Much of what African countries have experienced over the last two
decades, as a result of economic decline and the spread of HIV/AIDS, can be
seen as a generalized process where capital (as defined above) has been
decumulating. Since income is the return on wealth and wealth is the
capitalized value of income, the reduction in the stock of (all types of)
capital has led to lower rates of income growth (with declines in numerous
instances), lower savings and (despite large amounts of foreign aid) lower
investment.
Another factor that has endogenously lowered savings and investment has
been the general loss of confidence among (local and foreign) investors in
Africa. The process whereby that has happened can be understood in terms of
the theories of irreversibility and options values. Investment is defined as
an action that involves a certain present outlay in the expectation of a
future return. All investment is characterized by some sunk costs, or
"irreversibility." Those costs may be reflected in the purchase of
specialized equipment, the time taken to conceive of and develop the
investment project, or the opportunity cost of alternative investments
passed over once the decision to pursue a particular course of action is
taken. These costs, by definition, cannot be retrieved should the investment
be liquidated. Investors expect, however, that future returns from the
investment will amortize these sunk costs.
The prospect of irreversible costs adds to the uncertainty. It also
raises the question of the value of alternative options. These "option
values" exist because all investors have alternatives, the most obvious of
which is to do nothing, i.e., decide to wait. But, like all investment,
"waiting" has an opportunity cost as well. When there is growing
uncertainty, as has been the case in African countries where growth has been
so low, the costs of waiting are more likely to be small relative to the
potential costs associated with irreversibility. For many potential
investors in African countries, especially those with connections to
international financial markets, the alternative activities may be
lucrative, low risk and highly attractive.
Taken together, the notions of irreversibility and options values are
crucial considerations when both locals and foreigners contemplate investing
in African countries. Investors are more likely to wait when they have
information indicating that the spread of HIV/AIDS will affect adversely the
cost structure of any investment they are contemplating. Under these
circumstances, waiting provides time to reassess and re-evaluate their
options. Having a low-risk, secure (foreign) alternative investment reduces
the urgency of committing their resources.55 The outcome has been
devastating in terms of stimulating new activity across Africa. As the
perceived costs of dealing with the spread of HIV/AIDS rises, the rate of
investment tends to decline. This has reinforced the decline in the supply
of investible resources, already under pressure through falling productivity
due to the spread of HIV/AIDS.
Rising Opportunity Cost of Time: One microeconomic mechanism that has
reinforced the decline in savings has been the diminished incentive for
those who are HIV-positive to save. Although many people who are
HIV-positive also have lower incomes from which to save, especially as they
become increasingly debilitated, the two effects need to be analyzed
separately. The prospect of a premature death raises the opportunity cost of
time. There is now a wealth of research, dating back to Irving Fisher,
tracing the systematic changes that occur in the ratio of consumption to
saving as patterns of time preference change.56 For risk averse individuals,
the standard result is that time preference is inversely proportional to the
length of the decision horizon. The relevant decision horizon for people
with HIV/AIDS shortens dramatically. As noted in Annex B, it is normally
five to seven years from the diagnosis of HIV to the onset of AIDS. Death
comes within one or two years after that.
A similar response can be derived from the 'life-cycle' model of
savings.57 The underlying theory recognizes that there is a systematic shift
in the relationship between a worker's earning potential and the pattern of
family formation over his/her life cycle. Early in a person's working life,
income streams and resource demands are rising. With the expectation of a
"normal" life span of work, the person has the capacity to finance his/her
consumption and asset accumulation goals based on the expectation of higher
income flows later in his/her productive life. The spread of HIV/AIDS,
however, has sharply reduced what individuals can "normally expect" from
their (truncated) life span of earnings. Once an individual is HIV-positive,
his/her consumption rises while at the same time his/her potential future
earnings stream is cut. Thus, when viewed in terms of the 'life-cycle'
model, we would expect to see a major reduction in individual savings. To
the extent that the individual has accumulated wealth during his/her working
life, there will be dissaving.58
Much the same result can be derived from the notion of 'overlapping
generations.' The typical approach is to assume two generations - one
generation produces, the other consumes. (The effect is the same as a
continuous two-part 'life-cycle'.) In this formulation, accumulation and
growth occurs because the output (or income) produced by the generation of
producers exceeds the output absorbed (or expenditure) of the generation of
consumers. With the spread of HIV/AIDS, this balance between production and
consumption shifts in ways that reduce this inter-generational surplus,
thereby reducing accumulation. This effect is reinforced as the dependency
ratio (the ratio of workers to persons under 15 years of age) shifts. A
major feature of the HIV/AIDS epidemic has been the sharp rise in the number
of orphans.59 The implication is that the extent to which the generations
overlap has undergone a serious, systematic, and adverse change. The net
outcome is to reduce the capacity of economies with a high prevalence of
HIV/AIDS to save and invest, and ultimately to accumulate and grow.
Declining Labor Productivity: Factors that reduce the rate of investment
lower labor productivity by reducing the level and/or rate of growth of
physical capital with which labor is combined to generate output and income.
From the usual conditions attached to production functions (positive
productivity of all factors, diminishing marginal productivity of individual
factors), a decline in capital per worker reduces output per worker (all
other things being equal). Further reductions in productivity occur when
workers are demoralized and distracted.
Apart from those who are in complete denial, individuals with HIV
recognize that they face a premature death and a shortened decision horizon.
Under normal circumstances, individuals facing rising opportunity costs of
time would invest in labor saving capital and technology. This option is
often unachievable for those who are HIV-positive due to higher costs they
incur for (formal or traditional) medical services. Thus, while the spread
of HIV/AIDS induces the need for higher rates of investment to help maintain
worker productivity, it erodes the means by which such investment can be
financed.
If finance could be arranged (for example, through the efficient use of
foreign assistance), the strategic use of new technology would allow
countries to compensate for the loss of labor due to HIV/AIDS. The obvious
drawback to such a strategy is that it takes resources to acquire new
technologies and skilled personnel to operate them. Though a drawback, this
should not be an insurmountable problem, particularly if attention is given
to devising goal-oriented practical training.60
A further element reducing labor productivity is the reduction in real
effective demand associated with the decline in real per capita income. As
already noted, real incomes across Africa were declining well before the
onset of the HIV/AIDS epidemic.61 Adding HIV/AIDS to the equation has
reinforced the decline. This has occurred through a mechanism described by
Adam Smith over two centuries ago. But, while Smith described the benefits
(to output, productivity, and growth) of expanding the "extent of the
market," the circumstances in Africa have been equivalent to running the
mechanism he described in reverse. The decline in real effective demand
raises the unit costs of all domestic activities. For African countries
where infrastructure is poorly developed, one obvious change is an increase
in transport and distribution costs as a proportion of total costs. The
relevance of this point becomes increasingly evident when one distinguishes,
as T.W. Schultz did, between the economic and the physical supplies of
land.62 The latter is the land base of a country or region. The former is
the land area that, for given technology and market conditions, yields a
positive net rent. Declining real effective demand reduces the net rent from
all economic activities and by extension reduces the productivity of all
factors including the labor that contributes to these activities. The
outcome is contraction in the economic supply of land. Furthermore, since
land is an important component of wealth in African countries, the declining
economic supply of land reduces the value of wealth. This leads to a further
contraction in real effective demand, further lowering the productivity of
associated factors of production.
