, more than
four-fifths of enterprises in Region I (83.7%) and Region IX (Zamboanga Peninsula,
82.8%) reported average wages of less than P9,000. By comparison, only the NCR,
14
Region IV-A, Region V (Bicol), and Region X had enterprises reporting monthly wages
of over P27,000.
The average monthly wage tends to increase as the enterprise size increases. The share
of enterprises with average wages greater than P18,000 increased from 2.9%
for microenterprises to 10.4% for small and 20.0% for medium-sized enterprises.
Meanwhile, the share of enterprises with an average wage of not more than P9,000 fell
from 67.6% for micro to 29.9% for small and just 24.0% for medium-sized enterprises. A
similar trend was observable by sector. The share of enterprises with an average wage
of less than P9,000 fell from 71.7% for agriculture to 67.2% for manufacturing and 55.3%
for services. However, in all three sectors, more than half of the enterprises reported
average wages of not more than P9,000.
Nearly three-fourths (74.0%) of the respondents indicated that they used the Internet
in their daily business. The Internet use was lower for microenterprises (71.6%) than
for small (84.0%) and medium-sized (88.0%) enterprises. Agricultural firms reported the
lowest use of the Internet in their daily business with 62.6%. However, relative
to the national average, lower Internet penetration was observable in Region XII
(SOCCSKSARGEN, 48.3%), Region I (65.7%), Region XI (Davao, 66.7%), Region X
(67.3%), Region II (Cagayan Valley, 67.6%), and Region V (69.1%). Higher penetration
rates were apparent in Region VII (84.2%), the NCR (83.8%), and Region III (83.8%). 6
Less than one-tenth (9.8%) of the responding enterprises reported that they were
involved in a global value chain (GVC). The percentage of medium-sized enterprises
(28.0%) was higher than those of small (20.8%) and microenterprises (7.1%). A similar
pattern was also evident across industries. Manufacturing had the largest GVC share
(11.5%), followed by agriculture (10.1%) and services (8.8%).
5. FINDINGS FROM THE DESCRIPTIVE STATISTICS

AND ECONOMETRIC ANALYSES
The study offers a descriptive analysis based on the survey findings and uses the LPM
to estimate the impact on MSME operations, employment, and fiscal conditions. It
weighted the survey data based on the PSA list of establishments by firm size, sector,
and location.
The survey found that an average 73.1% of MSMEs were forced to close their business
a few weeks after the COVID-19 outbreak and the lockdown measures implemented.
This was more pronounced for small firms (76.4%) and those in services (72.7%) (Figure
1). They immediately faced delays in the delivery of products and services (average
50.5%, especially for medium-sized firms [71.8%] and agriculture [49.7%]), disrupted
supply chains (average 47.3%, especially in medium-sized firms [68.7%] and
manufacturing [54.9%]), and a sharp drop in the domestic demand (average 35.2%,
especially in small firms [39.1%] and manufacturing [41.8%]). Only 2.3% of MSMEs
on average reported no change in the business environment after the outbreak.
Meanwhile, some groups of MSMEs (average 9.3%) reported a better business
environment than before the pandemic, especially in microenterprises (14.5%) and
agriculture (20.7%), due to the increased demand from households and firms for
essential goods and services and healthcare.
6 We removed the Autonomous Region in Muslim Mindanao (ARMM) from this ranking as only one
enterprise responded.
15
Figure 1: Business Conditions during the Pandemic

A. By Firm Size B. By Sector
MSME = micro, small, and medium-sized enterprise.

Notes: 1,804 valid samples in the Philippines. Calculation based on weighted data from the rapid MSME survey in the
Philippines, 30 March–16 April 2020.
Table 5 shows the LPM estimate results based on the weighted data. Model (1) was
carried out in five areas (sales, revenue, employment, wages, and finance) with eight
dimensions that affect firms’ resilience to the COVID-19 pandemic and associated
government measures.
The eight dimensions are binary dependent variables:
a. sales1 denotes a dummy variable taking the value one for a firm with no sales in
March 2020 and zero for a firm with sales;
b. sales2 denotes a dummy variable taking the value one for a firm with a sales
decrease in March compared with February 2020 and zero for a firm with a sales
increase or no change;
c. revenue1 denotes a dummy variable taking the value one for a firm with no
income/revenue in March 2020 and zero for a firm with income/revenue;
d. revenue2 denotes a dummy variable taking the value one for a firm with an
income/revenue decrease in March compared with February 2020 and zero for a
firm with an income/revenue increase or no change;
e. employment denotes a dummy variable taking the value one for a firm with a
decrease in the number of employees in March compared with February 2020
and zero for a firm with an increase or no change in the number of employees;
f. wage1 denotes a dummy variable taking the value one for a firm with no wage
payments to employees after the COVID-19 outbreak (15 March 2020 as the
base date) and zero for a firm that has paid wages to employees;
g. wage2 denotes a dummy variable taking the value one for a firm with a decrease
in the total wage payments to employees after the virus outbreak and zero for a
firm with an increase or no change in wage payments; and
h. finance denotes a dummy variable taking the value one for a firm with no
cash/savings at the time of the survey or running out of cash/funds in a month
and zero for a firm that reported having enough savings, liquid assets, and other
contingency finance to maintain business at the time of the survey.
16
Table 5: COVID-19 Impact on MSMEs—Linear Probability Model

