hiv/aids and its impacts of on labor markets - world...
TRANSCRIPT
HIV/AIDS and its
Impacts of
on Labor Markets
World Bank Labor Market Policy Core Course
April 8, 2010
Damien de Walque
The World Bank
Development Research Group (DECRG)
2
The HIV/AIDS epidemic is
“the most formidable development challenge of
our time”
“By overwhelming the continent’s health and
social services, by creating millions of orphans,
and by decimating health workers and teachers,
AIDS is causing social and economic crises
which in turn threaten political stability.
In already unstable societies, this cocktail of
disasters is a sure recipe for more conflict. And
conflict, in turn, provides fertile ground for
further infections."
UN Secretary-General Kofi Annan, January 2000
3
Overview of Presentation
Part I - Introduction: Current status of the Epidemic
Part II - How is AIDS Affecting Costs to Business and
Labor? Findings from Southern Africa
Part III - How is AIDS Affecting Agricultural Labor
Productivity? The Case of Tea in Kenya
Part IV - How AIDS Treatment is Changing the Impact of
HIV/AIDS? Evidence from Kenya and Botswana
Part V - Conclusions: Treatment and Prevention
4
Part I —
Introduction:
Current Status of the AIDS
Epidemic
5
2007 Global estimates for adults
Source: UNAIDS Dec. 2007 Epidemiological Update
Adults
Women
New HIV infections
(adults)
Deaths due to AIDS
(adults)
30.8 million [28.2 – 33.6]
15.4 million [13.9 – 16.6]
2.1 million [1.4 – 3.6]
1.7 million [1.6 – 2.1]
Total: 33.2 (30.6 – 36.1) million
Western & Central Europe
760 000[600 000 – 1.1 million]
Middle East & North Africa
380 000[270 000 – 500 000]Sub-Saharan Africa
22.5 million[20.9 – 24.3 million]
Eastern Europe & Central Asia
1.6 million [1.2 – 2.1 million]
South & South-East Asia
4.0 million[3.3 – 5.1 million]Oceania
75 000[53 000 – 120 000]
North America
1.3 million[480 000 – 1.9 million]
Latin America
1.6 million[1.4 – 1.9 million]
East Asia
800 000[620 000 – 960 000]
Caribbean
230 000[210 000 – 270 000]
Adults and children estimated to be living with HIV, 2007
Estimated adult and child deaths from AIDS, 2007
Western & Central Europe
12 000[<15 000]
Middle East & North Africa
25 000[20 000 – 34 000]Sub-Saharan Africa
1.6 million[1.5 – 2.0 million]
Eastern Europe & Central Asia
55 000 [42 000 – 88 000]
South & South-East Asia
270 000[230 000 – 380 000]Oceania
1200[<500 – 2700]
North America
21 000[18 000 – 31 000]
Latin America
58 000[49 000 – 91 000]
East Asia
32 000[28 000 – 49 000]
Caribbean
11 000[9800 – 18 000]
Total: 2.1 (1.9 – 2.4) million
Over 6900 new HIV infections a day in 2007
• More than 96% are in low and
middle income countries
• About 1150 are in children under 15
years of age
• About 5750 are in adults aged 15
years and older
of whom:
— almost 50% are among women
— about 40% are among young
people (15-24)
9
Global estimates of Adult
Deaths
HIV/AIDS 1.7 million
Tuberculosis 1.6 million
Cardiovascular Diseases 16.7 million
80% of CVD deaths in developing countries
Smoking related 4.8 million
80% in men, 50% in developing countries
Violence and Injuries >5 million
BUT, compared to other diseases, AIDS tends
to kill relatively young and productive people.
10
Figure 1: Age profile of HIV prevalence. Males
0
2
4
6
8
10
12
14
15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59
Age group
HIV
pre
va
len
ce
(in
pe
rce
nt)
Burkina Faso Cameroon Ghana Kenya Tanzania
11
Figure 2: Age profile of HIV prevalence. Females
0
2
4
6
8
10
12
14
15-19 20-24 25-29 30-34 35-39 40-44 45-49
Age group
HIV
pre
vale
nce (i
n p
erc
ent)
Burkina Faso Cameroon Ghana Kenya Tanzania
12
Part II —
The cost of HIV/AIDS to businesses in
southern Africa
(Source: Rosen, Vincent, MacLeod et
al., AIDS 2004, 18: 317-324)
13
Objectives of the Study
1. Develop a methodology for estimating
the cost of HIV/AIDS to organizations.