Other factors contribute to the decline in labor productivity. The theory
of efficiency wages is based on the recognition that, because of fixed costs
of employment (hiring, training, settling-in), firms will have an incentive
to pay above-market wages in order to keep their employees. A further aspect
of the theory is that firms will pay workers above-market wages because of
the direct link between wages and worker productivity.63 Annex A reviews
what employers have been doing in Southern Africa through training and
reorganization of work schedules to maintain productivity. Annex B explains
the incentives employers have to providing training.64
In these matters, one can readily imagine that both these processes work
in reverse as the HIV/AIDS epidemic intensifies. Because of increasing
debility and absenteeism, the marginal value product of workers who are
HIV-positive is less than the wage being paid by the employer. Faced with
this situation, firms have an incentive to reduce employment and/or take
steps to reduce labor costs. Doing this, however, reduces the incomes of all
their employees whether they are HIV-positive or not.65 To the extent that
the efficiency theory of wages holds, overall efficiency declines.
The severity of the economic decline in Africa, however, has resulted in
some practices that tend to contradict the efficiency theory of wages. There
is mounting evidence that working hours have tended to increase across
Africa as a means of counteracting the decline in reward per hour. The
problem has been compounded by the inability of most African countries to
generate increases in net employment in their formal sectors. In Zambia, for
example, there has been no increase in formal sector employment for more
then two decades.66
This "informalization" of employment has increased average hours worked
in a number ways. Employers in the informal sector, by definition, are not
subject to government regulation. Individual workers have the option of
remaining unemployed or accepting the conditions offered. A second factor
has been the proliferation of "coping strategies." Such strategies are low
productivity, time-intensive activities that are characterized by extreme
risk aversion. They generate limited amounts of income but because the risks
of disruption created by government interference have been minimized, the
variability of income is low.
The trends are antithetical to efforts to raise growth since they work in
ways that tend to lower productivity. The growth accounting literature, for
example, has shown that a major source of productivity growth in developed
countries has been the reduction in hours worked.67 Several factors
contributed. Shorter working hours reduce worker fatigue, improve
motivation, and require employers to pay closer attention to organization
and management. Since much of the change in hours of work has come from
dropping the half-day on Saturday, there was also an increase in the ratio
of productive work time to travel time.
A negative link can emerge between the process of economic retrogression
and declining labor productivity. As economies regress, the lower level of
output systematically lowers the average product of labor. Unless real wages
fall at a rate faster than the decline in productivity, or the exchange rate
depreciates correspondingly, unit labor costs will rise. Without these
adjustments, the rising prevalence of HIV/AIDS will further undermine
international competitiveness across Africa. Since African countries have
already undergone considerable "marginalization" due to the inability to
compete in international markets68 and to grow,69 a continued loss of
competitiveness will undermine their capacity for sustained growth and
development.
There is an ironic twist in these changes. The typical argument has been
that globalization and premature liberalization lead to the loss of
competitiveness as goods produced by "cheap" labor out-compete local
products. The logic of the argument above is that, in countries where
HIV/AIDS is spreading rapidly, real unit costs of labor will need to decline
(i.e., labor will need to become correspondingly "cheaper") if these
countries are not to suffer adversely from competitive labor in countries
(both developed and developing) where HIV/AIDS has a less severe impact on
productivity.
We conclude this section by noting the special problems facing
agriculture. There is now widespread evidence pointed to exceedingly high
rates of HIV infection in many farming and fishing villages.70 The death of
large numbers of working age adults in rural areas has led to a dramatic
increase in the number of child-headed households and orphans living with
their grandparents.71 For years, agricultural specialists have argued that
in order to boost agriculture across Africa there has to be an expansion of
technology and knowledge-based production. Governments and donor agencies
have devoted large amounts of effort "extending" such technology and
knowledge through a variety of (often costly) programs. The emerging dilemma
is how the farming households headed by children and grandparents can raise
the resources and absorb the knowledge (and hence take the steps) needed to
raise agricultural productivity.
Agricultural scientists and professional extension workers were
encountering major difficulties formulating and conveying information about
appropriate "packages" to African farmers well before problems associated
with the spread of HIV/AIDS emerged. Their task is orders of magnitude more
difficult now, particularly since the loss of adult workers has lowered the
average educational levels in rural areas. Without potential solutions, the
productivity of agriculture will decline and the one sector in Africa that
has the prospect of stimulating growth and reducing poverty, namely
agriculture, will continue to falter.
Progressive Loss of Skills and Erosion of Institutional Capacity: A major
consequence of HIV/AIDS in organizations such as government departments and
agencies is the loss of "institutional memory." As more senior members of
the staff become debilitated and die, there are fewer people with the
experience to help younger staff members place their work in perspective.
There is also a decline in the amount and scope of on-the-job guidance
("mentoring") that is common to all large organizations. The loss of senior
staff affects the consistency and direction of organizations as well. For
example, budget offices and central banks depend for their effectiveness on
a set of well-established procedures being followed. The loss of key staff
from these (and other) organizations has undermined the ability of African
governments to provide the services crucial to operating a modern economy.
Countries where HIV/AIDS has spread rapidly have major difficulties
maintaining the integrity of these organizations and their procedures. In
Zambia, for example, the rate of staff attrition has been so high in these
key organizations that they have been progressively unable to fulfill their
roles in effectively managing the economy.72 Loss of staff is only one of
several issues. Because the Public Service Commission has been slow to
confirm staff in their positions, the number of "acting" appointments has
proliferated. This reinforces the sense of impermanence in the civil service
further reducing morale and performance. An added outcome has been increase
in irresponsible behavior and opportunism that (due to lack of monitoring
and declining morale) remains largely undetected and un-remedied.
In the limits of organization, Nobel laureate Kenneth Arrow explained why
modern economies and societies have large organizations. Arrow noted that
when viewed in purely economic terms "Organizations are a means of achieving
the benefits of collective action in situations in which the price system
fails."73 Their purpose, Arrow asserted, is "to exploit the fact that
(virtually all) decisions require the participation of many individuals for
their effectiveness."74 The pervasive character, persistence, and cost of
uncertainty imply a special need for cooperative decisions and,
consequently, for organizations. Arrow discussed how improvements in
organizational design that enhance the flow of information improve the
effectiveness of collective actions. He also noted the importance of
education (referring specifically to Pareto's notion of the "circulation of
elites", i.e., the highly skilled) as providing the basis for redefining the
agendas of organizations and raising their performance and thereby adding to
social welfare.75
As we have done earlier, one can readily imagine this process running in
reverse. As HIV/AIDS takes an increasing toll on skilled workers
(particularly those who provide the administrative and strategic
capacities), the performance of critical organizations will decline. In
extreme cases, the organizations may even become dysfunctional, especially
in cases where HIV/AIDS compounds the operational difficulties of
organizations that were already under serious economic stress. Obvious
examples are the many state-owned enterprises across Africa whose
effectiveness has diminished as they scale back to curtail their deficits
and pay off their accumulated debt.
Overview: The above points have highlighted some of the considerations
that underpin the approach taken here to the impact of HIV/AIDS on economic
growth. All of the mechanisms described draw on well-established theoretical
constructs. We have taken advantage of the fact that economic regress, like
economic growth, is a systematic process. Most of the familiar relationships
that we habitually see as fostering "progress," can also be reversed.
The problem with regression, however, is that once underway, there has
(so far) been no easy stopping point. Some African countries --- Ghana,
Zambia, Tanzania, Senegal, Cameroon, and Chad come readily to mind --- have
experienced long periods of economic decline and dissipation. Re-creating
the conditions for growth and development has been far more difficult than
simply reversing the direction. Attitudes shift, behavior changes, and the
demands for additional "safety-margins" and "security" intensify before
individuals will re-embrace the economic reforms. When the damage wrought by
the spread of HIV/AIDS is added, the difficulties are compounded.