(1) (2) (3) (4)
Variables sales1 sales2 revenue1 revenue2
Industry (base—agriculture, forestry, and fishing)
Manufacturing 0.0286 0.09650* 0.0597 0.0533
[0.0982] [0.0580] [0.0910] [0.0435]
Transportation and storage 0.1259 0.0785 0.0630 –0.0573
[0.1493] [0.0689] [0.1532] [0.0964]
Power and energy (e.g., electricity and gas) –0.4098*** 0.0949 –0.4256*** 0.0314
[0.1564] [0.0597] [0.1434] [0.0446]
Construction 0.3979*** 0.0686 0.3686*** 0.0236
[0.1017] [0.0638] [0.0983] [0.0513]
Wholesale and retail trade –0.0061 0.0714 0.0164 0.0214
[0.0930] [0.0587] [0.0867] [0.0444]
Information and communication technology 0.0752 0.0830 0.0260 –0.0185
[0.1376] [0.0734] [0.1352] [0.0758]
Accommodation and food service activities 0.1492 0.0937 0.1808** 0.0397
[0.0975] [0.0576] [0.0913] [0.0439]
Financial and insurance activities –0.1350 –0.2442 0.0413 –0.0941
[0.1956] [0.1938] [0.1926] [0.1856]
Real estate activities 0.0669 –0.1867 –0.0504 –0.3467*
[0.2670] [0.2292] [0.2467] [0.2101]
Professional, scientific, and technical activities 0.0281 0.0213 0.0087 0.0082
[0.1348] [0.0758] [0.1290] [0.0605]
Administrative and support service activities 0.1505 0.0960 0.1685 0.0463
[0.1139] [0.0618] [0.1110] [0.0492]
Education 0.4791*** 0.0754 0.4851*** 0.0372
[0.1135] [0.0624] [0.1051] [0.0490]
Human health and social work activities –0.0650 0.1123* –0.0325 0.0665
[0.2215] [0.0602] [0.2200] [0.0456]
Other service activities 0.1376 0.1145** 0.1196 0.0717
[0.1118] [0.0579] [0.1064] [0.0437]
Location (base—National Capital Region (NCR))
Autonomous Region in Muslim Mindanao (ARMM) –0.6801*** 0.04346* 0.3516*** 0.05893**
[0.0611] [0.0248] [0.0614] [0.0269]
Cordillera Administrative Region (CAR) 0.1595 0.0327 0.1793 0.05012**
[0.1215] [0.0232] [0.1246] [0.0255]
MIMAROPA Region –0.0877 0.0187 –0.0558 0.05500*
[0.1300] [0.0384] [0.1288] [0.0304]
REGION I (Ilocos Region) –0.1008 0.0118 –0.0158 0.0213
[0.0907] [0.0264] [0.0918] [0.0280]
REGION II (Cagayan Valley) –0.3472*** –0.0440 –0.3865*** –0.0319
[0.1242] [0.0762] [0.1098] [0.0768]
REGION III (Central Luzon) 0.0953 0.0052 0.0419 0.0181
[0.0752] [0.0401] [0.0799] [0.0409]
REGION IV-A (CALABARZON) –0.0779 0.0006 –0.0641 0.0007
[0.0610] [0.0322] [0.0609] [0.0346]
REGION V (Bicol Region) –0.1977 0.0172 –0.1864 0.0067
[0.1588] [0.0326] [0.1655] [0.0383]
17
Table 5 continued
(1) (2) (3) (4)
Variables sales1 sales2 revenue1 revenue2
REGION VI (Western Visayas) –0.1274 0.0447 –0.1091 0.0565
[0.1173] [0.0409] [0.1180] [0.0412]
REGION VII (Central Visayas) –0.1064 –0.0357 –0.0890 –0.0117
[0.0833] [0.0434] [0.0827] [0.0440]
REGION VIII (Eastern Visayas) –0.1715 0.0595 –0.1619 0.07255*
[0.1943] [0.0407] [0.1984] [0.0415]
REGION IX (Zamboanga Peninsula) 0.0568 0.0108 0.0585 0.0222
[0.1338] [0.0273] [0.1338] [0.0286]
REGION X (Northern Mindanao) –0.1727** –0.0105 –0.1281* 0.0298
[0.0785] [0.0370] [0.0775] [0.0295]
REGION XI (Davao Region) –0.0609 –0.0060 –0.0437 –0.0222
[0.1385] [0.0422] [0.1399] [0.0514]
REGION XII (SOCCSKARGEN) –0.4667*** –0.0058 –0.1704 0.0029
[0.1058] [0.0415] [0.1891] [0.0423]
REGION XIII (Caraga) –0.4919*** 0.0246 –0.4762*** 0.03542*
[0.1291] [0.0196] [0.1329] [0.0212]
Operating Period (base—0–5 years)
6–10 years –0.0368 0.0112 0.0353 0.0121
[0.0551] [0.0215] [0.0618] [0.0222]
11–15 years 0.0368 0.0232 0.0735 0.0274
[0.0738] [0.0252] [0.0748] [0.0253]
16–30 years –0.0326 0.0222 –0.0069 0.0097
[0.0786] [0.0311] [0.0771] [0.0332]
31 years and above –0.0382 0.0251 0.0105 0.02759*
[0.1257] [0.0165] [0.1193] [0.0162]
Gender of Owner (base—male owner)
Woman 0.09479** 0.0127 0.07358* 0.0043
[0.0414] [0.0207] [0.0437] [0.0215]
Involvement in GVC (base—non-GVC firms)
GVC firms 0.0019 –0.1092* 0.0378 –0.1045*
[0.0764] [0.0615] [0.0749] [0.0619]
Enterprise Classification (base—medium enterprise)
Microenterprise 0.2246* 0.1424 0.2791** 0.1580
[0.1275] [0.1301] [0.1305] [0.1341]
Small enterprise 0.0383 0.2012 0.1037 0.2080
[0.1295] [0.1283] [0.1327] [0.1325]
Constant 0.3668** 0.7301*** 0.2792* 0.7573***
[0.1628] [0.1415] [0.1616] [0.1421]
N 1,804 1,804 1,804 1,804
Pseudo R-square 0.1215 0.0826 0.0750 0.0594
18
Table 5 continued
(5) (6) (7) (8)
Variables employment wage1 wage2 finance
Industry (base—agriculture, forestry, and fishing)
Manufacturing –0.0995 0.1576* 0.2275*** 0.0495
[0.0852] [0.0920] [0.0821] [0.0734]
Transportation and storage –0.1678 0.1681 0.1536 0.1842**
[0.1681] [0.1473] [0.1238] [0.0830]
Power and energy (e.g., electricity and gas) 0.1369 –0.4952*** 0.1471 –0.0375
[0.2363] [0.1267] [0.1490] [0.2145]
Construction –0.3341*** 0.1053 0.0831 0.1831**
[0.1164] [0.1373] [0.1205] [0.0759]
Wholesale and retail trade –0.2400*** 0.0815 0.0771 0.0103
[0.0809] [0.0886] [0.0815] [0.0705]
Information and communication technology –0.3907*** 0.1347 –0.0345 –0.2158*
[0.1065] [0.1326] [0.1298] [0.1269]
Accommodation and food service activities –0.2073** 0.2621*** 0.2184*** 0.1379*
[0.0904] [0.0916] [0.0826] [0.0714]
Financial and insurance activities –0.4296*** 0.1504 0.3438*** 0.1548
[0.1324] [0.1894] [0.1008] [0.1060]
Real estate activities –0.4916*** –0.1943 –0.3451* –0.0466
[0.1649] [0.1904] [0.1968] [0.2131]
Professional, scientific, and technical activities –0.4559*** –0.0426 –0.1884 –0.0440
[0.0912] [0.1234] [0.1251] [0.1171]
Administrative and support service activities –0.3757*** 0.0989 0.0892 0.1054
[0.0902] [0.1121] [0.1065] [0.0937]
Education –0.0495 0.4664*** 0.3134** 0.1810
[0.2069] [0.1424] [0.1447] [0.1273]
Human health and social work activities –0.4450*** –0.0911 –0.2071 0.2514***
[0.1477] [0.2214] [0.2071] [0.0757]
Other service activities –0.2988*** 0.1197 0.0638 0.0274
[0.0969] [0.1123] [0.1084] [0.0919]
Location (base—National Capital Region (NCR))
Autonomous Region in Muslim Mindanao (ARMM) 0.7935*** –0.5639*** 0.2506*** 0.1840***
[0.0598] [0.0639] [0.0537] [0.0482]
Cordillera Administrative Region (CAR) 0.5829*** 0.2543* 0.1441 0.0653
[0.1579] [0.1305] [0.0884] [0.1010]
MIMAROPA Region 0.0224 0.1594 0.0059 –0.1261
[0.1069] [0.1199] [0.1196] [0.1237]
REGION I (Ilocos Region) 0.1020 0.0455 0.0524 0.0234
[0.0759] [0.0896] [0.0641] [0.0616]
REGION II (Cagayan Valley) 0.4326*** –0.0213 0.0474 0.0643
[0.1287] [0.1379] [0.1093] [0.0885]
REGION III (Central Luzon) 0.1572* 0.1051 0.0481 0.0067
[0.0829] [0.0829] [0.0713] [0.0683]
REGION IV-A (CALABARZON) 0.0650 0.0641 –0.0035 –0.0177
[0.0556] [0.0607] [0.0531] [0.0503]
REGION V (Bicol Region) –0.0670 –0.1703 0.0129 0.1697**
[0.1030] [0.1614] [0.0866] [0.0829]
19
Table 5 continued
(5) (6) (7) (8)
Variables employment wage1 wage2 finance
REGION VI (Western Visayas) –0.0739 0.0004 0.0412 0.0736
[0.0961] [0.1146] [0.0972] [0.0870]
REGION VII (Central Visayas) 0.1005 0.0369 0.0309 –0.0301
[0.0759] [0.0835] [0.0689] [0.0647]
REGION VIII (Eastern Visayas) 0.3136* –0.0381 0.1798** 0.0418
[0.1720] [0.1938] [0.0702] [0.1391]
REGION IX (Zamboanga Peninsula) –0.0947 –0.0089 –0.1797 –0.1962
[0.1200] [0.1643] [0.1507] [0.1801]
REGION X (Northern Mindanao) 0.0800 0.0440 0.0031 0.0117
[0.0725] [0.0779] [0.0657] [0.0584]
REGION XI (Davao Region) 0.3134** –0.0482 –0.0253 0.0325
[0.1573] [0.1443] [0.1175] [0.0888]
REGION XII (SOCCSKARGEN) 0.3754** –0.1251 –0.0946 0.0948
[0.1691] [0.2072] [0.1753] [0.0705]
REGION XIII (Caraga) 0.3073 –0.4973*** –0.1992 0.2053**
[0.1946] [0.0952] [0.1665] [0.0829]
Operating Period (base—0–5 years)
6–10 years 0.0332 –0.0173 –0.0335 –0.0166
[0.0577] [0.0631] [0.0540] [0.0424]
11–15 years 0.0388 0.0881 0.1297** –0.1511*
[0.0792] [0.0833] [0.0643] [0.0852]
16–30 years –0.0609 –0.0465 –0.0845 –0.0711
[0.0642] [0.0805] [0.0779] [0.0704]
31 years and above –0.0889 –0.0614 –0.1641 –0.1617
[0.1090] [0.1300] [0.1127] [0.1140]
Gender of Owner (base—male owner)
Woman –0.0126 0.0562 0.0325 0.0149
[0.0421] [0.0445] [0.0389] [0.0339]
Involvement in GVC (base—non-GVC firms)
GVC firms 0.0897 –0.0075 –0.0928 –0.0729
[0.0646] [0.0796] [0.0734] [0.0729]
Enterprise Classification (base—medium enterprise)
Microenterprise 0.0505 0.1322 –0.0759 0.0930
[0.1441] [0.1708] [0.1128] [0.1402]
Small enterprise 0.1779 –0.1268 –0.1734 0.0739
[0.1454] [0.1723] [0.1156] [0.1401]
Constant 0.4086** 0.2940 0.7157*** 0.6979***
[0.1711] [0.1991] [0.1446] [0.1584]
N 1,804 1,804 1,804 1,804
Pseudo R-square 0.1280 0.0791 0.0831 0.0727

