A Growing Crisis Work Workers and Wellbeing Myanmar

A Growing Crisis: Work, Workers, and
Wellbeing in Myanmar
Document of the World Bank

Produced by the Poverty and Equity Global Practice
Equitable Growth, Finance and Institutions Vice Presidency
East Asia and Pacific Region.
Author: Sutirtha Sinha Roy ssinharoy@worldbank.org

Giorgia Demarchi and Elizabeth Rhoads provided qualitative analysis based on the community welfare monitoring survey. The author gratefully
acknowledges support from Mariam Sherman, Hassan Zaman and Rinku Murgai. The author thanks Kim Alan Edwards, Yashodhan Ghorpade, Kemoh
Mansaray and Valens Mwumvaneza for their valuable review comments. Additional thanks to Lwin Lwin Aung, Chan Aye Myint and Roy van der Weide
for supporting data collection activities and to Lena Harris for excellent research assistance. The author is also thankful to Bart Minten and Ian Masias
from IFPRI for providing access to MHWS data. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views
of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included
in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of
The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. All errors and omissions are
attributed to the author. Image credit Sasin Tipchai from Pixabay. Retrieved from https://pixabay.com/photos/harvesting‐myanmar‐burma‐rice‐crust‐
1822493/
Executive Summary
Myanmar’s economy has been affected by numerous internal and external shocks since 2020. Myanmar’s
transition to an open market economy in 2011 propelled its GDP per capita by 52 percent between 2012 and
2019 and caused poverty to fall from 42.4 to 24.8 percent between 2010 and 2017. However, the Covid‐19
pandemic and subsequent military coup have now reversed these gains. Estimates of real GDP per capita grew
by 2 percent in 2020 and fell by 18.4 percent in 2021, the year of military coup. Economic growth recovered to
a tepid 3 percent in the following year, suggesting that output levels are still well below 2019 levels.
This report takes stock of the effect of these shocks on Myanmar’s workers and their well‐being using a reliable
new source of household data. The report compares employment indicators from 2017 and 2022 to spotlight
the extent of adversity faced by workers and households. Employment indicators for 2017 are based on official
Myanmar Living Conditions Survey (MLCS) data. However, no official data exists after 2020. A new set of surveys
—the Myanmar Subnational Phone Survey (MSPS) conducted by the World Bank—are now filling this
knowledge gap. MSPS provides reliable employment indicators at the sub‐national level for 2022 that are
comparable with baseline official data from 2017. MSPS provides reliable employment indicators at the sub‐
national level for 2022 that are comparable with baseline official data from 2017. MSPS’ large sample of 8500
households, covering approximately 95 percent of Myanmar’s population across all states and regions, ensures
that almost every demographic group is proportionally represented in this study. While anchored in the MSPS,
this report complements quantitative data with qualitative insights from the Community Welfare Monitoring
Surveys, in particular the March 2023 round, conducted by the World Bank since 2020.
Labor force participation and employment rate in Myanmar fell by 1.6 and 4.8 percentage points, respectively,
between 2017 and 20221. Although 9 million people were added to Myanmar’s working age between 2017 and
2022 due to normal population growth, only 2.5 million males and 0.8 million females were added to its labor
force. Female employment in Myanmar has particularly suffered in recent years. Compared to men, adult
females are more likely to be out of the labor force, more likely to be unemployed, and more likely to be out of
education or training.
Male workers were partially protected by their education and work experience. Employment rates for men with
up to middle school level of education fell in the past half decade, but those with high school or college
education either increased or held their employment rate at par with 2017 levels. Female workers with college
or higher levels of education, in contrast, experienced a fall in the employment rate during this time. Incidence
of “Not in Employment, Education or Training” (NEET) increased by 9.4 percentage points among college
educated women during this time.
The quality of jobs and labor earnings deteriorated sharply. Employment in mining, construction and the retail
sector have risen while agricultural employment has contracted between 2017 and 2022. Moreover, private
sector salaried employment opportunities have diminished considerably while casual and self‐employed roles
1
Through this document, we refer to employment rate as the share of employed individuals in the population (with ages
15 and above).
2
have flourished. Younger adults are more likely to be employed in retail services while older, more experienced
workers have switched to agricultural activities in 2022 – pointing to a misallocation of human capital towards
less productive sectors. Real wages of salaried workers have fallen by 15 percent during the same time.
Aggregate nationwide labor market trends, however, mask significant differences at the subnational level.
Employment rates have fallen by more than 15 percentage points between 2017 and 2022 in Kayah, Kayin and
Tanintharyi – along the eastern border and Sagaing in the north. In comparison, states and regions on the
western border and at the middle of the country experienced up to 5 and 10 percentage point decline in
employment rate respectively. A greater share of households in eastern states and regions incurred income
losses of more than 20 percent over the past 12 months, compared to other states and regions2.
Agriculture was a buffer against falling employment opportunities in western parts of the country. Between
2017 and 2022, rural employment growth outpaced urban levels in western areas, while in eastern states and
regions rural employment lagged growth in urban employment. Differential trends in rural and urban
employment are likely due to differences in patterns of agricultural growth across locations. Agricultural
activities expanded in the western parts of the country and shrank along the eastern border during 2022. States
and regions in the middle of the country, comprising of Mandalay, Nay Pyi Daw, Magway and Ayeyarwady,
experienced notable services sector growth but relatively smaller changes in other sectors. Private sector
salaried employment has reduced in all states and regions.
Reductions in wage earnings were near universal but states and regions that had higher average wages in 2017
experienced larger cuts in the following half‐decade. The shrinking of private sector employment opportunities
has meant that real wages in 2022 were lower than 2017 across all states and regions states. However, states
and regions with higher average wages in 2017 experienced larger reductions in real earnings until 2022 –
resulting in wage convergence across states. The universal drop in real wages and deep household income
losses are associated with over 70 percent of households in every state and region having to either sell assets,
deplete savings, or borrow.
Migration is rarely adopted as a coping strategy; when migration does occur, it is mostly within the country and
towards urban and more well‐off areas. Thirty percent of households that resorted to migration as a coping
strategy had a member that relocated to another rural area within Myanmar, while 56 percent of such
households had a member migrate to urban regions. International migration was limited to a select number of
households. Overall, 4.1 percent of the population changed their township of residence in the past 2 years (that
is, since February 2021) and 2.5 percent of the population changed townships across state and regional
boundaries (either voluntarily or due to conflict induced displacement). Compared to other households that
have been residing in the same township for more than 2 years, immigrant households from other states and
regions have fewer assets, less education, more precarious employment, and have experienced deeper income
losses in the past year.
2
We refer to Kayah, Kayin and Tanintharyi as eastern and Kachin and Shan as north‐eastern; Sagaing as the north‐
western; Chin and Rakhine western; and Mandalay, Nay Pyi Daw, Magway, Ayeyarwady, Bago, Mon and Yangon as states
and regions in the middle or the center of the country.
3
Increasingly, the place of residence within states and regions—not individual skills or endowments‐‐determines
employment outcomes for Myanmar’s workers. In 2017, worker attributes explained 12.6 percent of the
variation in wages, while 14 percent of its variation could be attributed to township level characteristics. But by
2022, the contribution of township level factors towards explaining wage variation doubled while that of
individual characteristics increased only 4.1 percentage points. Thus, wages are increasingly more influenced
by characteristics of where workers reside within states and regions, compared to abilities, skills, education
levels and other individual level factors. This is perhaps not surprising given the nature of location‐specific, yet
covariate, shocks that the country has faced. This report spotlights three township characteristics that have
strongly influenced employment outcomes.
Employment has particularly suffered in townships with higher incidence of conflict events. Our analysis shows
that a one percentage point increase in township share of conflict incidents between 2021 and 2023 is
associated with an average 2.6 percentage point reduction in employment rate. These events contributed to
adverse impacts on the employment status of higher educated male workers, between the ages of 25 to 35,
and living in urban areas.
Employment in the mining sector has increased in townships that had a pre‐existing mine, leading to poorly
diversified economic structures. Concentration of employment in mineral extraction industry could carry
substantial longer‐term costs and environmental risks across Myanmar. The share of mining and construction
jobs in townships with preexisting mines rose by 1.8 percentage points between 2017 and 2022 while
agricultural share in employment contracted by 8.6 percentage points in these areas.
Townships that have higher likelihood of opium production experienced large increases in agricultural activities
and a fall in mining and construction activities. According to the UN Office on Drugs and Crimes (UNODC),
Myanmar experienced a historic growth in poppy cultivation in 2022. Poppy production in the country is
concentrated in Shan, Kachin, Kayah and Chin states. Townships located at altitudes over 1000 meters above
sea level in these four states are offer ideal conditions for poppy cultivation. The share of agricultural
employment in these townships was 22.2 percentage points higher in 2022 than 2017, while mining and
construction employment in these areas fell by 13.3 percentage points during the past half decade. Average
agricultural wages in townships with high risk were also considerably higher during this time.
Faced with a multitude of shocks—electricity outages, escalating energy prices, conflict—Myanmar’s