Testing the Model
Some Preliminary Results
It is beyond the scope of this paper to formally develop and broadly test
a cross-country model of HIV/AIDS-induced retrogression described above. In
Annex C, however, we have estimated a small (four equation) econometric
model using Zambian data to determine if two variables most directly related
to the spread of HIV/AIDS, changes in life expectancy at birth and changes
in the dependency ratio, have had an independent effect on economic growth.
The model covers the period 1968 to 1998. It was estimated using
three-stage least squares (3SLS).76 The four equations explain the growth of
real income, the growth of real investment, the change in the nominal
exchange rate, and the change in the domestic price level. The last two
variables have been included to capture the effects of the underlying
economic disruption to which the problems created by the spread of HIV/AIDS
have added.
Zambia's present economic problems began in the mid-1970's, when in
response to a major output and price shock in the copper sector, the
government attempted to finance the resulting economic imbalances rather
than adjust. The combination of large amounts of foreign assistance (well
over 20 percent of GDP for extended periods), a series of failed adjustment
programs (eleven so far), large budget deficits, rapid monetary growth, and
manipulated exchange rates, have led to a sustained decline in real per
capita income in Zambia. Under Kaunda's second republic (1972-1991), that
decline was close to 50 percent. So far, under Chiluba's third republic
(1992 to the present), the decline has been around 30 percent.
The HIV/AIDS epidemic has been superimposed upon this broad-based and
(essentially) unrelieved pattern of economic regress. With HIV infection
rates for the adult population of well over 20 percent, Zambia's
experiences, unfortunately, offer an opportunity to gauge the value of the
endogenous growth model as a framework for linking the effects of HIV/AIDS
to economic growth.
The growth equation contains the change in real investment, the change in
the real exchange rate, the change in life expectancy at birth, and the
change in foreign aid. The investment equation includes the growth of real
income, the growth of the labor force, and the change in the real exchange
rate. The intention is to determine the direction and strength of the
relationships among growth, investment, life expectancy, and age dependency
within the context of a simultaneous system. A full explanation of these
variables and the results are reported in Annex C.
The change in the life expectancy ratio is included in the growth
equation to measure, to the extent possible, the independent effect of
HIV/AIDS. Since investment is a major determinant of growth and age
dependency has an important effect on savings behavior, we expect the link
from age dependency to growth to be indirect. Sharp changes in life
expectancy and age dependency have been evident in Zambia from the early
1990s, the period when HIV/AIDS became a full-blown epidemic. The labor
force growth rate (which began to decline during the 1990s) is included in
the investment equation to capture the effect on investment demand of the
growth of the labor force. A key element of all theories of economic growth
(dating from the Harrod-Domar model) is that employers attempt to maintain
the stock of capital per worker. Such behavior would keep investment
(adjusted for depreciation) growing at least at the rate of the labor force.
Preliminary results from this small empirical model show that change in
the life expectancy at birth and the age dependency ratio have statistically
important effects on the rate of economic growth with the expected
respective signs. The coefficient of the change in life expectancy in the
growth equation is 0.68 and is highly statistically significant. The
coefficient on age dependency in the investment equation is -1.159 and is
also highly significant.
Although they are explained in more detail in Annex C, these results show
that the dramatic drop in life expectancy recorded in Zambia during the
1990s has reduced the rate of economic growth. Furthermore, the increase in
age dependency has reduced investment. Because of the direct links from
investment to growth, this too has reduced the rate of economic growth.
b. Developing a Full Scale Model
While the results we have obtained are suggestive, there are important
issues not addressed here. What economic responses can be expected as the
HIV/AIDS epidemic intensifies? What is the feedback from declining growth
rates to the spread of HIV/AIDS? Is there evidence that the spread of
HIV/AIDS is creating barriers to economic recovery? Testing these ideas will
require the specification and estimation of a broader, more comprehensive,
comparative model. The remainder of this section discusses issues involved
in doing that.
From the outset, the endogenous growth framework we have proposed already
provides useful insights regarding the dynamics of HIV/AIDS. There is a
problem with using the model to make projections. Earlier, we noted that the
conventional estimates of growth scenarios based on estimates derived 'with
AIDS' and 'without AIDS' misrepresent the underlying dynamic relationship
between HIV/AIDS and economic growth. African countries have been
experiencing the progressive (and cumulative) effects of the epidemic for
most of the last two decades. Thus, there is no model of an African economy
that can be constructed without the impact of HIV/AIDS having already
influenced its basic structure. In practical terms, this implies that any
projections made using a model will need to be updated on a regular basis as
more information emerges on the changing structure of the underlying economy
due to the spread of HIV/AIDS.
Another factor hindering the construction of any model is the general
lack of data on the course of the epidemic. As noted in Annex A, the data
that are available are not highly reliable or complete due to the widespread
pattern of denial and official obfuscation. Faced with these circumstances,
modelers invariably turn to proxies as we have done above. Proxies are
variables that are highly correlated with the variables that cannot be fully
observed.
Obvious proxies are those used above (life expectancy and age
dependency). Many more could be used. Some of these include the growth rate
of population, trends in the production of staple foods (which in Africa are
typically highly labor-intensive), the gap between actual and expected death
rates, changes in the flow of foreign aid devoted specifically to health,
the number of extraordinary deaths among health workers and teachers, and
the increase in "acting" staff in the civil service. These variables would
be included in a broader econometric model for a particular country, or set
of countries, and examined for both the sign and significance of their
coefficients. As demonstrated by the example provided in Annex C, there are
now a host of small-scale models that include some of the key macroeconomic
variables containing the main elements of a model of retrogression.77
It is always difficult and somewhat presumptuous to predict the sign and
significance of variables in simultaneous equation models. That difficulty
is accentuated in an African context where many countries for many years
have had large internal and external imbalances. Because many of these
economies are in transition as they reform, empirical estimates often show
estimated relationships that run the wrong way. For example, it is common to
find that imports and the nominal exchange rate are positively related. In
this case, the problem is that most countries receive additional foreign
assistance on condition that they devalue their currency. Thus, a
depreciation of the exchange rate is often associated with a surge in
imports. Both variables, however, are highly correlated with a "third
variable", namely, foreign aid.
The role of simultaneous estimation is to help sort out the various
direct and indirect associations among the key variables. Nonetheless,
because most African countries have been operating well inside their
production possibility frontiers for so long, many of the normal
"trade-offs" that apply when the economy is in general equilibrium do not
hold. They only begin to take effect once most of the slack has been
eliminated from the economy.
Where does this analysis leave us? The theoretical discussion and
empirical results in Annex C suggest that that the model of retrogression
(based on endogenous spillover effects) is a useful direction to pursue.
Furthermore, as expected, the initial results suggest that there is a
curvilinear relationship between the course of the HIV/AIDS epidemic and
economic growth. Initial evidence from the model is that the elasticity of
income with respect to changes in life expectancy is less than unity. Does
this imply that some important thresholds (with respect to the supply of
skills or the performance of key organizations) have already been crossed?
Such details will have to await studies of individual countries.
In Zambia's case, however, we do know that the country has been
undergoing sustained economic decline since the mid-1970s. Despite almost a
dozen formal adjustment programs, successive Zambian governments have been
unable to reform the economy in ways that produce sustained economic growth
and development. What is clear from those who have studied the economy
closely is that the accelerating loss of skilled personnel and the
increasing pattern of institutional dysfunction among key organizations
(central bank, ministry of finance, ministry of agriculture) suggest that
economic recovery will not be automatic or rapid.78
What are the next steps? What does the above approach suggest about the
implications of HIV/AIDS on economic growth? Under present circumstances,
African countries will remain in a downward spiral so long as behavior
patterns do not change. These changes need to occur on three levels. First,
the personal and collective behaviors that lead to the spread of HIV have to
be modified. Second, governments have to ensure they devote their full
attention to promoting and sustaining economic reform. And third, foreign
agencies interested in helping African countries move beyond the ravages of
HIV/AIDS have to radically restructure their assistance so that it does not
overload the agendas of governments that are already severely
over-stretched.