Note: Robust standard errors in brackets. *** p<0.01, ** p<0.05, * p<0.10.
Source: Authors’ calculation based on the Asian Development Bank Rapid Survey for the COVID-19 Impact on Micro,
Small, and Medium-Sized Enterprises in the Philippines, 30 March–16 April 2020.
20
5.1 Impact on MSME Sales

The survey found that 59.9% of microenterprises had no sales in March 2020 due to the
temporary closure of business, followed by small firms (44.8%) and medium-sized firms
(35.8%), suggesting a more serious impact on microenterprises (Figure 2). MSMEs that
operated continuously during the lockdown faced a significant drop in sales: 27.7% of
micro, 43.6% of small, and 41.0% of medium-sized firms experienced a sales decrease
of over 30% in March from February (the month before the lockdown). All three industrial
sectors—services, manufacturing, and agriculture—also had no sales immediately after
the lockdown (more than half of MSMEs in each sector). However, some MSMEs had a
sales increase due to special demand during the lockdown, such as retail trade offering
daily goods and food.
Figure 2: Sales in March 2020

70% 70%
59.9%
60% 60% 59.1%
57.4%
50.5%
50% Zero (temporarily closed) 50% Zero (temporarily closed)
44.8% More than 30% decrease
43.6% More than 30% decrease
41.0% 21%–30% decrease 21%–30% decrease
40% 40%
35.8% 11%–20% decrease 35.4% 11%–20% decrease
1%–10% decrease 1%–10% decrease
30% 27.7% No change 30% 29.4%
27.7% No change
23.0% 1%–5% increase 1%–5% increase
20% 6%–10% increase 20% 6%–10% increase
More than 10% increase More than 10% increase
10% 8.8%
10% 6.3%
4.2% 4.7% 4.5%
3.1%
0% 0%
Micro Small Medium Services Manufacturing Agriculture

The LPM result provided a more detailed picture of the lockdown’s impact on MSMEs
(Table 5). By sector, the regression result (sales1) indicated that education and
construction were more likely to have no sales in March 2020 than agriculture (base) due
to temporary business closures. Specifically, in comparison with agriculture-based firms,
the number of MSMEs with no sales was 47.9 percentage points higher in education and
39.8 percentage points higher in construction, both with significance at the 1% level. As
the ECQ included school closures and limited business operations,
it makes sense that the lockdown affected education most seriously and that construction
(given the government’s build-build-build program) immediately slowed. The power and
energy sector (including electricity and gas) was less likely to report
no sales. The number of no-sales MSMEs in the power and energy sector was
41.0 percentage points lower than that of agriculture-based firms at the 1% significance
level due to the increased demand for electricity and gas from households under ECQ
stay-at-home orders.
For firms that operated continuously during the pandemic, the estimates (sales2)
indicated that other services (including tourism), human health and social work activities,
and manufacturing were more likely to have decreased sales in March 2020 than
agriculture. The number of MSMEs with decreasing sales was 11.5 percentage points
higher in other services (the 5% significance level), 11.2 percentage points higher in
21
human health and social work activities (the 10% significance level), and
9.7 percentage points higher in manufacturing (the 10% significance level) than in
agriculture. These industries continued operating, but their sales declined immediately
after the lockdown.
By region, compared with NCR-based firms (base), MSMEs in Caraga (Region XIII),
SOCCSKARGEN (Region XII), Cagayan Valley (Region II), and Northern Mindanao
(Region X) were less likely to have no sales in March 2020 or no business closures. The
number of no-sales MSMEs was 49.2 percentage points lower in Caraga than in the
NCR, 46.7 percentage points lower in SOCCSKARGEN, 34.7 percentage points lower
in Cagayan Valley, and 17.3 percentage points lower in Northern Mindanao at the 1%
significance level (except Region X at the 5% significance level). This suggests that
MSMEs in the NCR or capital city were more likely to see declines in sales than those in
the provinces. MSMEs in the Autonomous Region in Muslim Mindanao (ARMM) were
less likely to have no sales (68.0 percentage points lower) but more likely to have
decreased sales in March 2020 (4.3 percentage points higher) than NCR-based firms—
however, it should be noted that there was just one firm from this region. Therefore, the
analysis of MSMEs will exclude the ARMM for the rest of this paper.
Compared with young enterprises operating for up to 5 years, longer-established MSMEs
were less likely to have no sales (no closure of business) in March 2020, suggesting the
immediate negative impact on the sales of young firms after the lockdown due to
business closures. For instance, the number of no-sales MSMEs was 3.8 percentage
points lower in firms that had operated for more than 31 years than in those that had
operated for up to 5 years; however, those decreasing sales increased by 2.5 percentage
points in firms aged 31 years and over, although they were statistically not significant.
Women-led MSMEs seemed to face more serious effects of the pandemic and lockdown
in sales than men-led MSMEs. The number of women-led MSMEs with no sales due to
business closures was 9.5 percentage points higher than that of men-led MSMEs at the
5% significance level. For those that operated continuously during
the pandemic, women-led MSMEs with decreasing sales in March 2020 increased by
1.3 percentage points compared with men-led MSMEs, but this was not statistically
significant.
For MSMEs involved in GVCs, we did not see any statistically significant result for
no-sales GVC firms or immediate business closure after the lockdown. They were
less likely to have decreased sales in March 2020 than domestically focused MSMEs or
non-GVC MSMEs. The number of GVC MSMEs with decreased sales was
10.9 percentage points lower than that of non-GVC firms at the 10% significance level.
Internationalized MSMEs could survive the first month after the virus outbreak and
lockdown imposed more easily than domestically focused firms.
By firm size, compared with medium-sized firms, microenterprises were more likely to
have no sales after the outbreak and lockdown. The number of microenterprises with no
sales was 22.5 percentage points higher than that of medium-sized firms, with
significance at the 10% level. Small firms with no sales increased by 3.8 percentage
points compared with medium-sized firms, but this figure was not statistically significant.
We did not see any statistically significant results for firms with decreased sales by size.
5.2 Impact on MSME Revenue