households have endured deep income losses. Of these three, conflict is inflicting the greatest impact. A one
standard deviation increases in the incidence of conflict in a township increases the probability of households
experiencing deep income losses (of more than 20 percent) over the past year by 4.7 percent – the highest
among a range of individual and township level factors.
4
Introduction
Myanmar’s economy has been subdued by multiple shocks in recent years —COVID, high energy prices, and
conflict — stalling a stretch of rapid growth. Estimates show that Myanmar’s GDP contracted by 18 percent in
2021 and in July 2022, its local currency kyat, depreciated by 30 percent relative to the dollar even as inflation
rose to 19.5 percent. Given these challenges, World Bank (2023) had projected a real GDP growth of 3 percent
between 2022‐23 for the country, suggesting that future output levels will continue to be lower than in 2019.
The impact of Myanmar’s multiple crises has been uneven across states and regions. For instance, households
in Rakhine encountered 10 percent higher price of rice in December 2022 compared to a national inflation rate
of ‐3.0 percent (m‐o‐m; WFP, 2023). Similarly, levels of hunger amongst households in Kachin state are much
higher than the rest of the country (MAPSA, 2022). Even within states, households in townships (the third
administrative division in Myanmar) with more exposure to conflict incidents have experienced greater
adversity than in other locations. Thus, sub‐nationally representative indicators from reliable data sources are
needed to determine the impact of Myanmar’s multiple crises on household welfare.
Past household monitoring efforts have been impeded by the lack of reliable official data. Much of what is
known about household wellbeing in Myanmar since 2020 is based on field reporting by media publications,
qualitative community surveys using purposive sampling, or analysis that either uses simulations of official data
collected before 2020 or data from telephonic surveys. For instance, ILO’s rapid employment assessment for
2022 (ILO, 2022) are based on models calibrated to official labor force survey data from 2020. Others like the
World Bank and IFPRI have collected household data using high frequency telephonic surveys. The World Bank
initiated its data collection immediately at the onset of the pandemic and has shown rising levels of household
adversity across the country. However, owing to their small sample size and survey design, these surveys could
not capture the diversity of impact across states and regions. On the other hand, household surveys conducted
by IFPRI called the Myanmar Household Welfare Surveys (MHWS), have produced reliable sub‐national
household indicators since 2022. However, due to differences in questionnaire design, employment indicators
from MHWS may not be directly comparable with baseline employment estimates from official surveys such as
MLCS‐2017.
A new series of surveys by the World Bank ‐ the Myanmar Subnational Phone Surveys (MSPS), have filled this
gap. MSPS data allows periodic monitoring of regionally disaggregated shocks in Myanmar and can quantify
their effects on household well‐being based on location and other characteristics. The first round of MSPS,
conducted during November 2022 to March 2023, collected household demographics, education, employment,
exposure to adversity, coping strategies, consumption, and prices information. MSPS is comparable with past
sub‐nationally representative surveys in Myanmar, such as, MLCS‐2017. A large sample size exceeding 8500
households, drawn from all fifteen states and regions, and covering approximately 95 percent of Myanmar’s
population, ensures that almost every demographic group in the country is proportionally represented in MSPS.
Employment indicators from MSPS have been used to analyze labor market outcomes in this report and an
accompanying report by Bhatta, et al (2023) uses education and learning indicators from MSPS to take stock of
educational standards in Myanmar. A new round of the community welfare monitoring survey, which has been
5
following a sample of 17 rural communities across eight states and regions since 2020, was implemented in
March‐April 2023 to provide additional insights.
This report is organized as follows: chapter 1 provides a snapshot of overall labor market changes since 2017
and focuses on disparities by gender, employment type, industry of occupation and other worker characteristics.
Chapter 2 analyzes similarities and differences in level employment indicators by state and regions. Chapter 3
concludes by spotlighting key township level characteristics that have influenced workers, their work prospects,
and their overall well‐being.
6
Chapter 1: Snapshot of Myanmar’s labor market in
recent years
Key Messages:
 Labor force participation and employment rate in Myanmar have fallen by 1.6 and 4.8
percentage points, respectively, between 2020 and 2022.
 Female employment has faced considerable setbacks in recent years. Compared to men,
adult females are more likely to be out of labor force, more likely to be unemployed, and
more likely to be out of education or training.
 Higher education and more years of work experience protected employment status of male
workers. However, highly educated women were less likely to be working in 2022 than
2017.
 Employment in mining, construction and the retail sector have risen while agricultural
employment has contracted during 2017 and 2022.
 Quality of employment has suffered over the past half decade as private sector salaried
employment opportunities have diminished considerably while casual work and self‐
employment have flourished.
 Older, more experienced workers have switched to agricultural activities in 2022.
 Real weekly wages of salaried workers have fallen by 15 percent (~3.2 percent annually)
between 2017 and 2022.
Macroeconomic weaknesses are associated with deteriorating labor market conditions in Myanmar.
Labor force participation (LFP) in Myanmar was already on a gradual downward trend despite strong GDP per
capita growth between 2015 and 2020 (Table 1). LFP fell further by 1.6 percentage points between 2020 and
2022 – coinciding with 16.5 percent GDP per capita reduction over the same period3. The prevailing social norm
prescribing that women be primarily responsible for performing household work and tending to children and
elderly dependents (Asian Development Bank, 2016) has contributed to a higher share of women staying out of
the labor force in the past. Female LFP has continued to fall further in 2022 from a low baseline level, as less
than half of all adult women were engaged in the labor force.
In addition, the quality of jobs have suffered. The share of employed adults and share of wage employment
among employed workers have fallen by 4.8 and 6.2 percentage points, respectively, between 2020 and 2022.
3
Employment figures for 2022 are based on World Bank’s new Myanmar Subnational Phone Surveys (MSPS). The first round of MSPS
was conducted between November 2022 to March 2023 and collected detailed information regarding labor market participation,
education levels, consumption, migration, exposure to economic shocks and coping strategies. Additional information about MSPS
survey design and sampling strategy can be found in Sinha Roy (2023).
7
The share of formally employed workers (that is, those with contracts and pension benefits) in the worforce
almost doubled between 2017 and 2022 – likely reflecting the fact that workers without formal work
arrangements were more likely to experience job cuts. The fall in job quality standards suggests rising
informalization of work opportunities. Finally, share of agricultural employment dropped by 6.6 percentage
points between 2017 and 2022.
Table 1: Summary of key‐labor market indicators (as a share of 15+ population)

Share of
Share of wage Share of
formal
LFP rate Employment employment to agricultural
Year LFP rate employment
(Females) rate total employment
among wage
employment (15+)
employees
2015 64.7 51.6 64.2 35.3
2016 36.0
2017 62.0 48.5 61.2 36.8 13.0 49.5
2018 62.4 49.3 61.9 34.4
2019 60.5 47.1 60.3 35.1
2020 60.2 45.6 59.3
2021
2022 58.6 45.9 54.5 28.9 24.2 42.9
WDI, national WDI, national WDI, national WDI, modelled ILO MLCS‐2017,
Sources: MLCS‐2017, MSPS
estimates; MSPS estimates; MSPS estimates; MSPS estimates; MSPS MSPS
The past half decade has seen significant backsliding in employment prospects for women.
Although 9 million people were added to Myanmar’s working age population between 2017 and 2022 through
normal population growth, only 3.3 million people were added to its labor force. During the same period,
number of adults not in employment, education, or training (NEET) grew by 6 million while those in education
fell by 0.3 million. The 6 million additional NEET individuals between 2017 and 2022, comprised of 4 million
females and 2 million male adults. The rising incidence of NEET in Myanmar therefore disproportionately
comprised of women. Of the 3.3 million people that entered the labor force over the past half‐decade, 2.5
million were males and 0.8 million were females. Meanwhile, female unemployment has risen by 7 percentage
points while that of men has grown 2.5 percentage points. Differential population growth cannot explain the
adverse labor market outcomes for women as an almost equal number of male and female individuals of
working age (4 and 4.8 million respectively) were added to Myanmar’s population between 2017 and 2022.
8
Figure 1: Snapshot of Myanmar’s labor market in 2017 and 2022
Notes: Population figures are in million. Estimates for 2017 and 2022 are from MLCS and MSPS respectively. Sample include individuals aged 15 or over.
Deterioration in employment indicators since 2020 has been more pronounced among male workers.
Disaggregating labor force participation by gender and location shows a gap between female and male
participation rates already existed in the past (Figure 2). Between 2015 and 2020, LFP and employment rates
steadily fell among rural women but remained steady for urban females. Thereafter, female employment rate
fell by 3.4 and 3.2 percentage points in rural and urban areas until 2022, even as LFP remained steady. Thus,
employment conditions for rural women were downward trending prior to the COVID pandemic and the military
coup and worsened further in following years.
In contrast, the year 2020 marked a clear break in male employment trends. Participation rates and
employment rates among men were largely stable between 2015 and 2020. Thereafter, male LFP dropped by 5
and 1 percentage points in rural and urban areas, while employment rates fell by 7.7 and 4.2 percentage points,
respectively.
In general, smaller changes in LFP rates of both genders relative to employment rate suggests that the
willingness to find work among working age individuals in Myanmar has been less affected by the crises.
However, Myanmar’s unsupportive economy in 2022 has been unable to provide enough suitable jobs to
interested workers, leading to a relatively larger fall in employment rate.
9
Figure 2: Labor force participation rate and employment rate, by gender and location: 2015‐2022
a. LFP ratio b. Employment rate
85 85
80 80
75 75
70 70
65 65
60 60
55 55
50 50
45 45
40 40
2015 2017 2018 2019 2020 2022 2015 2017 2018 2019 2020 2022
Rural Male Rural Female Rural Male Rural Female