Are the prospects high that these three requirements will be met?
High-risk activities in many parts of the continent are being modified. This
may be happening too slowly for some specialists. Nonetheless, knowledge
about HIV/AIDS is widespread and national leaders are beginning to speak
out, after years of irresponsible and, for many of their citizens, fatal
silence. The challenge is to sustain the progress that is being made. Of
course, for some collapsed and collapsing states (the former Zaire,
Congo-Brazzaville, Sierra Leone, Zimbabwe, Angola, Somalia, Liberia, Rwanda)
much more will have to be done before progress of even the most rudimentary
form emerges.
On the second point, African governments have development agendas that
keep them over-committed.79 The result is that much of the efforts towards
economic reform across Africa have essentially been preprogrammed to fail.
This matter can only be resolved if African governments, and those whose
representatives assemble periodically in Brussels, Paris, London and
Washington to determine Africa's development agenda, reduce what is being
attempted so that whatever is achieved can be sustained. In this regard, the
experience of Asian countries over the last three decades has a clear lesson
for Africa.80 Focus on a few priorities that matter most. Make progress on
them, and build upon that.81
Regarding the third point, the majority of donor agencies continue to
pursue a muddle of jumbled initiatives and distorted priorities. Being the
largest and most prominent, the World Bank and the International Monetary
Fund, are perhaps the easiest to categorize. After fifty plus years of
operation, World Bank management is still groping to discover the types of
actions and activities essential to economic development.82 The World Bank
supports some excellent research. Nonetheless, there has been a major
disconnect between the lessons of that research and the conclusions drawn by
the Bank's management. This has greatly diminished the contribution to
development that the Bank could have made if its resources (financial and
intellectual) had been selectively and judiciously applied.
For much of the last three decades, the IMF has attempted to soften its
'sharp-pencil' image. Its latest venture in this direction is to encourage
developing countries to formulate and implement "poverty reduction and
growth strategies." There are serious doubts about whether the IMF can (or
even should) attempt to deal with poverty in a systemic way. At the very
least, the emphasis in the strategy is back to front. Without an emphasis on
growth, poverty reduction has proven to be impossible. Furthermore, as
recent debates surrounding the Meltzer report to the U.S. Congress have
shown, we are not alone in arguing that any focus by the IMF on "poverty
reduction" misconstrues the role of the Fund and pushes it into areas where
it has a major comparative disadvantage.83
The implication is clear: to borrow Wellington's phrase, much "hard
pounding" is needed. To be effective, African citizens, policy makers, and
the donors that wish to help will have to confine their attention (and keep
it confined) to the issues instrumental to growth. Since that has not
happened across Africa in the last three decades, what are the indications
that it will begin to happen?
Unless one begins to look beyond the depth and breadth of the tragedy
that has been unfolding with the spread of the HIV/AIDS epidemic, it would
be difficult to believe that circumstances can change dramatically. Yet, as
President Clinton noted in his inaugural speech "there is nothing wrong with
America than cannot be cured by what is right in America."84 Africa and
Africans are resilient. Some, like the editors of The Economist, may have
given up and branded the continent as "hopeless."85 For others, the
challenge is to continue striving. HIV/AIDS has exacted and will continue to
exact a terrible toll. After a belated response there are indications that
major efforts are being made to change behavior and contain the disease.
These efforts will slowly take effect and provide Africans with the hope
that they can begin to think about tackling the development challenges they
face as they move beyond HIV/AIDS.
Concluding Comments
This paper describes a conceptual framework for assessing the effects of
HIV/AIDS on economic growth. The framework we use has a number of novel
features. It dispenses with the conventional approach of comparing scenarios
that purportedly pertain to situations 'with AIDS' and 'without AIDS'. We
argue that such approaches are fundamentally flawed. In particular, they
grossly misrepresent the dynamics of the HIV/AIDS epidemic.
The approach we follow takes its inspiration from endogenous growth
theory. That theory explicitly recognizes the spillover effects that result
from increasing returns to the generation, dissemination, and use of
knowledge. The spillovers raise the rate of savings and investment, and
ultimately the rate of growth. The spread of HIV/AIDS has negative
spill-over effects that undermines a country's capacity to save and invest.
To measure the impact of the epidemic, we imagine an endogenous growth model
"running backwards." Using this mechanism, we can explain why and how
economic growth will decline as HIV/AIDS intensifies.
We further enhance the dynamics of the model by taking advantage of
several other theoretical concepts --- the opportunity cost of time,
irreversibility and option values, efficiency wages, over-lapping
generations (or 'life-cycle') theories of saving, and coping strategies.
These ideas help explain the various incentives that individuals and
employers confront as they deal with issues related to HIV/AIDS. We conclude
that when the prevalence of HIV/AIDS is high and rising, as has been the
case in most African countries, the incentives are skewed towards greater
consumption and reduced investment. The consequence is a decline in the rate
of growth.
Most African countries have been under extreme economic stress as a
result of the shocks of the mid- and late-1970s and weak (mainly failed)
attempts to adjust to these shocks. The HIV/AIDS epidemic has added to the
stress. By stripping countries of some of their best talent and undermining
the incentive to invest, these difficulties may be condemning African
countries to extended periods of stagnation and decline.
The challenge for policy makers and agencies that seek to help promote
growth and development across Africa is to understand how the epidemic is
affecting the behavior and activities of everyone in society. Special
attention needs to be given to the challenge of maintaining (and even
increasing) productivity of those who are HIV-positive. Doing this will
require a shift in attitudes and the re-orientation of current programs. A
further dimension will be to help stabilize the key organizations
responsible for managing the economy and maintaining the integrity of basic
social processes. This is an area where the donor community can make an
effective contribution. Though it appears to be "turning back the clock",
donor agencies have a special role in providing technical assistance to
strengthen the performance of these organizations. In the absence of such
support, it is difficult to see how African countries, on their own, can
break out of the pattern of regression that the spread of HIV/AIDS has been
reinforcing.
Annex A
Work Place Interventions in Response to HIV/AIDS
by
Deborah A. Hoover
1. Introduction
The following literature review highlights the experience that is
emerging on work place interventions in response to HIV/AIDS in Sub-Saharan
Africa. My specific emphasis is conditions in Southern Africa. Over recent
years, the material available on work place interventions has expanded
although the data remain limited in scope and reliability.
This paper also proposes a set of criteria for developing work place
interventions. The objective is to provide readers with a detailed analysis
of the existing data, and guidance regarding their value in mitigating the
effects of the epidemic. With more than one-quarter of the working age
adults infected with HIV/AIDS in several Southern African countries, the
private sector can no longer afford to neglect the medical reality. Excuses
that action is limited because companies lack resources, or have inadequate
knowledge of the issues, or that the topic is too sensitive, are no longer
acceptable.
The criteria developed in this study draw heavily upon the author's
experience with capacity building programs in several African countries and
a thorough review of work place interventions in response to HIV/AIDS in
Sub-Saharan Africa. The criteria take for granted that management decisions
concerning the work place should respect and enforce basic human rights.