The impact on MSME revenue was similar to the results for MSME sales. There were
61.7% of microenterprises with no revenue in March 2020 due to business closures,
followed by small firms (49.1%) and medium-sized firms (35.8%) (Figure 3). MSMEs that
22
operated during the lockdown had a sharp drop in revenue, with 26.5% of micro, 40.8%
of small, and 41.0% of medium-sized firms having more than a 30% revenue decrease
in March from February. All three industrial sectors also had no revenue after the
lockdown (more than 60% of MSMEs in services and manufacturing, respectively). Some
groups of MSMEs increased their revenue due to the special demand brought by the
lockdown.
Figure 3: Revenue in March 2020

70% 70%
61.7% 60.9% 60.4%

60% 60%
50.6%
50% 49.1% Zero (temporarily closed) 50% Zero (temporarily closed)
More than 30% decrease More than 30% decrease
40.8% 41.0% 21%–30% decrease 21%–30% decrease
40% 40%
35.8% 11%–20% decrease 34.7% 11%–20% decrease
1%–10% decrease 31.8% 1%–10% decrease
30% No change 30% No change
26.5% 26.7%
1%–5% increase 1%–5% increase
20% 6%–10% increase 20% 6%–10% increase
13.1% More than 10% increase More than 10% increase
9.9%
10% 10% 6.7%
3.8% 3.7% 3.8% 3.5% 3.9%
0% 0%

According to the LPM, the impact of COVID-19 and quarantine measures on MSME
revenues followed a similar pattern to MSME sales but differed slightly (Table 5).
By sector, the estimates (revenue1) indicated that education, construction, and
accommodation and food services were more likely to have no revenue in March 2020
than agriculture (base) due to temporary business closures. In comparison with
agriculture-based firms, the number of MSMEs with no revenue was 48.5 percentage
points higher in education (the 1% significance level), 38.9 percentage points higher
in construction (the 1% significance level), and 18.1 percentage points higher in
accommodation and food services (the 5% significance level). The results clearly
identified the impact on operations in the education and construction sectors. As
accommodation and food services include the tourism industry, the result also suggests
that tourism closures contributed to the immediate revenue decline in this sector. Similar
to the analysis of sales, the power and energy sector was less likely to have no revenue.
The number of no-revenue MSMEs in the power and energy sector was 42.6 percentage
points lower than that of agriculture-based firms at the 1% significance level due to the
household demand for electricity and gas during the stay-at-home order.
For firms that operated continuously after the outbreak, the estimates (revenue2)
indicated that real estate services were less likely to have decreased revenue in
March 2020 than agriculture. The number of MSMEs with decreasing revenue was 34.7
percentage points lower in real estate services than in agriculture (the 10% significance
level). The real estate sector continued operations, and the ECQ that required people to
stay at home somewhat ensured its revenues. This suggests that the lockdown forced
people to keep their housing/property lease contract during the pandemic, supporting
relatively stable revenues in the sector.
23
By region, similar to the analysis of sales, MSMEs in Caraga (Region XIII), Cagayan
Valley (Region II), and Northern Mindanao (Region X) were less likely to have no revenue
in March 2020 or no closure of business than NCR-based firms. The number of no-
revenue MSMEs was 47.6 percentage points lower in Caraga than in NCR,
38.7 percentage points lower in Cagayan Valley, and 12.8 percentage points lower
in Northern Mindanao at the 1% significance level (except Region X at the 10%
significance level). In other words, MSMEs in the NCR seem to have had no revenue as
compared with those in the provinces. However, for MSMEs that operated continuously
during the pandemic, the number of MSMEs with decreasing revenue was 7.3
percentage points higher in Eastern Visayas (Region VIII) than in the NCR,
5.5 percentage points higher in MIMAROPA Region, 5.0 percentage points higher in
Cordillera Administrative Region (CAR), and 3.5 percentage points higher in Caraga at
the 10% significance level (except CAR at the 5% significance level). Firms with declining
revenue had grown more widely outside the NCR.
By operating period, we did not see any statistically significant result for no-revenue
MSMEs or immediate business closure after the lockdown, but longer-established
MSMEs were more likely to have decreased revenue in March 2020. The number
of revenue-declined MSMEs that had operated for more than 31 years was
2.8 percentage points higher than the number that had operated for up to 5 years at the
10% significance level.
Women-led MSMEs, similar to the analysis of sales, faced a more serious impact from
the pandemic and lockdown than men-led MSMEs. The number of women-led MSMEs
with no revenue due to business closures was 7.4 percentage points higher than that of
men-led MSMEs at the 10% significance level. For those that operated continuously, we
did not see any statistically significant result for women-led MSMEs with decreasing
revenue in March 2020.
Again, we did not see any statistically significant result for no-revenue GVC firms
or immediate business closure after the lockdown, but GVC firms were less likely to have
decreased revenue in March 2020 than non-GVC MSMEs. The number of
GVC MSMEs with decreasing revenue was 10.5 percentage points lower than that of
non-GVC firms at the 10% significance level. The results supported the relatively higher
survival ratio of internationalized MSMEs than of domestically focused firms during the
survey period.
The firm size results were also similar to the analysis of sales, with microenterprises
being more likely to have no revenue after the outbreak and lockdown than medium-
sized firms. The number of microenterprises with no revenue was 27.9 percentage points
higher than that of medium-sized firms with significance at the 5% level. Small firms with
no revenue increased by 10.4 percentage points compared with medium-sized firms, but
this was not statistically significant. We did not see any statistically significant result for
firms with decreasing revenue by size.
5.3 Employment during the COVID-19 Pandemic

Generally, in most MSMEs, employment experienced no change during the first month
following the lockdown (65.7% of micro, 50.8% of small, and 66.3% of medium-sized
firms) (Figure 4), but MSMEs began laying off employees at a relatively early stage to
survive (68.0% of micro, 59.5% of small, and 78.6% of medium-sized firms), especially
in manufacturing (69.4%) and services (67.3%) (Figure 5). The larger the firm size,
the more working options were available. Among medium-sized firms, 41.1% reduced
their employee working hours, followed by 34.8% of small firms and 26.5% of
microenterprises. Work-from-home was not a preferred option for many MSMEs. Smaller
24
firms had much more difficulty in introducing this arrangement: 24.9% of medium-sized,
19.8% of small, and 11.3% of microenterprises.
Figure 4: Employment by MSMEs in March 2020