Urban Male Urban Female Urban Male Urban Female
Notes: Estimates for 2015‐2020 and 2022 from ILOSTAT and MSPS respectively. Sample include individuals over 15 years of age.
Education and experience shielded male workers from shocks. In contrast, educated women have withdrawn
and are more likely to be in NEET status in 2022 than in 2017.
Employment of urban males has been most resilient despite numerous shocks to Myanmar’s economy since
2017 across different groups (Figure 3, panel a). In comparison, employment rate for rural males between 25‐
35 years, and 36 to 65 years were 11 and 5 percentage points lower in 2022 than 2017. Thus, younger male
workers with fewer years of experience were more exposed to labor market shocks. On the other hand, older
females experienced a larger fall in employment rate compared to younger ones. Employment rate for women
between the ages of 36 to 65 (in rural and urban combined) fell by 12.3 percentage points compared to an 8.1
percentage point drop among younger female age group (25 to 35 years). Therefore, additional years of work
experience shielded male workers from employment shocks more than female ones.
Male workers also benefitted more from higher educational attainment. Employment rate for men with up to
middle school level of education fell in the past half decade (Figure 3, panel b), but those with higher education
either increased or held their employment rate at par. Female workers with college or higher levels of education
in contrast experienced a fall in employment rate between 2017 and 2022.
Additionally, sixty‐four percent of women with less than primary education and 55 percent with up to primary
education were in NEET category in 2022 (Figure 3, panel c). This represents a 7.6 and 13.3 percentage point
rise in female NEET shares over the past half‐decade. NEET shares have also risen by 9.4 percentage points
among college educated women. In contrast, males with college degrees are less likely to be in the NEET
category in 2022 than in 2017.
10
Qualitative data from World Bank’s community socio‐economic welfare monitoring from March‐April 20234
suggests that competition from returned students and migrants reduced employment opportunities available
to less educated women in rural areas. Male day laborers are paid significantly more than female day laborers
(20‐33 percent more on average across the sample villages), which may have led households to prioritize men’s
remunerated labor. Additionally, some agricultural tasks are seen as unsuitable for women and only offered to
men. Mobility restrictions and security concerns have disproportionately impacted women’s access to work and
employment. Qualitative interviews across the village sample indicate female day laborers are prevented from
travelling in search of work and from commuting to nearby towns or cities for non‐agricultural employment. To
avoid unofficial tolls at checkpoints, laborers in masonry, construction, and some forms of agricultural work
tend to stay at the job site until the work is completed, rather than commuting, making this type of work less
available to women due to cultural mores, security concerns and care responsibilities. Respondents in
community surveys have also noted that since April 2022, a citizenship scrutiny card is required for all internal
travel. According to the 2014 census, women are less likely to possess this documentation, with disparities rising
with age.
Figure 3: Employment rate by gender, age, and education levels

a. By age‐location‐gender b. By education‐gender
c. Share of individuals not in education, employment, or training
4
The community socio‐economic village‐level monitoring is a panel study running annually or bi‐annually since 2020.
Using the village as the unit of analysis, it draws on interviews in 17 villages across 8 of Myanmar’s states and regions and
18 civil society organizations and service providers active in rural areas in the same states and regions.
11
Notes: Source: MLCS‐2017 and MSPS‐2022. Sample include individuals over 15 years of age. Panel (a) shows average WPR for each age using lowess
smoothed locally weighted regressions. Estimates based on sampling weights. Non‐literate population is included under less than primary level of
schooling in panel (b).
The workforce has shifted from agriculture and manufacturing to construction, mining, and retail services.
Sectoral changes among female workers are more pronounced than men.
An expansion of retail sector jobs, combined with rising mining and construction work, provides indications of
rising informality levels in Myanmar since 2017. Male employment has increasingly shifted from agriculture to
mining, construction, and retail sector jobs (Figure 4). Changes in the sectoral composition away from
agriculture for female workers were even more pronounced, falling sharply by 10 percentage points between
2017 and 2022. The rise in services sector employment among both genders, does not signify a positive
structural transformation process that is generally associated with expansion of sectors with high value addition.
This is because most services sector growth over the past half‐decade occurred due to a 7‐percentage point
rise in share of wholesale and retail activities (Figure 5).
Figure 4: Distribution of employment sectors by gender and over time

a. Males b. Females
29% 33% 39%

45%
12%
14% 2%
8% 12% 3%
6%
13%
51% 47% 47%

37%
2017 2022 2017 2022
Agriculture Manufacturing Agriculture Manufacturing

Mining and Construction Services Mining and Construction Services
Notes: Agriculture includes forestry and fishing. Manufacturing sector includes electricity, water supply and waste management sectors. Services
includes all other sectoral codes. Sample include individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐2017 and
MSPS.
Figure 5: Disaggregating services into constituent sectoral groups
12
22 2017 2022
15
5 5
4 4
4 3 3
2
Wholesale and Transportation Hospitality, Professional, Adminisrative Education, Art, Others

retail trade communication, scientific and and public health, social entertainment
finance, real technical admin work
estate
Notes: Figure shows share of workers (male and female) by various services sectoral codes. Sample include individuals over 15 years of age. Estimates
are weighted by sampling weights. Source: MLCS ‐2017 and MSPS.
Work has become more precarious as opportunities for salaried employment that provide stable income have
declined.
Private sector salaried opportunities have diminished while self‐employment work opportunities have increased
for both genders (Figure 6). Similarly, casual work share increased by 8 percentage points for male workers but
only by 1 percentage points for females. Females are less likely to be compensated for their casual work
activities. Whereas unpaid casual work for both genders have risen during 2017 and 2022, for women, the share
of casual work performed in exchange of pay has fallen by 1.7 percentage points (Figure 6, panel c). Casual work
performed by men in exchange for pay, in comparison, almost doubled between 2017 and 2022.
Qualitative data from World Bank’s community socio‐economic monitoring survey suggest that women in the
private sector may have stopped employment due to security risks and challenges such as military checkpoints
between villages and their workplaces. Moreover, high transportation costs due to higher prices of petrol have
affected women’s private sector participation as they rely on transportation services to travel to work on a daily
basis. Combined, these changes make commuting from rural areas to towns or cities for private sector work
less feasible. Childcare responsibilities, increased due to school closures and security concerns, may also be
prompting women more than men to stop working, especially in urban areas. Qualitative data from rural
Myanmar did not capture this driver as strongly, given the underlying childcare strategies for rural households
remained (i.e. elder household members providing care, joining parents in the fields, etc.).
Qualitative monitoring also indicates that an oversupply of male casual labor in villages (exacerbated by return
migration of male villagers due to high inflation, loss of work or security concerns in cities) could have crowded
out female casual activities performed in exchange for pay. Traditional farm activities that were specifically
13
performed by women workers in the past are now being allocated to excess male casual labor as men are
traditionally seen to be able to do multiple tasks related to cultivation.
Figure 6: Employment type by gender

a. Males b. Females
36%
34%
49% 32%33%
44% 29%
37%
22%
22% 23%
15%
9%
5%
4% 5%
Salaried, Salaried, Self Casual work Salaried, Salaried, Self Casual work
private firm public firm employed private firm public firm employed
2017 2022 2017 2022
c. Paid and unpaid casual wage activity

Gender Casual‐ wage activity 2017 2022
Male Paid 4.5% 9.3%
Female Paid 9.4% 7.7%
Male Unpaid 10.7% 14.1%
Female Unpaid 22.3% 25.7%
Notes: Salaried jobs include apprenticeships and trainees. Self‐employed includes working as an own employer – with or without other employees.
Casual work includes operating another business or someone’s farm as well as helping without pay in a business or a farm. Sample include employed
individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐2017 and MSPS.
Shrinking of private sector salaried work and expansion of casual work was most pronounced in the mining
and construction sector.
The expansion of mining and construction activities in the past half‐decade, has coincided with shrinking share
of private salaried workers in the sector (falling 40 percentage points) and an expansion of casual employment
(rising 30 percentage points). Private sector employment also fell, by varying magnitude, in agriculture, services,
and manufacturing (15, 7 and 6 percentage points respectively). Finally, the high concentrations of self‐
employed workers in retail in 2017 has remained largely unchanged over the past half decade. This relative
stability in the retail sector during a tumultuous economic period could explain the 7‐percentage point growth
in retail employment shares since 2017.
14
Figure 7: Employment type by sector
Agriculture Manufacturing
Private, salaried Public, salaried Self employed Casual Private, salaried Public, salaried Self employed Casual
13% 14%
34% 40%
31% 34%
3% 5%
41%
48%
53% 47%
25%
10%
2017 2022 2017 2022
Mining and Construction