2. Calculate the costs of AIDS in the
workforce to companies of different
sizes and in different locations and
sectors.
14
Framework for Analysis
Increased expenses Lost productivity
From one employee
with HIV/AIDS
(individual costs)
Benefits payments
Medical care
Recruitment of replacement
Training of replacement
Increased leave and absenteeism
Reduced on-the-job productivity
Supervisor’s time
Vacancy
Inexperienced replacement
From many
employees with
HIV/AIDS
(organizational costs)
Accidents due to sick or inexperienced
employees
Litigation over benefits, dismissals, etc.
Production disruptions
Loss of institutional memory, experience
Breakdown of workforce morale, cohesion
Diversion of senior managers’ time
Deteriorating labour relations
From high HIV
prevalence in society
(market or external
costs)
Higher cost of material inputs
More security needed due to
breakdown in civil society
Higher wages due to shortage of skilled
workers
Reduced demand for products or services
Higher risk premium on investment
Higher cost of capital
Higher cost of transactions with government and
labour
Total costs of HIV/AIDS
Measured
Unmeasured
*
15
Timing of Cases, Costs, and Liability
Progression of HIV/AIDSin the Workforce
Cost to Company
Morbidity begins (some early mortality, some long-term non-progressors)
Employee becomes infected
Employee leaves workforce through death or retirement (some long-term survivors)
Company hires replacement employee
No cost to company at this stage
Morbidity-related costs are incurred(absenteeism, productivity, management time, medical care)
Termination-related costs are incurred (death and disability benefits, loss of morale, experience, & cohesion)
Turnover costs are incurred (vacancy, recruiting, training)
Timeline
Year 0
Year 2-8
Year 6-12
Year 6-12
Liability Acquired by
Company
Discounted sum of all costs from years 0-10
Employee remains asymptomatic and fully productive
No cost to company at this stageYear 0-8
16
Companies in the Study
Site Co.
A
Co.
B
Co.
C
Co.
D
Co.
E
Co.
F
Sector Heavy
industry
Agric. Mining Mining Retail Media
Location South
Africa
KwaZulu
Natal
Botswana KwaZulu
Natal
KwaZulu
Natal
South
Africa
Size of
workforce
>25,000 5,000-
10,000
500-1,000 500-1,000 <500 1,000-
5,000
Est. HIV
prevalence
8.8%
(1999)
22.9%
(1999)
31.6%
(2000)
24.0%
(2001)
7.9%
(2001)
10.2%
(2001)
Discount rate: 7% (real) Median survival time: 9 years
Assumptions:
17
$0
$15,000
$30,000
$45,000
$60,000
$75,000
$90,000
Co A Co B Co C Co D Co E Co F
Pre
sent valu
e p
er
infe
ction (
2001 $
US
)
Non-permanent Unskilled Skilled Supervisor Manager
Cost Per Incident HIV Infection
Present value per infection for males age 35-49
(Multiple of median annual salary)
(3.8 x)
(1.7 x)
(0.8 x)
(3.8 x)
(1.3 x)
(3.9 x)
18
Why Is the Cost Per Infection So Different?
Variable Higher cost to firm Lower cost to firm
Level of death and
disability benefits
Large; defined benefit;
benefit levels stable
(Co A, C, F)
Premiums capped; benefit
levels falling
(Co B, D, E)
Medical care Medical aid coverage for all
employees
(Co A, C, F)
Most use company clinics
and public hospitals
(Co B, D, E)
Status of unskilled
workers
Permanent employees with
full benefits
(Co A, C, D, F)
Many are contractors with
few benefits
(Co B, E)
Salaries (labor
productivity)
Higher, so absences and
turnover cost more
(Co A, C, D, F)
Lower, so absences and
turnover cost less
(Co B, E)
19
Distribution of the Cost
of an Incident Infection
Leave and absenteeism
Productivity loss
Retirement, death, and disability
Medical care
Recruitment and training
23%
15%
45%
8%
8%
23%
15%
45%
8%
8%
56%
36%4%
3%
56%
36%4%
3%
Company A (high cost) Company B (low cost)
20
Magnitude of the Costs
of a New Infection
Company A Company B
Actual Size
Part III —
The Impact of HIV/AIDS
on Labor Productivity in Kenya
(Source: Fox, Rosen, MacLeod et al.,
Tropical Medicine and International
Health 2004, 9: 318-324)
22
Objectives of the Study
1. Determine the distribution of deaths in the
workforce of a large agricultural firm in
western Kenya.