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
0.3% 0.3%
Micro 34.1% 65.7% Services 32.5% 67.2%
Small 47.6% 50.8% Manufacturing 49.3% 50.0%

1.6% 0.7%
Medium 33.7% 66.3% Agriculture 58.7% 40.1%

1.2%
Increase of employees Decrease of employees No change Increase of employees Decrease of employees No change

Figure 5: Change in Working Environment

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
26.5% 26.0%
Working hours reduced 34.8% Working hours reduced 32.9%
41.1% 37.7%
11.3% 12.1%
Work from home (teleworking) 19.8% Work from home (teleworking) 10.9%
24.9% 7.5%
4.6% 5.2%
Taking sick leave 8.8% Taking sick leave 3.5%
3.4% 4.1%
68.0% 67.3%
Temporarily laid off (staffing cut) 59.5% Temporarily laid off (staffing cut) 69.4%
78.6% 45.4%
1.8% 1.9%
Others 1.3% Others 0.7%
0.0% 2.0%

The LPM estimates (employment) indicated that most industrial sectors were less
likely than agriculture to decrease their employees in March 2020 or mostly retained their
employees (Table 5). In comparison with agriculture-based firms, the number of MSMEs
that decreased their number of employees was more than 20 percentage points lower in
real estate, professional services, human health and social work, financial services,
information and communication technology, administrative services, construction, other
services, wholesale and retail trade, and accommodation and food services, all at the
1% significance level. This suggests that no significant change in employment occurred
among MSMEs in the first month after the lockdown.
25
However, by region, as compared with NCR-based firms, MSMEs in CAR, Cagayan

Valley (Region II), SOCCSKARGEN (Region XII), Eastern Visayas (Region VIII), Davao
(Region XI), and Central Luzon (Region III) were more likely to decrease their number of
employees in March 2020. The number of MSMEs that decreased their number of
employees was 58.3 percentage points higher in CAR (the 1% significance level), 43.3
percentage points higher in Cagayan Valley (the 1% significance level), 37.5 percentage
points higher in SOCCSKARGEN (the 5% significance level), 31.4 percentage points
higher in Eastern Visayas (the 10% significance level), 31.3 percentage points higher in
Davao (the 5% significance level), and 15.7 percentage points higher in Central Luzon
(the 10% significance level). This suggests that provincial MSMEs were more likely to
decrease their employee numbers than those in the NCR.
The results by operating period, the gender of MSME owners, the involvement in GVCs,
and the firm size did not show any statistically significant results.
5.4 Wage Payments during the COVID-19 Pandemic

A large portion of MSMEs suspended wage payments to employees after the
COVID-19 outbreak and lockdown, and this was more pronounced for microenterprises
(59.8%) and manufacturing (61.0%) (Figure 6). More than 30% of small and medium-
sized enterprises reported a decrease in the total wage payments to employees after the
outbreak. By sector, more than 20% of MSMEs in manufacturing and agriculture reduced
their total wage payments.
Figure 6: Total Wage Payments after the Outbreak

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Micro 59.8% 16.8% 20.1% Services 57.9% 16.5% 22.2%

3.2% 3.4%
Small 36.2% 30.6% 31.3% Manufacturing 61.0% 25.5% 12.4%

1.9%
1.1%
0.0%
Medium 45.8% 34.8% 19.4% Agriculture 42.3% 20.6% 30.7% 6.4%
Temporarily no payment Decrease in payments Temporarily no payment Decrease in payments

No change Increase in payments No change Increase in payments

The LPM results (wage1) showed that education, accommodation and food services,
and manufacturing were more likely to have made no wage payments to employees after
the outbreak compared with agriculture, mainly due to temporary business closures
(Table 5). Compared with agriculture-based firms, the number of MSMEs
with no wage payments was 46.6 percentage points higher in education (the 1%
significance level), 26.2 percentage points higher in accommodation and food services
(the 1% significance level), and 15.8 percentage points higher in manufacturing
(the 10% significance level). Similar to the analysis of sales and revenue, the power and
energy sector was less likely not to pay employee wages. The number of MSMEs with
26
no wage payments in the power and energy sector was 49.5 percentage points lower
than that of agriculture-based firms at the 10% significance level, supported by the
demand for the electricity and gas supply.
For firms operating after the outbreak, the estimates (wage2) indicated that financial
services, education, manufacturing, and accommodation and food services were
more likely to have decreased total wage payments to employees after the outbreak than
agriculture. The number of MSMEs with decreasing wage payments was
34.4 percentage points higher in financial services than in agriculture (the 1%
significance level), 31.3 percentage points higher in education (the 5% significance
level), 22.8 percentage points higher in manufacturing (the 1% significance level), and
21.8 percentage points higher in accommodation and food services (the 1% significance
level). Meanwhile, real estate services were less likely to have decreased wage
payments than agriculture. The number of MSMEs decreasing their wage payments was
34.5 percentage points lower in real estate services than in agriculture (the 10%
significance level).
The COVID-19 pandemic and lockdown measures hit education, accommodation and
food services (including tourism), and manufacturing the hardest in sales and revenue,
and they moved to cut wages at the same time. Financial services also moved to cut
their internal costs. Meanwhile, the stay-at-home orders created a new demand for
electricity and gas and real estate services for households, leaving a more favorable
business and employment climate in these sectors than in other sectors.
By region, the wage payment situation varied. MSMEs in CAR were more likely to
promote no wage payments during the pandemic, while those in Caraga (Region XIII)
were less likely to do so than NCR-based firms. The number of MSMEs with no wage
payments was 25.4 percentage points higher in CAR (the 1% significance level)
and 49.7 percentage points lower in Caraga than in the NCR (the 10% significance level).
MSMEs in Eastern Visayas (Region VIII) were more likely to decrease their wage
payments than NCR-based firms (18.0 percentage points higher at the 5% significance
level).
Regarding firms’ operating period, there were no statistically significant results
for MSMEs stopping wage payments after the lockdown, but those operating for
11–15 years were more likely to decrease their wage payments (13.0 percentage points
higher than those operating for up to 5 years at the 5% significance level). We did not
see any statistically significant results by gender of MSME owners, involvement in a
GVC, and firm size.
5.5 Financial Condition during the COVID-19 Pandemic

Most MSMEs immediately lacked the funds necessary to maintain their business after
the lockdown. The magnitude of the working capital shortage was more serious in
smaller firms. There were 43.8% of micro, 23.2% of small, and 3.9% of medium-sized
enterprises with no cash or savings at the time of the survey (Figure 7). In addition, 61.4%
of medium-sized, 53.1% of small, and 37.8% of microenterprises reported that they
would run out of funds within a month. Only 14.8% of micro, 19.0% of small, and 28.7%
of medium-sized enterprises had enough cash and savings to keep their business going
for 1 month. All the sectors faced a serious lack of funds to operate. There were 43.7%
of service MSMEs with no cash or savings at the time of the survey, followed by
agricultural (42.1%) and manufacturing MSMEs (34.2%). Many manufacturing MSMEs
(47.8%) stated that their funds would run out in a month, followed by those in services
(37.4%) and agriculture (35.1%). Some 22.5% of agricultural MSMEs also reported that
they had enough cash and savings to maintain their business for a month.
27
Figure 7: Financial Condition after the Outbreak