A major barrier to comparative analyses of the impact of HIV/AIDS in
Southern Africa is the inaccuracy, lack of detail, and nonexistence of
employee medical statistics.86 Even when personnel office or human resource
records are accurate regarding length of employment, they often do not
include reasons for worker separation or termination of their employment.
The continuing widespread denial and stigma associated with HIV/AIDS has
compounded this problem. Employer's records of the cause of debility or
death typically treat AIDS as an opportunistic infection. As a result,
information related to the effects of AIDS interventions in the work place
cannot be analyzed by sector, profession (or occupation), employment grade,
length of service, or gender.
With relevant statistical data so scarce, researchers have turned to
interviews as the primary methodology for determining work place practices
and responses. The interviewees usually have been managing directors, human
resource managers, or personnel officers. Medical officers and peer-group
educators are poorly represented even when those in the latter group manage
the AIDS education program. In almost every instance in the literature
surveyed, the interviewer has been working under the auspices of an HIV/AIDS
prevention organization such as UNAIDS (Joint United Nations Program on
HIV/AIDS) or AIDSCAP (Family Health International's AIDS Control and
Prevention Project). The limitations of this methodology could easily result
in the information derived being dismissed as unsubstantiated, undocumented,
inaccurate, and opinionated. Yet, aside from anecdotal information, these
are the principal data available. Similar difficulties emerge in efforts to
determine the costs of work place interventions since many of these are part
of joint activities arising from employee education, training, or medical
care.
In some cases, work place interviews have been supplemented with focus
groups or worker discussion groups. The literature, however, does not
indicate whether HIV positive employees, peer educators or medical personnel
were included in these groups. There is some evidence showing that
interviews with employee groups and labor unions contradict data from
management, especially with regard to issues such as HIV screening,
confidentiality of records, and discrimination practices.87 Similarly,
in-house medical personnel often have different views about the prevalence
of HIV among the firm's workers than the views held by management.88
Perhaps the major limitation in studies of work place interventions is
that because HIV/AIDS has only reached epidemic proportions over recent
years, companies have not been dealing with the disease for extended
periods. The relatively short period involved has also limited efforts to
measure the comparative effectiveness of different work place interventions.
Consequently, the impact of many of the activities discussed here has yet to
be evaluated for effectiveness. An added consideration is that because of
the generally ad hoc responses that have emerged as HIV/AIDS has spread,
many of the work place interventions were not specifically designed to
respond to the challenges posed by the disease. In numerous cases, concerns
about HIV/AIDS have been grafted onto ongoing education and/or training
programs. In other cases, those concerns were added to schemes designed to
restructure (often to "down-size") organizations.
2. Literature Review of Work Place Interventions
Until very recently, most work places across Sub-Saharan Africa had done
nothing to directly address the HIV/AIDS epidemic. The predominant response
often has been perfunctory and ad hoc efforts to create awareness. Most
organizations have done little more than encourage the distribution of
posters with warnings of the nature of HIV/AIDS and urging their workers to
exercise caution in personal behavior.
The principal reason for this limited response is denial. Much has been
written and said on this subject.89 In the present circumstances, some
repetition seems worthwhile because denial exists on so many levels.90
Despite the variety of HIV/AIDS activities noted below, and despite the fact
that companies now have work place interventions more than ever before, only
a minority of companies have taken the initiative with respect to HIV/AIDS.
Most companies, whether or not they have programs to address HIV/AIDS,
continue to publicly deny or disregard the disease.
Three examples illustrate this point among the business community in
Sub-Saharan Africa. In a 1995 profile of the Botswana Diamond Valuing
Company (BDVC), management reported "no visible impact of HIV/AIDS on the
company." Yet, actions taken by the company suggest otherwise. In 1993, BDVC
established one of the most extensive prevention programs of any business in
Botswana, including the hiring of an AIDS coordinator and the development of
an in-house drama program. It also sponsored condom vending and peer
educators.91 The second example is taken from Uganda where a study of the
productive labor force concluded that HIV/AIDS issues were not being
incorporated into personnel policies in the work place.92 The third example
concerns the National Business Initiative (NBI), an organization whose
membership includes 170 leading South African and international companies
that operate in South Africa. The NBI's principal objective is to enhance
the business contribution to South Africa's success by assisting member
companies in addressing some of South Africa's most pressing socioeconomic
problems.93 There is no mention whatsoever of HIV/AIDS in the two most
recent annual reports.94 The inference is that HIV/AIDS was not seen as a
"pressing socioeconomic problem." As the information presented at the July
2000 Durban conference on HIV/AIDS demonstrated, the weight of evidence and
expert opinions suggest otherwise.
Although HIV/AIDS may not have severely affected certain industries in
the past due to its five-to-seven year incubation period, the disease is now
a major epidemic, especially in Southern Africa. Indeed, Myron Essex of the
Harvard AIDS Institute has recently referred to the "new" AIDS epidemic in
Southern Africa associated with the spread of the highly virulent HIV-1C
sub-type.95 Despite the widespread evidence of the disease,96 there are
cultural and economic reasons why many companies and organizations continue
understating or denying the reality of HIV/AIDS. Evidence of this is the
general absence of forward-looking intervention strategies.
The majority of the work place interventions described below are
characterized by avoidance. Businesses have avoided facing the systemic
issues related to the spread of HIV/AIDS by finding ways to rid the employee
pool of HIV-positive employees. The majority of company interventions
typically focus on a combination of preventing new infections and avoiding
and/or reducing the costs associated with existing and probable
infections.97 A smaller number of firms have adopted interventions that
raise their costs through adjustments to employment, training and benefit
schemes. Finally, even fewer businesses have adopted what I refer to below
as 'socially responsive' interventions.
Business responses to the HIV/AIDS epidemic has been initiated by company
management, often in consultation with one or more non-governmental
organization. Examples where employees have initiated AIDS activities remain
rare and largely undocumented. This is further evidence of the stigma
attached to HIV/AIDS.98
The work place interventions described below are organized according to
four categories: prevention of new infections; cost avoidance and/or cost
reduction; adjustments to employment, training and benefit schemes; and
other, socially responsive interventions.
a. Prevention of new infections
If companies have responded to HIV/AIDS, the activities most often begin
(and end) with an HIV prevention program. The size, longevity, and funding
of these programs vary widely. The reasons given for starting such programs
differ. In some instances, management has perceived that the prevalence of
HIV/AIDS will have a significant effect on profits. Other companies suggest
a more altruistic motive. They "want to show a concern for their workers" or
"promote national awareness of the epidemic." Other companies indicate
financial cost/benefit reasons for developing prevention programs, noting
that such programs "are relatively inexpensive and well worth the cost." At
the same time, a study of the commercial sector in Zimbabwe showed that
there is little, if any, belief in the efficacy of AIDS prevention
campaigns, and little interest in prevention activities.99 Due to the lack
of hard data, there is no convincing evidence that prevention activities are
either effective in reducing HIV or cost-effective for the company.
Companies with the most promising activities for the prevention of new
infections invariably have a history of external support, whether from the
government or a donor agency such as AIDSTECH or AIDSCAP. When outside
funding is discontinued, the program often ends.100 Well-intentioned
programs frequently suffer from repetition and lack of new information and
materials, and gradually decrease to a token effort. One company complained
that its employees "got bored" with the HIV prevention activities.101 In
other cases, peer educators, working with few incentives and usually without
re-education programs "burn-out" and lose interest.102 Other companies that
had prevention programs curtailed them due to economic difficulties and
downsizing.103 When costs needed to be cut, HIV prevention program were
frequently treated as dispensable.104
The primary means of establishing AIDS prevention programs in the work
place is through education and awareness programs (including information and
discussion groups, and media presentations), HIV/AIDS counseling, testing
and treatment, the distribution of condoms, control of work place
organization to decrease the risk of HIV, and in-kind donations.