Retail
Private, salaried Public, salaried Self employed Casual Private, salaried Public, salaried Self employed Casual
3%
10% 20% 18%
33%
16% 59% 61%
86%
47%
21% 20%
2017 2022 2017 2022
Services
Private, salaried Public, salaried Self employed Casual
7% 11%
36% 32%
21% 28%
36% 29%
2017 2022
Notes: Sample include employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐2017 and MSPS.
High quality human capital is being misallocated to least productive sectors: workers with more education and
years of work experiences are increasingly turning to agricultural activities.
In 2017, 55 percent of college educated individuals worked in non‐retail services sector jobs and 7.5 percent in
agriculture (Figure 8). But by 2022, the share of highly educated individuals working in non‐retail services fell
by 11 percentage points even as share of agriculture in the group rose by an equivalent amount. This result
points to significant misallocation of human capital as the most educated workers in the economy are
increasingly turning to low productivity agricultural activities. The sectoral reallocation was associated with a
fall of 5 percentage points share of private sector salaried jobs among highly educated workers and a rise in
self‐employment of an almost equivalent amount (Figure 9).
In addition to education, employment patterns also vary by years of experience in two notable ways (Figure 10).
First, younger workers are much less likely to be in agriculture in 2017 than in 2022. At the same time, share of
agriculture among older workers has risen – further evidence that the most productive human capital is being
diverted to low productivity sectors. Second, retail employment shares have risen the most for younger workers
15
and fallen for the oldest cohorts. Changes in mining and construction, manufacturing and services are almost
uniform across all age‐groups.
The rise in agricultural employment among older workers is associated with higher growth in casual work and
fall in self‐employment compared to all other age groups. The large increase in retail sector jobs among younger
cohorts on the other hand have led to increases in self‐employment and casual work status of comparable
magnitude. Lastly, private sector salaried work opportunities have fallen almost equally across all age categories.
Qualitative data from World Bank’s community monitoring surveys suggests that students and younger return
migrants could have taken‐up retail employment in their home villages. On the other hand, civil servants who
have given up their public sector positions have turned to subsistence farming as a coping strategy. Other
sectors of employment are not feasible for such individuals because their identity cards identify them as
‘government staff’ and many employers do not want to take risks in employing them. Given a relatively high
share of more educated women were working in the public sector (often as teachers, nurses, clerks, etc.),
women are particularly likely to have been affected.
Qualitative interviews have also highlighted that farmers are planting less acreage and lowered demand for
casual laborers. Due to the high cost of agricultural inputs and transporting items to markets, farmers rely more
on family labor, including that of out of school children and youth. Villages in the qualitative study pointed to
security concerns as another potential explanation: rising insecurity has meant fewer work opportunities either
due to active fighting making farms inaccessible, or checkpoints within or between village tracts where workers
are subjected to searches or fees to pass through. Lastly, some respondents have highlighted that in heavily
mined areas, even farm labor could be a fatal activity.
Figure 8: Industry by education level

a. 2017 b. 2022
Figure 9: Employment type by education level

a. 2017 b. 2022
16
Figure 10: Changes in sector and type of employment (2017‐2022)

a. Sectoral employment shares b. Share of employment type
Notes: Plotted lines are lowess smoothed locally weighted regressions. Salaried jobs include apprenticeships and trainees. Self‐employed includes
working as an own employer – with or without other employees. Casual work includes operating another business or someone’s farm as well as helping
without pay in a business or a farm. Sample include employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source:
MLCS ‐2017 and MSPS.
Workers without written employment contracts or pension contributions were most likely to lose their
salaried private sector job.
The share of salaried workers possessing a written employment contract or receiving pension contributions
from their employers has risen in 2022 across worker characteristics (Figure 11). These changes occurred likely
because salaried workers with formal work arrangements had a higher chance of surviving deep reductions in
private sector salaried work reported in the earlier sections. As a result, their share within salaried employment
is higher in 2022 than 2017, even as opportunities for salaried work have considerably diminished during the
past half decade.
Figure 11: Share of wage workers with written contracts and pension contributions
17
Willingness to undertake additional work remains high but number of hours of work performed has reduced.
Number of hours of work performed by males per week was 10.7 percent lower in 2022 than in 2017 (Figure
12) compared to 2.3 percent reductions among females. Urban, highly educated, and more experienced
workers were likely to work fewer hours in 2022 than their rural, younger and undereducated counterparts.
The fall in working hours among younger workers since 2017 is unlikely to be related to their willingness to work.
Approximately 1/5th of all younger workers reported willingness to perform additional duties – the highest
among all worker categories – but were unable to find suitable work in the past 7 days (Figure 12, panel b).
Figure 12: Hours of work performed per week and underemployment

a. Hours of work performed per week b. Inability of find additional jobs in the
past week, despite willingness to
perform more duties
Notes: Sample include employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐2017 and MSPS.
18
Real wages have dropped at an annualized rate of 3.2 percent between 2017 and 2022, with higher reductions
among more educated and experienced workers.
Average weekly wages have dropped by 15 percent (~3.2 percent annually) between 2017 and 20225. The
reduction in real weekly earnings coincides with a significant fall in private sector work opportunities identified
in earlier sections. Weekly wages have fallen across the board, with some variation across sectors and worker
characteristics such as education, age, and earnings levels (Figure 13). The fall in earnings is most pronounced
for workers in urban areas and with middle school level of education. Both groups have experienced wage losses
of upto 5 percent annually.
Organizing results by sectors, shows broad based earnings reduction of approximately 5 percent per year across
percentiles in mining and construction. This coincides with about 40 percent reduction in share of salaried
workers in the sector. In manufacturing, retail and services, earnings at the top and the bottom of the earnings
distribution have fallen but are flatter at the middle of the distribution. Finally, despite a 15 percentage point
reduction in share of salaried workers within agriculture, earnings in the sector have changed the least and even
risen at the top of the distribution. This is could be a likely explanation for more educated and experienced
workers switching into agriculture (alongside other factors like security, and inflation making renting
accommodation and buying food in the urban areas too costly for migrants, and transportation too expensive
for commuters).
Figure 13: Annualized changes in earnings by percentile rank

a. Worker characteristics b. Sector of employment
Notes: Percentile ranks assigned on real weekly wages observed in the two surveys, after trimming 1st and 99th percentile outliers. Wages are available
only for employed workers. Sample includes employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐
2017 and MSPS.
Paradoxically, despite broad‐based reduction in wages, 40 percent of households that had a wage‐earning
member reported that they did not experience challenges related to their employment activity. An additional
30 and 26 percent of households reported health or safety related challenges pertaining to their wage work.
5The estimate is based on weekly earnings of regular salaried workers since MLCS 2017 carries earnings information only for such
workers. The average earnings data is calculated after trimming the 1st and 99th percentile of the income distribution due to the
presence of large outlier amounts in both surveys.
19
Moreover, despite observed reductions in number of hours of work performed, only a fifth of such households
reported facing challenges in the number of hours of work performed.
In constrast, over 80 percent of households that had members working in a non‐agricultural or agricultural
sector (either as a self‐employed or casual wage worker) reported challenges related to prices of raw
materials. About 60 percent of households in both categories reported challenges related to avaiability of
factor inputs. Sixty percent of households with members working in non‐agricultural related business cited
weakness in their customer’s ability to purchase goods as a detriment to activities. Finally, 70 pecent of
households with members engaged in the agriculture sector reported weather, pests or security‐related
challenges in accessing their farms.
Figure 14: Challenges related to work activities

a. Wage workers b. Self‐employed workers
c. Agricultural workers
Notes: Samples in panel a, b and c are restricted to households that had members working a regular salaried job, members working in non‐agriculture
related sectors and members engaged in agriculture related industries respectively (either as self‐employed or casual wage workers).
20
Chapter 2: Subnational Differences in
Employment, Earnings, and Coping Strategies
Key Messages:
 Employment losses and income contractions were highest among states and regions along
the eastern border. Males and more educated individuals were less likely to lose employment
across all states and regions.
 In contrast, urban employment losses were higher in eastern, and lower in western states and
regions compared to rural areas. This was likely due to an expansion of agricultural activities
along the western border and rise in non‐agricultural work in eastern states and regions.
 Real wages in 2022 were below 2017 levels across all states. However, wage reductions were
more pronounced in states and regions that had higher wages in the baseline period.
 Households resorted to different coping strategies across states and regions. Asset sales,
savings depletion and borrowing were most common across locations while some household
members resorted to migration.
 Households that migrated across states since February 2021 were more likely to be poorer,
undereducated, and employed than other households residing in the same township.
States and regions along the eastern border have registered a large fall in employment since 2017. Share of
households reporting income losses exceeding 20 percent over the past year is highest in this area compared
to other states and regions.
Employment rates have fallen by more than 15 percentage points in Kayah, Kayin and Tanintharyi – along the
eastern border and Sagaing in the north (Panel a, Figure 15). In contrast, states and regions on the western
border and at the center of the country experienced up to 5 and 10 percentage point decline in employment
rate respectively. The share of the population that is not in employment, education or technical education
(NEET) has also risen in eastern states (Panel b, Figure 15). This indicates a sizable reduction in labor force
participation rates in such areas.
21
Figure 15: Changes in worker share and NEET in percentage points (2017‐2022)
a. Worker share b. NEET
Notes: Sample include employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐2017
and MSPS.
A high proportion of households in eastern states and regions also reported income losses exceeding 20 percent
in the past 5 years (Figure 16). For instance, over 40 percent of families in Kayah, Kayin, and Tanintharyi in the
south‐east and Sagaing in the north‐west, reported income losses exceeding 20 percent in the past year as
compared to 30 percent of households reporting a similar figure across the country. In comparison, over 50
percent of households in Bago, Mon, Mandalay, Yangon, and Nay Pyi Daw (central states and regions) reported
either no change or a small increase in household incomes (Figure 16, panel c) and only about a quarter of
households experienced income reduction over 20 percent (Figure 16, panel b).
Figure 16: Changes in household incomes over the past year

a. Losses below 20% b. Losses exceeding 20% c. No change or rise in
incomes
22
and MSPS.
Individuals with higher educational attainment were more protected from shocks and were less likely to report
employment losses in all states and regions, including those on the eastern side where aggregate employment
losses have been the highest (Table 2). Similarly, gender patterns in employment are consistent across states
and regions – female adults have experienced a larger fall in employment rates than males across locations. In
contrast, employment trends differ considerably across states and regions based on households’ rural and
urban status. Urban households in western states and regions (Sagaing, Chin and Rakhine) experienced a larger
fall in employment rate than in rural areas. Urban households along the northern and eastern borders (Kachin,
Shan, Kayah and Kayin) on the other hand, registered smaller reduction in employment rate than rural. These
urban‐rural differences are indicative of differential growth in agricultural activities across states and regions.
Table 2: Changes in employment rate by states and regions (2017‐2022)