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Micro 14.8% 37.8% 43.8% 3.6% Services 15.3% 37.4% 43.7% 3.6%
Small 19.0% 53.1% 23.2% 4.6% Manufacturing 13.7% 47.8% 34.2% 4.3%
0.3%
Medium 28.7% 61.4% 3.9% Agriculture 22.5% 35.1% 42.1%
5.9%
Enough saving, liquid assets, and other contingency budget to maintain business Enough saving, liquid assets, and other contingency budget to maintain business
Cash/fund to run out in a month Cash/fund to run out in a month
Already no cash and savings Already no cash and savings
Others Others

The LPM results (finance) indicated that human health and social work, transportation
and storage, construction, and accommodation and food services were more likely to
lack working capital after the outbreak than agriculture (Table 5). Compared with
agriculture-based firms, the number of MSMEs that had no cash/savings or would run
out of cash/funds in a month was 25.1 percentage points higher in human health and
social work (the 1% significance level), 18.4 percentage points higher in transportation
and storage (the 5% significance level), 18.3 percentage points higher in construction
(the 5% significance level), and 13.8 percentage points higher in accommodation and
food services (the 10% significance level).
Meanwhile, the information and communication technology sector was less likely to face
a lack of working capital. The number of MSMEs in this sector that had no cash/savings
or would run out of cash/funds in a month was 21.6 percentage points lower than that of
agriculture-based firms at the 10% significance level, supported by the demand for
Internet connections caused by the stay-at-home order and teleworking (work-from-
home) arrangements.
By region, MSMEs in Caraga (Region XIII) and Bicol (Region V) were more likely
to lack working capital during the pandemic than NCR-based firms. The number of
MSMEs with no cash/savings or cash/funds that would run out in a month was 20.5
percentage points and 17.0 percentage points higher in Caraga and Bicol, respectively,
than that of NCR-based firms (both at the 5% significance level).
By age, MSMEs that had operated for over 6 years were less likely to lack working capital
than those that had operated for up to 5 years, though this was not statistically significant
except for MSMEs operating for 11–15 years (15.1 percentage points lower at the 10%
significance level). This suggests that young start-ups faced a more serious lack of funds
during the pandemic. The gender of MSME owners, involvement in a GVC, and firm size
produced no statistically significant results.
28
Figure 8: Funding after the Outbreak


Figure 9: Borrowing Small Amounts Quickly
Can you borrow P50,000? Difficulty of Borrowing
Easier now
3.0%
Same as
last year
14.8%
No Yes Don't know

71.0% 29.0% 34.8%
More difficult
47.4%

The descriptive analysis showed that more than half of MSMEs on average (44.2%
of micro, 54.4% of small, and 56.9% of medium-sized firms) used their own funds
or retained profits to survive after the lockdown was implemented (Figure 8).
However, this was not sufficient to continue operating. Relatively large numbers of
microenterprises (33.1%) relied on borrowing from close relatives. Meanwhile, small and
medium-sized firms could apply for and obtain bank loans (8.4% of small and 14.9% of
medium-sized firms). This suggests that financial accessibility differs by firm size.
Ensuring immediate working capital is crucial for MSMEs to survive, but most MSMEs
(71.0%) faced difficulty in raising even small amounts of funds (P50,000) quickly (Figure
9). The survey found that the majority of MSMEs (77.3%) needed up to P10 million
($200,000) to survive during the pandemic.
29
6. POLICY IMPLICATIONS
The survey also asked about MSMEs’ main concerns and obstacles should the
pandemic continue for over a month from the time of the survey. The top-ranked concern
was a lack of working capital, especially for microenterprises (78.7%) and
all sectors (77.5% of services, 81.6% of manufacturing, and 81.7% of agriculture) (Figure
10). This was followed by supply chain disruptions (especially for small [60.9%] and
medium-sized firms [61.4%] and manufacturing [63.1%] and agriculture [61.3%]), loan
repayments, and the decline in the domestic demand (especially for medium-sized firms
[73.9%]). MSMEs’ concerns strongly focused on financing matters.
Figure 10: Concerns and Obstacles Faced by MSMEs

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
40.4% 39.9%
Decline in domestic demand Decline in domestic demand
48.8% 47.2%
continuing over time continuing over time
73.9% 52.4%
5.4% 5.3%
Decline in foreign demand continuous Decline in foreign demand continuous
11.3% 8.7%
over time over time
28.5% 5.8%
43.9% 41.9%
Disruption of production/supply Disruption of production/supply
60.9% 63.1%
chain/business networks chain/business networks
61.4% 61.3%
34.6% 34.6%
Requirements of tax payments 45.8% Requirements of tax payments 40.7%
67.3% 32.8%
78.7% 77.5%
A lack of working capital to maintain or A lack of working capital to maintain or
69.5% 81.6%
restart business restart business
67.0% 81.7%
48.2% 49.0%
Repayment of loans 57.3% Repayment of loans 48.4%
54.2% 34.3%
Micro Small Medium Service Manufacturing Agriculture

Figure 11: Actions Considered by MSMEs

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
19.4% 18.9%
Cancel contracts with suppliers 19.1% Cancel contracts with suppliers 22.0%
15.6% 18.5%
55.6% 55.5%
Request the government to delay tax Request the government to delay tax
69.2% 62.9%
payments payments
88.2% 55.1%
52.6% 53.0%
Ask financial institutions for the Ask financial institutions for the
59.8% 53.8%
delayed repayment delayed repayment
63.0% 46.3%
30.9% 31.1%
Reduce staff (layoffs) 54.9% Reduce staff (layoffs) 41.4%
65.5% 37.9%
25.2% 26.7%
Reduce employees' wage/salary 36.4% Reduce employees' wage/salary 21.3%
39.6% 24.3%
14.1% 14.8%
Apply for bankruptcy 13.6% Apply for bankruptcy 9.1%
0.9% 14.5%
Micro Small Medium Service Manufacturing Agriculture

If the pandemic lasted for over a month from the time of the survey, more than half
of the MSMEs surveyed wanted a deferment of tax payments and loan repayments
30
(Figure 11). More than half of small (54.9%) and medium-sized firms (65.5%) would
consider further staff layoffs, and a third (36.4% of small and 39.6% of medium-sized
firms) would seek employee wage cuts. This was a common trend across the industry.
Among 21 policy options, we found that most MSMEs strongly desired financial support
from the government. The top-ranked policy measure that firms selected was “zero
interest rate and collateral free loans” and “subsidies and grants for business recovery
of MSMEs” (86.1% indicated a strong wish) (Figure 12).
Figure 12: Policy Measures Desired by MSMEs

0% 20% 40% 60% 80% 100%
Tax relief/deferred tax payments/corporate tax reduction 72.6%
Subsidy for business recovery/cash transfer/grants 86.1%
Assistance to pay salary for employees 71.1%
Simplified procedures/eased requirements for public procurement 63.5%
Suspending payment on government contracts 40.5%
Business development and advisory services 67.0%
One-stop service windows to support MSME exporters/importers 59.1%
Removing restrictions to foreign investments in domestic MSMEs 38.6%
Mentoring/business literacy programs for MSMEs 62.0%
Providing assistance on teleworking arrangement 43.2%
Sector-specific support measures (e.g., tourism) 59.0%
Comprehensive information platform on govt assistance programs 75.1%
Special refinancing facility/low-interest rate loans 81.3%
Zero interest rate/collateral-free loans 86.1%
Special credit guarantees (partial or full coverage) 68.7%
Faster approval of bank loans (simplified loan procedures) 79.3%
Loan repayment moratorium/ease of loan repayment conditions 73.8%
Facilitating access to new financing models (DFS) 58.5%
Support MSMEs in access to trade finance and SCF 71.9%
Business restructuring fund 73.9%
Debtor-in-possession financing for MSMEs in financial distress 54.4%
Strongly want Somewhat want Neutral Somewhat don't want Want least