AIDS Education and Awareness Programs The goal of these programs is to
promote awareness about HIV and sexually transmitted diseases (STDs) for all
employees, or more usually, a selected subset of employees. In most cases,
education/awareness programs are operated at low cost to the company by a
peer educator or by a designated AIDS coordinator. The latter may be the
human resource officer or the company nurse, upon whom additional
responsibilities are placed often without additional resources to cover
salary, space, programming, or materials. Program funding (if any) usually
comes from outside sources. Consequently, education and awareness programs
vary considerably in scope and quality, and may consist of hosting a
one-time education program operated and funded by an outside agency.
Attendance is voluntary and the programs are inevitably directed at
employees below the management level.
Peer education programs These are one of the more common educational
awareness programs because they tend to be created by external agencies,
such as an NGO that is sub-contracted by the company. Such programs can be
operated at low cost or even no cost to the company. Peer educators are
company employees, often on a voluntary basis, and are usually from the
middle or lower level staff. They are rarely from management.105 Peer
educators are frequently responsible for the company's entire efforts in
relation to HIV/AIDS including condom distribution, information and
discussion groups, and the development of media programs.
A major drawback of work place education programs in response to HIV/AIDS
is the narrowness of the information provided. HIV/AIDS education alone is
insufficient to help people adopt preventive measures. A broader context is
needed, including facts about methods of HIV transmission, safe sex
alternatives, problems associated with drug use, prostitution, promiscuity,
and matters related to gender inequality and violence against women. The
last two, of course, raise broader social issues that are often intractable
because they threaten existing social arrangements and (male)
preferences.106
Work place education and awareness programs take one of two forms --
lectures and discussion groups, and media communication. Each is described
briefly below. Another way to classify these programs is according to their
target audience. These include: new employees, all employees, management,
and medical and human resource personnel. Each group will be discussed in
more detail below. It is most common to find only one of these audiences
targeted, or only one of these audiences at any one time.
It should be noted that, although companies frequently cite their
education programs as their primary effort in HIV/AIDS prevention, their
rhetoric often overstates the reality. The programs tend to be limited in
scope; they often take place only once or twice; they are rarely ongoing;
and they tend to be poorly attended.
Lecture and discussion groups Lecture/presentations, seminars, and
discussion groups are commonly mentioned as HIV/AIDS education and awareness
activities. Details about the content of these programs are rarely provided
in the literature. In isolated cases, companies operate health education
programs that include family planning; some companies have had discussion
groups dealing with gender inequality and discrimination, nutrition, stress
and alcohol abuse.107
Media presentations Companies use a broad variety of media communication
programs to educate and develop awareness of HIV/AIDS. The type of program
adopted correlates with the rank and education level of the employee. For
workers who are illiterate or semi-literate, folk and electronic media
presentations are common. These include videos, puppet shows, dramatic
productions, songs/choirs, and traditional dance groups. Such methods have
been effective as communication tools and sources of entertainment in
campaigns dealing with other health matters. A number of companies have
printed and distributed a cartoon booklet produced by David Whitehead
Textiles of Zimbabwe, for their semi-literate and literate employees.108
Printed material is more common for skilled employees and management.
Poster displays and brochures may be available at the health clinic, and the
company may print and distribute a free monthly or quarterly HIV/AIDS
newsletter.
HIV/AIDS education programs for new employees These activities are often
part of the induction or orientation courses for new employees. They consist
of lectures, seminars, discussions, or media presentations, often with an
HIV/AIDS education component included.
General work place HIV/AIDS education and awareness programs These
programs typically have been organized by peer educators, company nurses or
human resource personnel. They range from regularly scheduled discussion
groups to media presentations and one-time AIDS awareness activities during
National AIDS Week or on World AIDS Day. General employee programs are often
held during lunch or tea breaks, so as not to cut into work time. Attendance
is voluntary but the company may provide refreshments to improve
participation. Only occasionally are supervisors encouraged to attend and/or
lead discussions.
HIV/AIDS programs also have been incorporated into ongoing company
training sessions. In some companies, AIDS education outreach programs are
made available to employee dependents and/or members of the surrounding
community.109
Education of medical and/or human resource development personnel Programs
in this category involve sending relevant staff on training courses. The
training typically covers methods of creating AIDS awareness, peer
education, pre-and post-counseling techniques and the development of an
in-house HIV/AIDS education program. It is usually assumed that such
training will lead to the implementation of more broad-based work place
interventions. The correlation between the two, however, appears to be
weaker than might be expected. .
Education programs oriented specifically for company managers and/or
supervisors Such activities are not common. While all companies recognize
that the loss of skills at senior levels can be highly disruptive, very few
of them report HIV prevention initiatives for senior staff.110 When such
programs occur, they are usually held at the work place and are organized by
the human resource development officer, sometimes with the assistance of
outside experts. The focus is AIDS awareness, including a basic
understanding of HIV and AIDS, and acquaintance with national statistics on
HIV/AIDS prevalence. Programs include one-day seminars and one-time courses
for supervisors. For management, the activities are often confined to
special sessions during National AIDS Week or to celebrate World AIDS Day.
Briefings to management on the company's prevention program (as a means of
encouraging management acceptance, knowledge of, and continued support) have
been limited.
Many companies tend to assume that if management is familiar with
statistics about the incidence of HIV/AIDS, there will be support for
prevention activities. There is little evidence to substantiate this
view.111
HIV/AIDS Counseling, STD Testing, and Treatment Some companies offer free
or affordable HIV/AIDS counseling, as well as testing and treatment for
sexually transmitted diseases. The services are often provided in-house and
on a voluntary basis at company clinics. Some work places also offer
voluntary testing and treatment for sexually transmitted diseases and
opportunistic infections such as tuberculosis. Others offer this service for
the employee's sexual partner if the latter takes the initiative to come to
the company health clinic. Some mining companies have provided testing and
treatment to commercial sex workers within the community adjacent to the
mine.112 Other companies will also cover the salary of employees when they
participate in treatment programs. In one instance, a company clinic
arranged for employees to visit a traditional healer and receive traditional
medicines during working hours.113
Distribution of Male Condoms Probably the most widely adopted work place
intervention is the distribution of male condoms. Practices vary widely as
to whether they are distributed free of charge, or at low cost. Most
frequently, condoms are obtained from donor agencies at a minimal charge or
supplied by the Ministry of Health at no cost. Condoms are often available
through dispensing machines at company stores, rest rooms or clinics.
Otherwise, they are available through the company nurse or peer educators --
sometimes on a limited basis. Regardless of how the condoms are distributed,
demand for them has been consistently high. Companies in urban communities
often do not distribute condoms if they are readily available, reliable and
affordable in the surrounding community.
It should be noted that the literature does not provide any evidence that
companies have made female condoms available to their employees. Those are
typically distributed through donor-supported social marketing projects.
Control of Organizational and Environmental Factors Increasing the Risk
of HIV Diminishing the organizational aspects of employment and the
environmental factors that increase risk of HIV can be highly effective. One
company that has shown this is the Jwaneng Mine, Debswana, Botswana.114 In
former times, employees were provided with room and board with no provision
for their families and/or partners. The company now offers employees shared,
self-catered housing, and permits spouses to live with the employee.
Major development projects often require that large numbers of male
workers live apart from their families for extended periods of time, thus
increasing the risk of HIV/AIDS from commercial sex. Companies and/or donors
can reduce this risk by redesigning these projects by creating special
villages where workers can live with their families.115
Another concern relates to work places such as hospitals or laboratories
where there is a risk of contact with contaminated blood. Special
precautions are needed to raise the awareness of all employees to the safety
issues involved.