Gender Rural/Urban Education
Below middle Above middle
States Male Female Rural Urban
school school
Kayah ‐30% ‐32% ‐33% ‐20% ‐32% ‐25%
Kayin ‐20% ‐24% ‐24% ‐5% ‐27% ‐2%
Tanintharyi ‐13% ‐16% ‐15% ‐19% ‐23% 8%
Nay Pyi Taw ‐12% ‐13% ‐11% ‐16% ‐15% ‐7%
Sagaing ‐10% ‐19% ‐15% ‐18% ‐19% ‐4%
Rakhine ‐8% ‐10% ‐7% ‐17% ‐8% ‐10%
Mandalay ‐7% ‐15% ‐14% ‐6% ‐16% 0%
Magway ‐5% ‐14% ‐11% ‐5% ‐14% 4%
Ayeyarwady ‐5% ‐9% ‐8% ‐7% ‐9% 0%
Shan ‐3% ‐15% ‐14% 3% ‐13% 7%
Kachin ‐2% ‐1% ‐4% 7% ‐9% 17%
Bago ‐1% 0% ‐1% 5% ‐3% 9%
Yangon 1% ‐6% ‐4% ‐3% ‐8% 3%
Mon 1% 3% 4% ‐1% 1% 6%
Chin 2% 8% 9% ‐23% 1% 15%
and MSPS.
Non‐agricultural employment has intensified in eastern parts while agricultural share in employment has risen
in western states and regions. Growth in non‐agricultural work is associated with rising self‐employment in
eastern states while agricultural growth in western states and regions led to an increase in casual activities.
Eastern states and regions – Kachin, Shan, Kayah and Kayin – experienced a considerable fall in agricultural
employment as share of manufacturing, mining and construction sector jobs expanded (Figure 17). Mining and
construction jobs accounted for most of these increases. Growth in non‐agricultural sectors in these areas were
23
also accompanied by a small rise in retail services activities. The fall in agricultural activities in these states have
likely resulted in rural employment lagging urban areas, as reported in Table 2 above.
States on the western border, most notably Sagaing and Chin, experienced large increases in agricultural
employment (which includes forestry and timber related sectors) and a contraction in manufacturing, mining
and construction and services employment. Agricultural growth expansion combined with contraction of non‐
agricultural activities in this cluster of states and regions corresponds to a larger fall in urban employment than
rural sectors (Table 2). Finally, states in the middle of the country, comprising of Mandalay, Nay Pyi Daw,
Magway and Ayeyarwady, experienced notable services sector growth but relatively smaller changes in other
sectors.
Figure 17: Sectoral changes in employment by state and regions

a. Agriculture b. Manufacturing, mining, c. Services
and construction
Notes: Sample include employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐
2017 and MSPS.
The growth in non‐agricultural activities in the eastern cluster of states and region has contributed to increased
self‐employment in these areas (Figure 18). In contrast, rising agricultural work on the western cluster of states
and regions has led to a rise in casual activities. Private sector salaried job opportunities have fallen in all parts
of the country but most intensely in central states and regions which had the highest shares of such jobs in
2017. Workers in these regions have increasingly switched to casual jobs.
24
Figure 18: Changes in employment types
a. Self‐employed b. Casual work c. Private sector salaried
Notes: Sample include employed individuals over 15 years of age. Estimates are weighted by sampling weights. Source: MLCS ‐
2017 and MSPS.
A pattern of strong wage convergence is observed across states but average wages in 2022 were lower than
2017 in every state.
The drop in private sector opportunities in areas led to a convergence of wage earnings across the country
(Figure 19). Areas with high wages in 2017 experienced a larger reduction in wage earnings between 2017 and
2022 (wage information is available only for salaried workers in MLCS‐2017). In contrast, wages have fallen by
a smaller amount in areas where earnings were low to begin with in 2017. In other country contexts, wage
convergence often signifies a reduction in inequality across locations as low wage areas grow at a faster rate
than higher income locations. However, in the case of Myanmar, the observed reduction in inequality coincides
with significant losses in household wellbeing as all states.
Figure 19: Convergence in wage earnings between 2017 and 2022
Notes: Sample include wage workers over 15 years of age. The wage data excludes top and bottom 1 percent of all workers
due to the presence of outliers.
25
Given widespread income and wage losses, over 70 percent of households in every state and region had to
either sell assets, deplete savings, or borrow to cope with losses.
Households in Kayah have been most affected by income and wages losses. Sixty percent of households in the
region reported income losses exceeding 20 percent over the past year (Figure 16) while experiencing the
largest fall in average wages since 2017 across all states and regions (Figure 19). To cope with such high losses,
over 90 percent of households in Kayah resorted to selling assets, depleting savings or borrowing from non‐
family members. High level of adversity in Kayah is expected as UNHCR estimates about a third of its population
comprises of internally displaced persons 6 . Even in Ayeyarwady, where average wage losses were minimal
(Figure 19) and 62 percent of households reported no change or small increase in incomes over the past year
(the highest across all states, see Figure 16 panel (c)), 74 percent of households had to reduce expenditures in
agricultural inputs, health, education, or food. In states like Kayin, Tanintharyi and Rakhine, households were
more likely to reduce expenditures than sell assets, borrow, or deplete savings to cope with earning shocks.
Long term migration (measured as absence of a member for more than 6 months from the household unit) as
a coping strategy was employed by a limited number of households.
However, data from the World Bank’s community monitoring suggests that a particular form of migration,
displacement, is frequent and repeated. Villages in Magway, Sagaing, and Kayin in particular have faced
frequent localized displacement, with villagers keeping bags packed with essentials in case they need to flee at
a moment’s notice. Displacements were reported to last 1‐30 days at a time, with villagers often returning to
the village during the day or to the fields but spending the night in the forest. Such frequent and repeated
displacement puts tremendous strain on households, leading to negative coping strategies such as reducing
food, borrowing, or selling assets. Those that are displaced for political reasons and forced to leave the village
permanently often sell their assets to finance travel and renting accommodation in another area.
Figure 20: Coping strategies across states and regions
6
UNHCR estimates 95,100 IDPs in Kayah with a total population of approximately 300,000 (UNHCR, 2023)
26
Notes: Estimates are weighted by sampling weights. Source: MSPS.
Most migration occurred within Myanmar and a limited number of households reported a member that
migrated to Thailand and Malaysia. Households that migrated within Myanmar since 2021 were less likely to
own assets, experience deeper income losses and were more likely to be working than other households within
the same township.
Thirty percent of households that resorted to migration as a coping strategy had a member that relocated to
another rural area within Myanmar, while 56 percent of such households had a member migrate to urban
regions. International migration as a coping strategy was limited: only 7 and 4.3 percent of households reported
a member having relocated to Thailand and Malaysia respectively. MSPS also allows us to identify households
that have relocated across townships since 2021. According to UNHCR, 1.499 million people were internally
displaced between February 2021 and May 2023, representing 2.6 percent of Myanmar’s population, and
adding to the more than 300,000 people internally displaced prior to February 2021 who remain displaced. In
comparison, MSPS finds that 4.1 percent of population moved across townships in the past 2 years and 2.5
percent of the population changed townships across state and regional boundaries. The share of migrants in
MSPS is higher because it likely includes internally displaced populations registered with UNHCR in addition to
other migrant families.
Sinha Roy (2023) identifies considerable differences in migration patterns across states. Up to 4 percent of
Mandalay, Mon and Yangon’s population is observed to be migrant households that have moved townships
across state/region boundaries. In contrast, states like Kayah and Kayin have limited migrants from other
states/regions. Compared to others that reside in the same township that serves as a destination for migrants
and have been living in the area for more than 2 years, migrant households from other states/regions are less
likely to own assets, experience deeper income losses in the past year, be more likely to be undereducated,
more likely to be employed and work in casual sector work (Table 3).
Data from qualitative monitoring found that villagers from Magway, Sagaing, Ayeyawaddy and Rakhine are
migrating to Mandalay, Yangon, Shan State, mining areas in Kachin, and fishing areas in Tanintharyi for work.
Those migrating internationally are going to Thailand, Malaysia, and India. Yet, there are significant new trends
in migration that are important to monitor – family migration from conflict‐affected areas; migration of out of
school teenagers and youth to safer locations or for education; CDM participants and political activists migrating
abroad for safety; an increase in illicit migration and payment of brokers and smugglers; and an increase in
underage youth migrating abroad.
Table 3: Characteristics of immigrant households and other families living within the same township
Characteristics Shares among households Shares among households Difference
that immigrated to the residing in the township for
township from another more than 2 years
state/region
Assets:
Rice cooker 65% 62% 0.03
Refrigerator 34% 34% ‐0.001
TV 48% 65% ‐0.174***
Wardrobe 42% 63% ‐0.212***
Car, motorcycle, scooter 42% 73% ‐0.304***
27
Computer, laptop 91% 88% 0.033
Agricultural land 14% 45% ‐0.31***
Urban 54% 28% 0.258***
Household income changes in the past year
Losses below 20 percent 8% 14% ‐0.063***
Losses above 20 percent 53% 32% 0.211***
Increase or no change 39% 53% ‐0.149***
Education
Less than primary 20% 17% 0.031**
Upto primary 27% 35% ‐0.073***
Upto middle school 24% 20% 0.044***
Upto high school 17% 19% ‐0.016
Upto college or above 11% 10% 0.015
Employed 61% 54% 0.068***
Notes: Estimates are based on sampling weights. * p < 0.10, ** p < 0.05, *** p < 0.01
28
Chapter 3: Micro-determinants of
employment and household wellbeing
Key Messages:
 Between 2017 and 2022, township characteristics played a greater role in influencing
employment outcomes than individual level conditions.
 Employment levels were considerably lower among townships that experienced high levels
of conflict events
 Employment in mining sector has increased in townships that had a pre‐existing mine,
leading to poorly diversified economic structures. The concentration mineral extraction
industry could carry substantial longer‐term costs and environmental risks across Myanmar.
 Myanmar experienced a historic rise in poppy cultivation in 2022. Townships that have
higher likelihood of poppy cultivation, experienced higher agricultural employment growth,
contraction of mining and construction activities, rise in casual work, a fall in self‐
employment activities, and a significant increase in wage incomes.
 Although Myanmar has experienced a series of internal and external shocks, greater
incidence of conflict events at the township level is the strongest correlate of household
level adversity among a range of individual and township level characteristics.
Employment outcomes in 2022 were more influenced by townships level characteristics than education or skill
profiles of workers.
Employment status and wage earnings are jointly determined by worker characteristics (such as gender,
education, years of experience, industry of work, language, etc.) as well as localized factors in the vicinity of
where people reside and work. For example, denser areas are often associated with agglomeration related
externalities that confer higher wages on workers with similar ages, skills, and experience levels. Decomposing
variation of individual level wages in Myanmar into worker and township level factors can help reveal whether
locational characteristics, at the levels below state/regions, exert a greater influence on employment outcomes
than individual level factors.
Table 4 shows the dispersion in real wages across the country has fallen between 2017 and 2022 7 . This
observation is consistent with convergence of wages reported in Figure 19. In 2017, individual level
7
That is, the variance of (log) wages in Myanmar dropped from 0.419 to 0.315 between the two years.
29
characteristics8 explained 12.6 percent of the variation in wages, while 14 percent of the differences in wages
could be attributed to township level characteristics (Table 4, first row). However, by 2022, the contribution of
individual level characteristics increased by 4.1 percentage points to 16.7 percent, but the share of variation in
wages that can be explained by township level characteristics has doubled to 28 percent. Thus, wages are
increasingly more influenced by locational characteristics of where workers reside in Myanmar, compared to
their abilities, skills, education levels and other individual level factors.
Table 4: Decomposition of variation in log wages by individual and location characteristics