The LPM estimates found that the COVID-19 pandemic and lockdown measures
had created two streams of business clusters: (i) contracting firms hurt badly by
the pandemic and lockdown and (ii) firm groups that benefitted from the lockdown.
The first group includes manufacturing, transportation and storage, construction,
accommodation and food services (including tourism), education, and human health and
social work. The second group includes power and energy, information and
communication technology, and real estate.
In particular, education, construction, accommodation and food services (represented by
tourism), and manufacturing were the hardest hit sectors in sales and revenue.
The ECQ included school closures, travel bans, and limited business operations. It
immediately affected education and tourism-related sectors and slowed construction
(which the government’s build-build-build program promoted). It largely limited or shut
down manufacturing production. These industries, however, did not cut their workforce
31
immediately but quickly moved to internal cost controls (wage cuts) to survive
during the pandemic. Financial services also cut costs by reducing wage payments.
Transportation and storage, construction, accommodation and food services, and human
health and social work also faced a lack of working capital immediately after the virus
outbreak and lockdown.
The stay-at-home orders under the lockdown generated a new demand for electricity and
gas, Internet connections, and real estate, creating a more favorable business
and employment environment in the power and energy sector, information and
communication technology, and real estate services than in other sectors. MSMEs
operating in power and energy were less likely to have no sales or revenue and less
likely to impose wage cuts. The stable demand for household energy use supported
them. MSMEs in information and communication technology had relatively sufficient
working capital during the first month of the lockdown. MSMEs in real estate largely
benefited from stay-at-home orders under the lockdown as their revenues were
somewhat ensured. The LPM estimates showed no statistically significant results in the
wholesale and retail trade sector, except employment (less likely to reduce the number
of employees).
By region, MSMEs in the NCR were more likely to have no sales than those in the
provinces. However, MSMEs with declining revenue were more evident outside
the NCR. Provincial MSMEs cut employees more than those in the NCR. Trends in
MSMEs’ wage payments and financial condition differed by region.
The lockdown immediately affected young firms (operating for up to 5 years), causing no
sales and business closures. Longer-established MSMEs (operating for over
31 years) did not immediately close their business and maintained some sales in the first
month following the lockdown, but they tended to lose revenue. MSMEs operating for
11–15 years had more working capital but initiated wage cuts to save funds
to survive. Young start-ups faced a more serious lack of funds for survival during
the pandemic.
Women-led MSMEs faced more serious impacts from the pandemic and lockdown
in sales and revenue than men-led MSMEs. Although the LPM results showed
no statistically significant results in their financial condition, the ADB rapid surveys
indicated that “women-led MSMEs had more difficulty in raising enough working capital
through formal financial services than men-led MSMEs, partly caused by the lack of
women owning immovable assets (land and buildings) as loan collateral” (ADB 2020c).
This suggests that a more serious lack of working capital contributed to greater losses in
sales and revenue among women-led MSMEs.
MSMEs that were involved in GVCs continued to have sales and revenue in the
first month following the lockdown compared with domestically focused MSMEs or
non-GVC MSMEs. Internationalized MSMEs could survive the first month after the
outbreak more easily than domestically focused firms. The LPM results also showed no
statistically significant results regarding their financial condition, but the ADB surveys
revealed that GVC MSMEs had “sufficient cash and savings to maintain operations and
had better access to bank credit and funding support from business partners” (ADB
2020c). This suggests that a relatively sufficient amount of working capital helped
internationalized MSMEs to survive during the initial stage of the pandemic.
Overall, microenterprises had more temporary business closures, no sales, and no
revenue compared with larger firms after the outbreak and lockdown imposed. This
suggests different coping abilities to the impact by firm size.
32
Considering the LPM results and the survey findings, it is possible to view the Philippine
government’s initial response to the pandemic as being in the right direction given the
urgent need to contain the virus and continue some economic activity. The government
launched a large-scale, four-pillar socioeconomic strategy in May 2020 (P1.74 trillion or
9.1% of the annual GDP) (ADB 2020d): (i) emergency support
for vulnerable groups and individuals (P595.6 billion or 3.1% of the GDP), including wage
subsidies, soft loans and credit guarantees for MSMEs, and cash assistance;
(ii) expanded medical resources to fight COVID-19 (P58.6 billion or 0.3% of the GDP),
including assistance for purchasing medical equipment and supplies along with special
compensation for healthcare workers; (iii) fiscal and monetary actions to finance
emergency initiatives including the central bank’s purchase of government bonds to fund
COVID-19 response measures; and (iv) an economic recovery program focused on
creating and sustaining jobs (P1.1 trillion or 5.7% of the GDP for (iii) and (iv)). This
package covered some of what MSMEs desired and the concerns that they raised in the
survey (especially financial assistance and wage subsidies). However, based on the
LPM findings, the following two policy approaches require further examination as
recovery begins.
6.1 Identifying Focus Groups for Assistance

The ECQ and other lockdown measures included travel bans and temporary closures of
schools and businesses. To ease the impact on businesses, the government increased
spending to support tourism. To support employment, the government provided one-off
financial assistance to the most affected workers (P5,000), a temporary employment
program for informal workers, P1.2 billion to support displaced workers, and
unemployment insurance. To increase access to working capital financing, the central
bank offered a 30-day grace period for loan repayments and reduced loan interest rates
by 25 basis points. However, most of these policy measures came after the fact.
The LPM estimates showed that the lockdown measures hit MSMEs in education,
construction, accommodation and food services, and manufacturing the hardest. The
government already knew the importance of focus group discussion and addressed
this in its assistance package. However, some groups slipped through the net. The LPM
results suggested that the government should address education, construction,
accommodation and food services, and manufacturing as focus groups for assistance.
Moreover, it could expand these efforts to identify those cities, municipalities, provinces,
and regions where the LPM results found geographical differences in the impact. The
LPM results also showed that young firms (operating for up to 5 years) and women-led
MSMEs faced a more serious pandemic impact than others; they could be included in
the focus group as well.
To identify needy focus groups and adjust policy support directions, it is crucial to monitor
MSMEs’ business conditions periodically. Focus group assistance is critical to optimize
the effective use of the limited budget and support the hardest hit groups
of firms, sectors, and areas in a timely manner. This is particularly true given the
uncertainty over containing the pandemic and building a strong recovery. Assistance
should be well balanced between financial and nonfinancial measures in accordance
with those desired by focus groups, referring to the survey findings.
6.2 Differentiating Policy Measures by Firm Size