In-Kind Services Some companies will contribute in-kind services such as
food and beverages to enhance and encourage participation in their HIV/AIDS
education programs. Most will also grant workers permission to attend
programs during working hours. Occasionally companies will provide
transportation for HIV/AIDS activities.116
b. Cost Avoidance and/or Cost Reduction
The most widespread work place intervention used by businesses in SSA has
been to avoid or reduce the probability of hiring an employee who is HIV
positive, or from a group "likely to get AIDS." Realizing that employees can
easily contract HIV once they are on the payroll, companies will often
combine this strategy with the reduction of benefits available to infected
workers.117 Such strategies are illegal and discriminatory and are
frequently neglected or avoided in the literature. Nonetheless, they tend to
be pervasive.
What is not clear from the sources is whether the interventions described
below have been instituted as a direct response to HIV/AIDS, or to other
economic concerns. There have been numerous crosscutting themes. In many
countries in SSA, the dramatic rise in HIV/AIDS infections has coincided
with rising labor costs, affirmative action, sharp increases in health care
costs, and exposure to competitive global markets. Regardless of the cause
of these interventions, the effect on employees remains the same.
Companies Avoid Hiring Infected or High-Risk Employees Pre-employment
medical exams routinely have been a condition for job applicants. Many
companies now use these obligatory exams to screen for HIV and other
infections such as tuberculosis, malaria, and so on. Applicants found to be
HIV-positive are told that they fail their interviews.118 Many managers do
not seem to be aware that pre-employment HIV testing violates the 1988
WHO/ILO guidelines on non-discrimination against workers.
Similarly, HIV screening is often camouflaged within the regular,
long-standing medical examination procedures required for older employees.
Those who fail are dismissed.119 Labor legislation in South Africa currently
provides that incapacity and debilitation for any cause can be treated as
grounds for dismissal.120
As a matter of policy, business organizations claim that there is no
mandatory HIV blood screening for potential employees, and that employees
who are infected with HIV remain in employment until such a time when they
develop full-blown AIDS and cannot continue working. The situation differs
in practice. Uganda's National Organization of Trade Unions (NOTU) has
claimed that some employers subject prospective employees to a "quiet" HIV
screening test before recruitment while others do it openly. In a survey of
eighteen firms in Zambia, fifteen used medical examinations to screen for
HIV in new job applicants and old employees alike.121
NOTU also reported that most employees who have HIV/AIDS do not retain
their jobs - a finding that other researchers have verified. Although there
is some indication that firms are less likely to dismiss or replace middle
and senior-level managers, there have been some examples.122 Some research
indicates that employees who are chronically ill are advised or coerced into
retirement "on medical grounds."123 According to NOTU, other workers leave
their job due to fear of isolation by other workers. This saves the company
medical and possibly funeral expenses. Furthermore, although organizations
claim that there is no job insecurity for employees who develop AIDS, once
sick leave has expired many workers are laid off.
NOTU also found that the Structural Adjustment Program in Uganda was used
as an excuse for dismissing staff.124 Although laws protect workers who are
HIV-positive from being summarily fired, the economic recovery programs
often pre-suppose retrenchment as part of the restructuring of industry.
Many companies have dismissed HIV infected employees under this guise.
Modified Benefits Many employers have been reducing their benefits to
employees and their dependents. One such practice is to externalize medical
costs by making employees pay for more of their medical care expenses, such
as those of their dependents. Other companies, such as Quick Print, in
Botswana, will not pay medical expenses for any illnesses specifically
related to HIV/AIDS.125 In at least one case, medical loans to employees
have been recovered from their salary or terminal benefits.126
Other companies use their life insurance benefits for employees as a
mechanism for determining whether to invest further in their mid- level and
senior employees. To receive more insurance coverage, a medical exam is
required that includes a test for HIV. Employees who will not be tested, or
those who fail the exam, are denied further training and promotion.127
With pension funds and other benefit schemes threatened by early
depletion, some companies have introduced new funding arrangements for
employees hired after a certain date. These new pension and benefit funds
are employer-managed individual retirement accounts, with benefits based on
the contributions made into it over the period of employment.128 In other
cases, it is the health insurance providers who are reducing their ceilings
for HIV-related claims.129
"Counseling" Employees who are known or suspected of being HIV positive
are "counseled" to retire.130 In most companies, early retirement is
synonymous with discontinued medical aid and life insurance. However, since
employees are aware of this, most will opt to stay on the payroll, and make
every effort to show up at the work place so they cannot be dismissed. Cases
have been reported of employees being carried to work.131
Modification of Funeral Ceremonies In instances where employees live on
company estates, managers have been insisting that funerals be held on
weekends to reduce absenteeism. Furthermore, official company mourners are
being chosen from those employees who are off-duty on the day/s of the
funeral.132
Out-Sourcing Production Activities Some companies have begun to control
the organization of the work place in ways that will decrease the risk of
HIV. The most common means of doing so is to out-source production
activities, particularly where those activities involves workers in
high-risk groups. Transportation and logistics are examples. One large
company in South Africa dissolved its shipping department and hired
independent contractors as truck drivers. This avoided paying benefits
(which had been rising rapidly) to its drivers.133
Evidence shows that companies are also increasingly hiring staff on
casual or rolling short-term contracts, thus decreasing the need to pay
medical, disability, or terminal benefits.134 Other companies are
eliminating their unskilled workforce entirely, as the same services can be
secured at a lower cost from outside contractors.
Shifting to Capital-Intensive Production Technologies Where feasible,
companies have considered the benefits of shifting from labor-intensive to
capital-intensive production technologies.135 As losses from HIV/AIDS
increase, those businesses with large numbers of unskilled and semi-skilled
laborers, such as commercial farms or mines, have begun to mechanize.136
c. Adjustments to Employment, Training and Benefit Schemes
Some companies are exploring, or have put in place, extra recruitment,
training and benefit options. These include additional hiring, increased
insurance coverage for key positions, multi-skilling strategies, and
'succession guidance' and training. The literature does not note
specifically that these training and employment schemes were developed in
response to the rising incidence of HIV/AIDS. On the contrary, many were
ongoing company programs, or were adopted as part of downsizing.137
Nonetheless, they have been well-suited for dealing with the effects of
HIV/AIDS in the work place.
Additional Hiring In contrast to the point made earlier about capital
intensification, some companies have found it useful to adopt more
labor-intensive production methods. Two or three people have been employed
to operate machines that would normally require only one person. The
redundancy reduces the risk that if one of these people becomes ill or dies,
the work schedule can be maintained.138 For semi- and unskilled positions,
some employers have begun to recruit additional employees, or to retain a
pool of contract employees to substitute for those on prolonged sick
leave.139 Others, such as British Petroleum and Barclays Bank, have begun
the practice of "double-hiring" for key, skilled positions.140 Although this
adds to costs, it insures that the company will not have to close down
temporarily or disrupt its operations while recruiting a replacement.
Increased Insurance Coverage In industries where the loss of one highly
skilled employee could threaten the whole production process, companies have
purchased "key man" insurance to cover the costs of recruiting replacements
if they die.141
Multi-Skilling Strategies As companies experience, or foresee, gaps in
their production or services due to increased illness, they are devising
strategies to widen the skill base of their employees. One such strategy is
to promote multi-skill training at most levels of their operation.142
Employees are usually trained in-house to acquire a broader range of skills
that will enable them to fill important gaps as the need arises. This
flexibility often works to the advantage of the employee, who then has a
greater chance for promotion.