Variance of log wages Proportion of variation Proportion of variation
explained by individual level explained by township level
characteristics characteristics
2017 0.419 12.6% 14.0%
2022 0.316 16.7% 28.0%
Notes: Variance calculations use sampling weights. Outlier wages (below 1st percentile and above 99th percentile) are excluded from
2017 and 2022. Sample includes individuals above 15 years. Source: MLCS for 2017 and MSPS for 2022.
In the following sections, we spotlight three township level factors that bear a strong relationship with
employment related indicators: (1) rising exposure to conflict events since the military coup at the township
level; (2) proximity of a township to known mines and quarries and (3) a historic rise in opium production in
select townships that have higher likelihood of poppy cultivation.
Townships with high incidence of conflict events have fewer employed individuals. These events had large
adverse impacts on the employment status of higher educated male workers, between the ages of 25 to 35,
and those living in urban areas.
According to data from Armed Conflict Location and Event Data Project (ACLED), there were 42 conflict events
per township between 2017 to 2020. Between 2021 and 2023, average incidents per township rose to 68.
Conflict events in our study includes battles, explosions, and violence against citizens from the ACLED dataset.
To study how these events affected employment, we overlaid ACLED data with MSPS and MLCS surveys at the
township level. Specifically, we calculated township level share of conflict events across Myanmar that occurred
between 2021 and 2023 and merged it with individual level employment data from MLCS and MSPS surveys.
The horizontal axis of Figure 21 shows the township level share of incidents while the vertical axis denotes the
average employment rate across townships. The divergence between the blue and red trend lines suggests that
townships with the highest concentration of conflict incidents between 2021 and 2023 have experienced the
biggest fall in employment since 2017.
8We consider the following worker characteristics for this analysis: female, urban, log‐age, log‐age‐squared, number of children
between 0 to 18 in the worker’s household, number of older individuals above 65 years in the household, female headed households,
education level of the head of the households, whether the household is employed, dummy variables for below primary, primary,
middle school, high school and college educated workers, Buddhist households and dummy variables for agriculture, manufacturing,
services and mining or construction.
30
Figure 21: Township level employment and share of conflict events between 2021 and 2023
Notes: The figure denotes binned scatter plots. Each dot represents mean values of horizontal and vertical indicators at 20
equally sized bins. The fitted lines are calibrated to actual data. 2017 trend line is based on MLCS‐2017; 2022 trendline is
based on MSPS round 1.
To test this relationship more formally, in Table 5, we regress township level share of conflict incidents
on an individual’s employment status. The results show that a one percentage point increase in township
share of conflict incidents is associated with an average 2.6 percentage point reduction in employment
rate in 2022 (column 1). Male employment rates are more affected by such incidents, falling 4.2
percentage point for a percent increase in township share of conflict incidents (female employment is
insignificantly affected by these events). Rising conflict also adversely impacted employment outcomes
of urban, more educated workers and those between the ages of 25 to 35 living in the township.
Table 5: Correlation between share of conflict events (2021‐2023) at the township level and employment rate
Employment rate in 2022 All Male Female Rural Urban Below Above Ages: Ages:
middle middle 25 to 35 36 to 65
school school
TS level share of conflict ‐0.026* ‐0.042* ‐0.014 ‐0.018 ‐0.031* ‐0.015 ‐0.033** ‐0.047** ‐0.023
(0.015) (0.022) (0.014) (0.020) (0.016) (0.018) (0.015) (0.023) (0.016)
Constant 0.545*** 0.684*** 0.423*** 0.566*** 0.529*** 0.497*** 0.597*** 0.689*** 0.498***
(0.006) (0.009) (0.007) (0.008) (0.008) (0.008) (0.008) (0.011) (0.007)
Observations 27552 12864 14688 11512 16040 14602 12950 6942 20610
Notes: Clustered Standard errors at township level. * p < 0.10, ** p < 0.05, *** p < 0.01
Employment in mining sector has boomed in townships that had a pre‐existing mine, leading to poorly
diversified economic structures. The concentration of mineral extraction industry could carry substantial longer‐
term costs and environmental risks across Myanmar.
31
World Bank (2022), citing reports of rising extractive sector activities such as mining of jade, gold, and rare earth
minerals, highlighted the substantial environmental risks and longer‐term costs associated with such work. This
section examines the employment effects of rising extractive industries in townships that are most exposed to
such activities. The analysis overlays survey data from MLCS and MSPS on historical maps of mines and quarries
obtained from Myanmar Information Management Unit (MIMU). These maps identify mining locations using
digitized high‐resolution imagery from Google Earth and Bing Aerial from 2014‐15. The dataset contains a field
indicating the certainty with which a location can be classified a mine. Figure 22 illustrates mining locations
identified with high certainty; 35 percent of all townships in Myanmar (117 townships in total) are observed to
contain a known mine or a quarry location – suggesting that mining takes place in a sizable portion of the
country. We regress this township‐level indicator, characterizing whether the location has a preexisting mine,
on sectoral composition of employment within the township from MLCS 2017 and MSPS 2022 in the next step.
Figure 22: Preexisting mines detected with high certainty in MIMU’s geospatial dataset
Notes: Scatter points represent centroids of mines detected with high certainty in MIMU’s geospatial data
The results are reported in table 6. Insignificant coefficients in the first row of the table mean that in 2017,
sectoral employment in townships with a historical mine or quarry was well‐diversified across sectors. That is,
despite the presence of mines and quarries, local employment was not dominated by just the mining and
construction sector in such areas. However, the third row of table 6 shows that in 2022, mining and construction
employment share in such townships rose by 1.8 percentage points while agricultural share in employment
contracted by 8.6 percentage points in these areas. Rising mining employment share was also accompanied by
a 15.8 percentage point increase in retail activities.
The rising concentration of employment in mineral extraction, drawing labor away from agriculture, can expose
workers to shocks in the mineral markets, in addition to carrying substantial environmental risks. This can be
inferred on the basis of other country contexts where a boom in extractive sectors without proper resource
management strategies is associated with rent capture (Asher and Novosad, 2023) and resource curse (Ross,
1999). Given that a sizable share of townships are exposed to such vulnerabilities in the extractives sector, the
risk could carry over to Myanmar’s aggregate economy.
32
Table 6: Sectoral share of employment in townships with known mines and quarries in 2017 and
2022
(1) (2) (3) (4)
Mining and Agriculture Manufacturing Retail
Construction
TS with known mines ‐0.012 0.048 ‐0.003 0.003
(0.013) (0.030) (0.011) (0.010)
Year=2022 0.012* ‐0.054*** ‐0.011 0.158***
(0.006) (0.012) (0.007) (0.009)
TS with known mines # Year=2022 0.018* ‐0.086*** 0.013 0.022
(0.010) (0.024) (0.011) (0.014)
Gender ‐0.111*** ‐0.051*** 0.048*** 0.060***
(0.005) (0.007) (0.005) (0.003)
Education below middle school 0.034*** 0.284*** ‐0.015*** ‐0.040***
(0.004) (0.011) (0.005) (0.005)
Constant 0.221*** 0.285*** 0.039*** 0.040***
(0.010) (0.019) (0.010) (0.008)
District Fixed Effect Yes Yes Yes Yes
Observations 39930 39930 39930 53865
Notes: Clustered Standard errors at township level. * p < 0.10, ** p < 0.05, *** p < 0.01. Gender and education are included as controls.
“TS with known mines” indicates townships that had a mine with high level of certainty in MIMU’s mine geospatial data.
The historic rise in poppy cultivation during 2022 is reflected in households reporting high agricultural
employment amongst townships that offer ideal conditions for opium cultivation. Mining and construction
activity shrank in these areas even as worker wages increased.
The Myanmar Opium Survey of 2022, conducted by the UN Office on Drugs and Crimes (UNODC), has found
that the total area under poppy cultivation in Myanmar has risen by 33 percent between 2021 and 2022 ‐‐
reversing a downward trend in cultivation that began in 2014. Similarly, total production and average yield for
poppy is estimated to have increased by 100 and 41 percent respectively in the past year. UNODC also found a
70 percent reduction in incidences of eradication and seizures and projects average farm gate prices of opium
to have risen by 69 percent during this time. Thus, by all yardsticks, poppy cultivation in Myanmar appears to
have experienced a historic boom in 2022.
According to UNODC, cultivation of poppy is concentrated in Shan, Kachin, Kayah and Chin regions of Myanmar.
Within these states, townships at altitudes above 1000 meters above sea level offer ideal conditions for crop
cultivation (Figure 23). In comparison, the likelihood of poppy cultivation among townships that are below 600
meters is quite low. This variation altitude is likely due to changes in climate conditions that support propagation
of the poppy crop. Studies also suggest that higher altitude locations in Southeast Asian countries offer good
soil quality and acidity levels to the poppy plant through optimal exposure to sunlight and rainwater drainage,
which is critical for the plant’s growth (US‐DOJ, 1992).
33
Figure 23: Townships with altitude levels that offer ideal conditions for plant propagation
Notes: Myanmar Opium Survey, 2022 by UNODC
The historic rise in poppy cultivation during 2022 is confirmed in MSPS, by examining if townships at altitude
levels exceeding 1000 meters, located in the four states that offer ideal conditions for plant propagation, have
reported higher share of agricultural activities than other locations within the same states. The blue trend line
in figure 24, panel (a) shows that in 2017, average agricultural employment share was comparable across
townships that are above and below the 1000‐meter altitude threshold. However, coinciding with the historical
rise in poppy cultivation of 2022, the employment share in agriculture among townships above the 1000‐meter
altitude threshold continuously rises at higher altitudes (as indicated by the upward trending maroon line).
Further, panel (b) shows that a similar relationship between agriculture and altitude is not observed around the
500‐meter threshold, neither in 2017 nor 2022. Therefore, in panel (a) we confirm UNODC’s finding of historic
rise in poppy cultivation during 2022 amongst townships at altitudes exceeding 1000 meters and in panel (b)
we find support for the observation that townships at lower altitude levels may be at less likely to grow such
crops.
Figure 24: Changes in agricultural and non‐agricultural employment share in townships that have high
likelihood of poppy cultivation
a. Share of agricultural employment b. Share of agricultural employment
(at 1000 meter threshold) (at 500 meter threshold)
Notes: Samples restricted to four states that offer ideal conditions for plant propogation: Shan, Kayah, Kayin and Chin. The figure
denotes binned scatter plots. Each dot represents mean values of horizontal and vertical indicators at 20 equally sized bins. The fitted
lines are calibrated to actual data. 2017 trend line is based on MLCS‐2017; 2022 trendline is based on MSPS round 1. Trend lines are
fitted using weighted regressions.
34
We use these insights to study how incidence of crop cultivation at the township level could impact employment
outcomes. Table 7 contains the main results. Column (1) of the table shows agricultural share of employment
in 2022 is 22.2 percentage points higher than 2017 in townships above 1000 m altitude. Correspondingly, the
share of mining and construction employment in the same areas has fallen by 13.3 percentage points during
the past half decade. Similarly, self‐employment is observed to have fallen by 26.1 percent in townships that
have higher likelihood of poppy cultivation while casual work in these areas have risen by 27.4 percentage points
between 2017 and 2022. Differences in manufacturing, services and private sector employment are insignificant
between townships. Finally, average agricultural wage earnings in townships with higher likelihood of poppy
cultivation have risen considerably between the two years.
Qualitative data from World Bank’s community surveys support these findings, while providing additional
insights into agriculture sector dynamics in these areas (see Box 1). Households in Chin State reported that
employment in the poppy fields mostly involves casual employment to carry out activities such as clearing fields,
planting, and harvesting. These workers reported working for 1‐2 weeks per season and receiving three times
the usual daily wage for agricultural workers for a day of work clearing land for poppy fields.
Table 7: Risk of crop cultivation at the township level and its influence on employment outcomes
(1) (2) (3) (4) (5)