2. Estimate the decline in on-the-job labor
productivity associated with HIV/AIDS
(impaired presenteeism).
3. Estimate the additional paid and unpaid leave
taken by workers with HIV/AIDS.
4. Calculate the earnings loss caused by
HIV/AIDS.
23
Study Site
Tea estates owned by large agribusiness in Kericho District in the highlands of western Kenya.
Company has ~10,000 tea pluckers (~ 79% male), paid by kilogram of tea leaf plucked.
Labor supply adjusted for tea leaf supply daily.
Free on-site medical care for all workers and dependents; workers get paid sick leave.
VCT available but few accepted at time of study.
Treat opportunistic infections, provide palliative care.
No antiretrovirals at this time (started Oct. 2004).
24
Number of Deaths Per Year
• 600 natural cause deaths and medical retirements at company hospital
• 198 workers, 402 dependents• Many HIV+ workers and dependents leave voluntarily and are
not counted here.
220
164
135
92
70
93
55
0
20
40
60
80
100
120
140
160
180
200
220
1997 1998 1999 2000 2001 2002 2003
Nu
mb
er
of
de
ath
s
> expected malaria and
diarrheal disease in rainy year
estimated based on
Jan-May 03
25
Distribution of Deaths by Age
16%
5%
3%
9%8%
6%4%
2%1% 1% 1% 1%
18%
6%
2%
19%
0
20
40
60
80
100
120
<1
1-4
5-9
10-
14
15-
19
20-
24
25-
29
30-
34
35-
39
40-
44
45-
49
50-
54
55-
59
60-
64
65
>
Unk
nown
Nu
mb
er
of
de
ath
s Effect of HIV/AIDS
26
Distribution of Deaths by Cause
All natural cause deaths, 1997-2002
0% 20% 40% 60% 80% 100%
Dependents (inc.
children) (n=402)
Workers (n=197)
HIV
TB
Other
27
Impact of HIV/AIDS
on Labor Productivity
28
Study Population
―Cases‖ (n=54) are tea pluckers who
died at company health facilities of an AIDS-
related cause (n=43)
were medically retired due to HIV/AIDS (n=11).
―Controls‖ are pluckers who were working in
the same fields over the same time period as
the cases (4:1).
29
Effect of HIV/AIDS on
Kgs Plucked/Day (Adjusted)
Differences are
adjusted for years of
service and unit.
Cases’ shortfall
increases consistently
starting 1.5 years
before death.
YBD § Shortfall
(Kg/day)
Shortfall
(%)†
p-value
2.5 -0.5 1.% 0.834
2.0 -2.4 6.% 0.220
1.5 -4.1 10% 0.028
1.0 -5.6 13% 0.004
0.5 -6.9 16% 0.002
0.0 -7.9 19% 0.003
§ Years before death
† Expressed as a percent of the average amount plucked by
controls, 41.7 kgs/day
30
Effect of HIV/AIDS on
Kgs Plucked/Day (Adjusted)
30
32
34
36
38
40
42
44
46
48
3.0 2.5 2.0 1.5 1.0 0.5 0.0
Years before death
Av
era
ge
kg
s p
luc
ke
d/d
ay
Cases Controls
3.9 kgs/day
9.3 kgs/day
31
5.4
8.8
21.4
33.2
17.116.4
0
5
10
15
20
25
30
35
40
45
50
Sick Leave Annual Leave Unpaid Leave
Nu
mb
er
of
lea
ve
da
ys
Controls
Cases 2-3 Years Before Death
Cases 1-2 Years Before Death
Cases 0-1 Year Before Death
Effect of HIV/AIDS on
Paid and Unpaid Leave
In 3rd year before death, cases were absent 32 days more than controls.