The LPM and survey results found that it is necessary to differentiate policy measures
by firm size, given the different abilities to cope with the impact among micro, small, and
33
medium-sized enterprises, respectively. In particular, microenterprises had more serious

difficulties than small and medium-sized firms according to the LPM results. The survey
found that larger firms could more successfully arrange a new working environment
during the pandemic and lockdown than smaller firms, such as work-from-home
arrangements and reduced employee working hours. The government could adjust or
customize policy measures by firms’ size according to their abilities to respond to the
pandemic’s impact. It is also important to strengthen the assistance for informal sectors,
including distributive trade, sole proprietorship, and family-run home businesses; they
are typical microenterprises and affected groups.
7. CONCLUSION
This paper examined the initial 1-month impact on MSMEs in the Philippines after the
ECQ or lockdown began. It described the effect of the initial policy measures and some
policy implications with evidence obtained through the nationwide rapid survey
conducted from the end of March to mid-April 2020. The COVID-19 pandemic and
lockdown measures created two streams of business clusters—contracting firm groups
that were devastated by the lockdown and those that benefitted from the lockdown. The
LPM regression identified the MSME groups that were hurt most and those
that benefited most. It found that education, construction, accommodation and food
services (including tourism), and manufacturing were hurt most; power and energy,
information and communication technology, and real estate coped better. The findings
of this paper addressed the importance of two policy approaches in the early stage
following the lockdown—timely identification of focus groups for assistance and
differentiation of policy measures by firm size.
Toward the year 2021, Asian economies gradually shifted to the recovery stage.
However, the overall business environment has yet to adjust; there is a continued sharp
drop in demand and revenue according to the follow-up survey in the Philippines
covering August and September 2020. Business and employment conditions differ by
firm size, but MSMEs are evolving under a new normal that requires a more contactless
society. Assistance for MSMEs to shift their business to digital transactions is another
policy priority given that their traditional business model requires physical and personal
contact. Six months after the outbreak, MSMEs have started introducing work-from-
home arrangements. However, working capital shortages are continuing to rise, as the
follow-up survey found. There is increased need among MSMEs for further financial
assistance from the government. COVID-19 containment will continue into 2021. Given
the different abilities of MSMEs to adjust by firm size, the government could pay more
attention to a phased approach and differentiate policy measures by firm size and sector.
Now is the time to consider an optimal approach that offers targeted assistance yet
ensures fiscal sustainability in a post-COVID-19 environment.
34
REFERENCES
Abrigo, Michael R. M., Jhanna Uy, Nel Jason Haw, Valerie Gilbert T. Ulep, and Kris
Francisco-Abrigo. 2020. Projected Disease Transmission, Health System
Requirements, and Macro-economic Impacts of the Coronavirus Disease 2019
(COVID-19) in the Philippines. Philippine Institute for Development Studies
Discussion Paper Series No. 2020-15. April. Manila: Philippine Institute for
Development Studies.
Amemiya, Takeshi. 1977. “Some Theorems in the Linear Probability Model.”
International Economic Review 18 (3): 645–50.
———. 1985. Advanced Econometrics. Harvard University Press. Cambridge, MA.
Angrist, Joshua D., and Jörn-Steffen Pischke. 2008. Mostly Harmless Econometrics:
An Empiricist’s Companion. 1st ed. Princeton University Press. New Jersey.
Asian Development Bank (ADB). 2020a. Asian Development Outlook 2020 Supplement
(December 2020). Manila: ADB.
———. 2020b. Asia Small and Medium-Sized Enterprise Monitor 2020 Volume I:
Country and Regional Reviews. Manila: ADB.
———. 2020c. Asia Small and Medium-Sized Enterprise Monitor 2020 Volume II:
COVID-19 Impact on Micro, Small, and Medium-Sized Enterprises in
Developing Asia. Manila: ADB.
———. 2020d. The COVID-19 Impact on Philippine Business: Key Findings from the
Enterprise Survey. Manila: ADB.
Hausman, Jerry. A., Jason Abrevaya, and F. M. Scott-Morton. 1998. “Misclassification
of the Dependent Variable in a Discrete-Response Setting.” Journal of
Econometrics 87 (2): 239–69.
Horrace, William. C., and Ronald L. Oaxaca. 2006. “Results on the Bias and
Inconsistency of Ordinary Least Squares for the Linear Probability Model.”
Economics Letters 90 (3): 321–7.
International Monetary Fund (IMF). 2020a. “Chapter 1: The Great Lockdown.” In World
Economic Outlook, April 2020. Washington, DC.
———. 2020b. “Policy Responses to COVID-19. Policy Tracker.” Accessed on
24 June 2020. https://www.imf.org/en/Topics/imf-and-covid19/Policy-
Responses-to-COVID-19.
Wooldridge, Jeffrey M. 2002. Econometric Analysis of Cross Section and Panel Data.
The MIT Press. Cambridge, MA.
World Bank. 2020. “Map of SME-Support Measures in Response to COVID-19.”
14 April 2020. https://www.worldbank.org/en/data/interactive/2020/04/14/map-
of-sme-support-measures-in-response-to-covid-19.
35

Asian Development Bank Institute: ADBI Working Paper Series

Uploaded by

Copyright:

Available Formats

Asian Development Bank Institute: ADBI Working Paper Series

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Asian Development Bank Institute: ADBI Working Paper Series

Uploaded by

Copyright:

Available Formats

ADBI Working Paper Series

COVID-19 IMPACT ON MICRO, SMALL,

Shigehiro Shinozaki and

Asian Development Bank Institute

Please contact the authors for information about this paper.

Email: sshinozaki@adb.org, nagrajrao@adb.org

Asian Development Bank Institute

© 2021 Asian Development Bank Institute

Keywords: COVID-19, economic crisis, economic impact, MSMEs, SME development,

JEL Classification: D22, G20, L20, L50

2. INITIAL NATIONAL RESPONSES TO THE COVID-19 PANDEMIC IN ASIA ............. 2

3.1 Data Structure ....................................................................................................7

4. PROFILE OF THE SURVEYED MSMEs..................................................................... 13

5. FINDINGS FROM THE DESCRIPTIVE STATISTICS AND ECONOMETRIC

5.1 Impact on MSME Sales ...................................................................................21

6. POLICY IMPLICATIONS .............................................................................................30

6.1 Identifying Focus Groups for Assistance ......................................................... 33

2. INITIAL NATIONAL RESPONSES TO THE COVID-19

Table 1: Initial MSME Support Measures Responding to COVID-19

continued on next page

3.1 Data Structure

Table 2A: Comparison between ADB and PSA Surveys

3.2 Weighting the Survey Data

continued on next page

A simple correction of the existing bias induced by self-selection and non-response

3.3 Analytical Approach

3.4 Linear Probability Model

Yi = α + β Indi + δ Regi + ϕ Opsi + ψ Womi + η GVCi + τ MSMEi + ϵ (1)

Table 3: Model Specification

4. PROFILE OF THE SURVEYED MSMES

Table 4: Industry Classification

5. FINDINGS FROM THE DESCRIPTIVE STATISTICS

Figure 1: Business Conditions during the Pandemic

MSME = micro, small, and medium-sized enterprise.

Table 5: COVID-19 Impact on MSMEs—Linear Probability Model

continued on next page

continued on next page

continued on next page

MSME = micro, small, and medium-sized enterprise.

5.1 Impact on MSME Sales

Figure 2: Sales in March 2020

MSME = micro, small, and medium-sized enterprise.

5.2 Impact on MSME Revenue

Figure 3: Revenue in March 2020

61.7% 60.9% 60.4%

MSME = micro, small, and medium-sized enterprise.

5.3 Employment during the COVID-19 Pandemic

Figure 4: Employment by MSMEs in March 2020

Small 47.6% 50.8% Manufacturing 49.3% 50.0%

Medium 33.7% 66.3% Agriculture 58.7% 40.1%

MSME = micro, small, and medium-sized enterprise.

Figure 5: Change in Working Environment

Micro Small Medium Services Manufacturing Agriculture

MSME = micro, small, and medium-sized enterprise.

However, by region, as compared with NCR-based firms, MSMEs in CAR, Cagayan

5.4 Wage Payments during the COVID-19 Pandemic

Figure 6: Total Wage Payments after the Outbreak

Micro 59.8% 16.8% 20.1% Services 57.9% 16.5% 22.2%

Small 36.2% 30.6% 31.3% Manufacturing 61.0% 25.5% 12.4%