Agriculture Mining and Self‐ Share of Log earnings
employment construction employment casual among
share employment share workers agriculture
share workers
Township altitude above 1000 m 0.060 ‐0.011 0.020 ‐0.007 ‐0.232**
(0.062) (0.010) (0.023) (0.029) (0.093)
Year=2022 ‐0.293*** 0.147*** 0.239*** ‐0.198** 0.316
(0.065) (0.051) (0.076) (0.076) (0.530)
TS altitude above 1000 m x Year=2022 0.222** ‐0.133** ‐0.261*** 0.274** 3.544**
(0.099) (0.057) (0.094) (0.109) (1.531)
Constant 0.632*** 0.042*** 0.299*** 0.574*** 10.425***
(0.051) (0.009) (0.020) (0.018) (0.062)
Observations 10153 19940 4577 4577 788
Notes: Clustered Standard errors at township level. * p < 0.10, ** p < 0.05, *** p < 0.01. Samples restricted to four states at high risk:
Shan, Kayah, Kayin and Chin. Source: 2017 data from MLCS‐2017 and 2022 data from MSPS 2022. Estimates based on weighted
regressions. Earnings data is available only for wage workers.
35
Box 1: A Deeper dive into agriculture
Trends in agricultural employment from MSPS can be summarized as follows. The share of workers engaged in
agriculture has fallen by 6.6 percentage points in the past half decade with larger reductions among female
workers than men. Workers with higher levels of education and more years of experience were more likely to
work in agriculture in 2022 than in 2017. This suggests that highly qualified labor in Myanmar is increasingly being
misallocated to a low‐productivity sector. A quarter of all agricultural workers had a regular salaried employment
in 2017. By 2022, the share of regular salaried workers engaged in agriculture fell by 15 percentage points. Thus,
there has been a considerable rise in informal employment in the sector over the past half‐decade. At the same
time, regular salaried workers in agriculture that managed to keep their formal positions, were able to better
protect their wage levels (in real terms) as regular wage earners in the sector experienced the smallest reductions
in wages; some at top of the earnings distribution even reported a small rise in real wages.
States and regions along the north and eastern borders experienced a decline in agricultural employment
between 2017 and 2022 but high‐altitude townships in these areas reported a rise in agricultural activity. A
potential reason for this difference could be rising incidence of poppy cultivation at higher altitudes. Since 2022
was a historical year of poppy cultivation in Myanmar, high‐altitude townships along the northern and eastern
border that were most likely to engage in poppy cultivation, experienced an increase in agricultural activity. The
activity was highly lucrative as average wages in such townships were considerably higher than others.
Qualitative evidence from community welfare surveys show that additional dynamics are also at play. They
suggest that increases in the costs of agricultural inputs (including fertilizers and petroleum) and inflation more
generally have led to farmers in some areas to reduce farmed areas. Others have switched from high‐input crops
like paddy to lower input crops such as beans and pulses; used family labor (including out‐of‐school children) in
agriculture to save on labor costs; or in some cases left farming altogether (reflecting drop in share of agriculture
work in MSPS in the past half‐decade). These coping strategies reduced demand for casual agricultural labor.
Pests in Sagaing and floods in Chin State led to crop loss and therefore less demand for casual labor during the
latest harvest season.
Qualitative interviews also revealed that restrictions on travel and frequent military checkpoints, particularly in
conflict areas, reduced mobility of casual agricultural laborers to areas with higher labor demand. The same
mobility restrictions have impacted the agriculture marketing network. For example, livestock owners in Magway
cited reduced market price of cows due to border market closures, hindering their sales abroad.
Qualitative data also shows that returned students, educated youth, and civil servants participating in the civil
disobedience movement (CDM) have increasingly been involved in agriculture. Their work in the sector either
as subsistence farmers or as casual laborers, echoes findings from the MSPS on the increased share of educated
workers in agriculture. CDM participants in particular have used farming as a coping strategy for their loss of
civil service salary and their limited freedom of movement due to their identity cards designating their
occupation as ‘government staff’.
Displacement in rural areas, particularly in Sagaing, Magway, and Kayin (the qualitative study did not include
Kayah) has impacted not only employment in agriculture, but also farmers’ agriculture marketing strategies.
Farmers in Sagaing reported that they were harvesting and immediately selling crops even if at low prices, as
they could not wait for higher prices 1‐2 months after harvest for fear of future attacks destroying stored crops
or forced displacement causing farmers to abandon their harvest.
36
Myanmar’s economy has experienced two major shocks since the pandemic: high energy prices and conflict
since 2021. While both shocks have adversely impacted household wellbeing in Myanmar, conflict incidents in
the aftermath of the coup have the strongest association with households reporting deep income losses.
Although petrol prices are high across Myanmar, there is considerable variation at the subnational levels. Petrol
prices from MSPS are observed to be higher in states/regions with higher levels of headcount poverty (Figure
25) likely because of their remoteness (locations with less market access have higher average prices and higher
incidence of poverty). For instance, households in Chin (headcount poverty rate of 58 percent – highest among
all states) reported petrol prices to be approximately 3700 kyats per liter – about 50 percent more than the
office price of petrol in Yangon according to the central statistics office9. The concentration of high petrol prices
amongst remote states, which are also the poorest, indicates that the cost‐of‐living crisis has likely had a
disproportionate impact among poorest households in Myanmar.
Figure 25: Average petrol prices are highest in the poorest states and regions
4000 70%
Poverty rate in 2017