In 2nd year before death, cases were absent 35 days more than controls.
In final year before death, cases were absent 31 days more than controls.
All differences significant at .05 level except annual leave in year 2-3 before death.
32
Overall Loss of Productivity and
Earnings Due to HIV/AIDS
In each of the last two years before death, a tea plucker with HIV/AIDS:
Is absent from work 31 days more often (an increase of 87 percent)
Spends 22 more days on light duty (an increase of 66 percent)
Plucks an average of 7.1 kg less tea leaf per day spent plucking (a decrease of 17 percent)
Has reduced annual earnings of 18%.
Produces 35.1% less tea in final year
33
Part IV
How AIDS Treatment is Changing the
Impact of HIV/AIDS on labor markets
and productivity?
Evidence from Kenya and Botswana
34
The impact of providing
antiretroviral treatment (ART) to
patients Source: Thirumurthy, Graff Zivin, and Goldstein:
The Economics of AIDS Treatment: Labor
Supply in Western Kenya,
The intervention
Provide ART to patients, rule is for a CD4<200, but
initial rationing
Treatment for opportunistic infections
Nutrition supplementation, but not much until later
Data – two rounds of household survey –
random sample & patients
35
Treatment: Evaluation Design
We know what happens to counterfactual
group: Medical evidence: patients continued decline in health
and death. Zero labor supply in round 20
10
20
30
Mean h
ours
work
ed in p
ast w
eek
-50 0 50 100 150 200 250 300 350 400Days on ARVs
treated counterfactual
36
CD4 Counts before and after
treatment50
100
150
200
250
300
Med
ian
CD
4 co
unt
-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90Weeks Before/After ARV Initiation
Source: Goldstein, Graff Zivin and Thirumurthy 2007
37
Body Mass Index and Labor Force
Participation Before and After
ARV Therapy
Source: AMPATH Medical Records System – data as of March 2005.
.6.7
.8.9
Fra
ctio
n p
art
icip
atin
g in
la
bo
r fo
rce
-8 0 8 16 24 32 40 48 56Weeks on ARVs
Source (right): Household survey data.
19
20
21
22
Media
n B
MI
-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90Weeks Before/After ARV Initiation
38
Treatment:
Main Findings – Labor supply
Very large (and rapid) increases in patient labor
supply
16.7 percentage points LFP; 6.9 Hours
Among other household members
Sizable decreases in labor supply of young
boys
22.7 percentage points LFP; 8.6 Hours
No significant response for girls, others living in
the household
Patient earnings are close to the cost of
treatment
39
HIV/AIDS, ARV Treatment and
Worker Absenteeism
Evidence from a large firm in Botswana
Source: Habyarimana, Mabakile and
Pop-Eleches, 2007.
40
CD4 count and absenteeism
41
Absenteeism before and after
treatment initiation
42
Main results from Botswana
Enrolled workers miss about 20 days in the
year prior to treatment initiation
Five times the annual absence duration from
illness among other workers
The introduction of ARV treatment is followed
by a large reduction absenteeism 6-12 months
following treatment inception.
Absenteeism 1 to 4 years after treatment is low
and comparable to other workers.
43
Part V
Treatment and Prevention
44
How to sustain ART in the long run? Example:
Thailand
Total Cost of Public ART (NAPHA)
$0
$50
$100
$150
$200
$250
$300
$350
$400
$450
$500
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
22
20
23
20
24
20
25
Millions
Cost of Public ART_1 line_asy Cost of Public ART_1 line_sym
Cost of Public ART_2 line_asy Cost of Public ART_2 line_sym
Source: Revenga et al, 2007
45
Treatment in the long run?
The greater the success of treatment, the more patients
Second line therapy is much more expensive.
Adherence crucial: role of health infrastructure and staff
Still…
Prevention, prevention, prevention!
46
Impact of condom use on the
number of new HIV infections in India
(Over et al. 2004)
0
1,000,000
2,000,000
3,000,000
4,000,000
5,000,000
6,000,000
7,000,000
8,000,000
1995 2000 2005 2010 2015 2020 2025 2030 2035
Year
Nu
mb
er
of
HIV
in
fecti
on
s p
er
ye
ar
40% condom 50% condom 70% condom 90% condom