mean petrol price headcount poverty 60%
3500 50%
Petrol Price
40%
3000
30%
2500 20%
10%
2000 0%
In the concluding section, we examine if deep income losses among households are more strongly influenced
by higher energy prices at the township level, higher incidence of conflict events in an area or other household
level characteristics10. Table 8 reports these main results. Holding all other factors constant11, urban households
are 1.8 percentage points more likely to report households’ income losses exceeding 20 percent. The probability
of income losses increases by 3.3 to 3.4 percentage points amongst households that have more undereducated
members. Households that have more older members (above 65 years) are able to reduce the probability of
experiencing deep income losses by about 2.1 percentage points. Gender of the household head, their
education level, their employment status, the size of the household and the number of children in the
household are not significantly correlated with households experiencing more than 20 percent income loss in
the past year.
9
https://www.csostat.gov.mm/Statistics/MarketPrice#rice_table accessed on 26 April 2023
10
All individual and township characteristics are standardized with a mean of zero and 1 standard deviation. The
coefficients in table can therefore be interpreted as the impact of 1 standard deviation change in an indicator on the
probability of households reporting more than 20 percent income losses over the past year.
11
The inclusion of district fixed effects means that comparisons are being made across units (households, individuals or
townships) within the same district. For instance, the estimate on the conflict variable shows the effect of such events on
household income losses across townships that are located in the same district but had a differential incidence of such
events.
37
On the other hand, township level characteristics have a stronger influence on household wellbeing than
individual level factors (consistent with the results reported in Table 4 above). Increases in the share of daily
power cuts and high average petrol prices within the township increases the likelihood of households
experiencing deep income losses by 4.4 and 3.4 percentage points respectively. However, the largest impact on
household wellbeing across a range of individual and location characteristics is township’s exposure to conflict
events. A one standard deviation increase in this variable increases the probability of households experiencing
deep income losses by 4.7 percent – the highest among all other factors. Thus, exposure to such events in the
township is the strongest predictor of adverse shocks of household wellbeing despite Myanmar facing a
multitude of external and internal shocks.
Table 8: Share of households reporting income losses of more than 20 percent in the past year
Share of female members 0.001
(0.010)
Urban household 0.018**
(0.009)
Share of members with less than primary education 0.033***
(0.010)
Share of members with up to primary education 0.034***
(0.013)
Share of members with up to middle school education 0.015
(0.011)
Share of members with up to high school education 0.026***
(0.010)
Share of members with college and above education ‐0.034***
(0.009)
Share of members between ages 0 to 18 ‐0.009
(0.008)
Share of members over 65 years ‐0.021**
(0.008)
Female headed household ‐0.007
(0.011)
Household head educated over high school 0.011
(0.012)
Household head employed ‐0.012
(0.011)
Buddhist household ‐0.006
(0.009)
Household size 0.006
(0.008)
Share of power cuts over a 24‐hour period 0.044***
(0.009)
Petrol prices 0.034***
(0.012)
Township with a known mine or quarry ‐0.016
(0.012)
Share of conflicts in the township 0.047***
(0.014)
Constant 0.295***
(0.006)
District fixed effect Yes
Observations 4652
Notes: Clustered Standard errors at township level. * p < 0.10, ** p < 0.05, *** p < 0.01. “Townships with known mines” indicates
townships that had a mine with high level of certainty in MIMU’s mine geospatial data. All variables are standardized with mean 0 and
standard deviation of 1.
38
References
Asher, S., & Novosad, P. (2023). Rent‐seeking and criminal politicians: Evidence from mining booms.
Review of Economics and Statistics, 105(1), 20‐39.
Asian Development Bank. “Gender equality and women’s rights in Myanmar: A situation analysis.”
Mandaluyong City, Philippines: Asian Development Bank. (2016)
Bhatta, S. D., Katwal, Sinha Roy, S., Van der Weide, R., Sharma, U., Kyaw, A. P., and Thwin, M. M.
2023. Education in Myanmar: Where are We Now? Washington, DC: World Bank.
International Labor Organization. “ILO Brief: Employment in Myanmar in the first half of 2022: A
rapid assessment”. (2022) https://reliefweb.int/report/myanmar/ilo‐brief‐employment‐myanmar‐
first‐half‐2022‐rapid‐assessment
Myanmar Agriculture Policy Support Activity (MAPSA). 2022. The state of food security and nutrition
in Myanmar: Findings from the Myanmar Household Welfare Survey 2021‐2022. Myanmar SSP
Research Note 88. Washington, DC: International Food Policy Research Institute (IFPRI).
https://doi.org/10.2499/p15738coll2.136466
Raleigh, C., Linke, A., Hegre, H., & Karlsen, J. (2010). “Introducing ACLED: An armed conflict location
and event dataset: Special data feature”. Journal of Peace Research, 47(5), 651‐660.
https://doi.org/10.1177/0022343310378914
Ross, M. L. (1999). The political economy of the resource curse. World politics, 51(2), 297‐322.
Sinha Roy, Sutirtha. (2023). “Myanmar Subnational Phone Surveys (MSPS) of the World Bank:
Coverage, Reliability and Representativeness”. Washington, DC: World Bank
World Bank. “Macro poverty outlook for Myanmar”. (2023)

https://thedocs.worldbank.org/en/doc/c6aceb75bed03729ef4ff9404dd7f125‐
0500012021/related/mpo‐mmr.pdf
World Food Program. “WFP Myanmar Market Price Update (November 2022)”. (2023)
https://reliefweb.int/report/myanmar/wfp‐myanmar‐market‐price‐update‐november‐2022
UNODC. “Myanmar Opium Survey 2022”. (2023)
US Department of Justice. “Opium Poppy' Cultivation and Heroin Processing in Southeast Asia” (2022).
https://www.ojp.gov/pdffiles1/Digitization/141189NCJRS.pdf
39

A Growing Crisis Work Workers and Wellbeing Myanmar

Uploaded by

Copyright:

Available Formats

A Growing Crisis Work Workers and Wellbeing Myanmar

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

A Growing Crisis Work Workers and Wellbeing Myanmar

Uploaded by

Copyright:

Available Formats

A Growing Crisis: Work, Workers, and

Document of the World Bank

Author: Sutirtha Sinha Roy ssinharoy@worldbank.org

Faced with a multitude of shocks—electricity outages, escalating energy prices, conflict—Myanmar’s

Table 1: Summary of key‐labor market indicators (as a share of 15+ population)

Rural Male Rural Female Rural Male Rural Female

Figure 3: Employment rate by gender, age, and education levels

c. Share of individuals not in education, employment, or training

Figure 4: Distribution of employment sectors by gender and over time

29% 33% 39%

51% 47% 47%

2017 2022 2017 2022

Agriculture Manufacturing Agriculture Manufacturing

Figure 5: Disaggregating services into constituent sectoral groups

Wholesale and Transportation Hospitality, Professional, Adminisrative Education, Art, Others

Figure 6: Employment type by gender

2017 2022 2017 2022

c. Paid and unpaid casual wage activity

Mining and Construction

2017 2022 2017 2022

Figure 8: Industry by education level

Figure 9: Employment type by education level

Figure 10: Changes in sector and type of employment (2017‐2022)

Figure 12: Hours of work performed per week and underemployment

Figure 13: Annualized changes in earnings by percentile rank

Figure 14: Challenges related to work activities

Figure 16: Changes in household incomes over the past year

Table 2: Changes in employment rate by states and regions (2017‐2022)

Figure 17: Sectoral changes in employment by state and regions

Figure 19: Convergence in wage earnings between 2017 and 2022

Figure 20: Coping strategies across states and regions

Table 4: Decomposition of variation in log wages by individual and location characteristics

Notes: Myanmar Opium Survey, 2022 by UNODC

(1) (2) (3) (4) (5)

Poverty rate in 2017

World Bank. “Macro poverty outlook for Myanmar”. (2023)

UNODC. “Myanmar Opium Survey 2022”. (2023)

You might also like