Succession Guidance and Training Numerous strategies have been developed
to address losses of experienced labor by combining career counseling with
in-house training or apprenticeship programs.143 These schemes vary
considerably in their application. In Botswana, the Sanitas Garden and
Nursery Centre has organized a tabular training form for each department
with a list of skills and product knowledge for each position. This informs
employees of the skills they need to move to different positions. Each skill
has an identified trainer. Employees shadow them to learn the job. Employees
are encouraged to map out a plan for their own career development, and
training is offered to those who are most highly motivated.
Other companies have adopted multi-tiered succession plans for identified
positions within the company.144 For each designated position, several
potential successors are identified and the necessary skills are outlined.
Employees are encouraged to take the initiative to acquire the necessary
skills through training programs offered at the company, or elsewhere. This
promotes a healthy atmosphere of competition within the company while
simultaneously offering career path counseling for employees.
The Botswana Development Corporation has adopted a plan that combines
training and lost work time strategies with succession planning, but with
compensation rather than competition as an added incentive. Certain
positions have a nominated assistant, who is usually in the grade
immediately below. This "assistant" is trained on-the-job to learn the
responsibilities of their designated superior. If the latter is absent, the
"assistant" will act on their behalf, and continue his/her own job as well.
Special pay is provided to compensate for this doubling up.
In other companies, the in-service and on-the-job training programs have
been expanded to encourage employees to broaden their skill base, but a
formal career path is not outlined.
d. Other (Socially Responsible) Interventions
The majority of these interventions relate to circumstances where the
employee already has HIV. Rather than seek to deflect the costs involved, a
number of organizations have adopted initiatives that support and sustain
their workers.
Assurance of Care and Non-Discrimination Companies support employees
infected with or affected by HIV. This enables them to receive appropriate
counseling, medical and social support. They also ensure that top management
enforces non-discrimination policies regarding HIV/AIDS, and that there are
clear procedures for dealing with personnel issues such as job termination.
Adjustments to Employee Benefits Company-provided benefits and programs
have an important role in preserving the dignity of employees who are HIV
positive, or who are suffering from AIDS, by helping them maintain normal,
productive lives for as long as possible. Some companies are restructuring
their insurance policies and benefit packages to meet the needs of
terminally ill workers and their families without bankrupting the
company.145
Examples include the extension of medical insurance to include coverage
of HIV/AIDS, and the introduction or enhancement of occupational health
clinics, pension funds, death benefits, funeral transportation and other
costs, and subsidized loans. Other companies have agreed to pay terminal
benefits to employees who are certified as terminally ill, so employees can
"retire" to die in peace, without losing these benefits.146
Hospice Programs A number of companies have established facilities for
home-based care and/or home visiting of ill people while others have
provided training for families in home-based care by company medical
personnel or through contracts with other health and/or Red Cross staff.147
In some cases, the company has provided financial support to community
home-based care schemes. (This has usually been limited to community-based
industries.) In other cases, the human resource personnel organize
volunteers to operate hospice programs in support of colleagues who are ill.
Company Foundations and Fund Raising Efforts Several companies have set
aside special funds for combating AIDS. Examples are Zimbabwe's Southampton
Life and Defy Industries, and South Africa's ALUSAF (Billiton Bayside,
Hillside Aluminum). Southampton Life has an obvious strategic interest in
dealing with AIDS. It has established a small foundation. Defy Industries
has organized fund raising to provide terminal care. Another group of
companies in Zimbabwe is reported to be establishing a trust to support
local hospitals.148
Return-to-work programs As an alternative to forcing or encouraging
employees with HIV/AIDS to retire or resign, some far-sighted companies have
designed flexible "return to work programs" which identify new job
placements for these workers who are unable to fulfill the requirements of
their former positions.149 The company will also provide them with the
necessary training. Although this option is costly, it represents an
explicit attempt to allow the worker to continue contributing in a
productive way.
3. Criteria for Developing Work Place Interventions in Response to
HIV/AIDS
The following criteria refer to a broad range of business enterprises
within the formal sector. They range from large agricultural estates and
mining companies to smaller commercial firms and public agencies. They also
draw in part from the extensive data available related to HIV/AIDS
activities in private and public enterprises in the United States, and from
literature concerning educational awareness campaigns and public health
programs.
All work place interventions should be proceeded by a planning process
A planning committee should include representatives from management,
employees, labor unions, the health profession, and local AIDS service
organizations such as NGOs. Whenever possible, it should also seek the
collaboration of employees who are affected by HIV/AIDS. Guidance should be
sought from outside organizations such as the Ministry of Health, the
National AIDS Coordination Program and the World Health Organization. The
planning process should emphasize dialogue between employees and management
rather than a top-down approach. The planning exercise cannot be a one-time
event; it needs to be ongoing.
Critical dimensions considered should be the initial identification of
circumstances in the work place that result in HIV infection,
company-specific obstacles and opportunities, and priorities for action.
This approach will help ensure that the program is designed according to
local need, and not externally driven based on available funding or donor
interest. The objective of the planning process should be the development of
a comprehensive HIV/AIDS policy and an HIV/AIDS awareness and education
program.
All HIV/AIDS programs or interventions should be guided by an HIV/AIDS
policy
Desirable features for a HIV/AIDS policy include:
The policy statement should be a written document.
It should include measures for communicating its contents to all
employees.
The company's position in respect to HIV and AIDS should be clearly
stated.
Activities proposed should reflect a commitment to the principles of
equity, confidentiality, non-discrimination, and medical accuracy.
The policy should be consistent with national HIV/AIDS policies and labor
laws, and SADC guidelines.150
An HIV/AIDS education and awareness program should be developed
An effective HIV/AIDS education and awareness program will have several
features. It should:
be ongoing
include a built-in system for program monitoring, evaluation and updating
be adequately funded
address prevention, control, and management of HIV/AIDS
include formal and informal education programs about HIV/AIDS
include input and/or involvement of staff members living with HIV/AIDS
involve peers, since they are the most powerful agents of change
require the mandatory participation of all employees
include the continuing engagement of management
extend beyond the work place to include the local community, where
appropriate
Programming needs to be comprehensive
Research shows that HIV/AIDS information alone is not sufficient to
change behavior that predisposes individuals and groups to HIV infection.151
The information provided needs to be placed in context. This implies that
work place education and awareness programs need to go beyond information
about HIV transmission, safe sex alternatives, condom availability, STD
treatment and medical care, and HIV testing and counseling. They must also
include information about alcohol and drug use, gender inequality, and
violence against women.
Programming should be sustained and supported
Work place education programs should be reviewed regularly and updated in
order to maintain accurate, current information, and to encourage continued
participation. Likewise, peer educators should be provided with on-going or
repeat training, and incentives or additional forms of motivation in order
to maintain their interest and energy.
In practice, employers tend to expect far too much of peer educators,
since their regular employment requirements are rarely reduced. If peer
education programs are scheduled for evenings, this absorbs the employee's
own time, is generally uncompensated, and takes away from time with their
family. Moreover, it is invariably difficult for educators from a lower
level in an organization to educate those at higher levels. Consequently,
management rarely benefits from these programs. It is one reason why company
managers are indifferent to these programs. Furthermore, since peer
educators are often trained by outside agencies, managers tend not to take
them seriously and do not provide them with adequate resources. If peer
education programs are used, they need to be sustained and supported from
the highest levels of the enterprise.
Programming should be regularly monitored and evaluated
The effectiveness of any work place intervention is increased if it is
regularly monitored and evaluated. Important objectives include the
increased involvement of employees in HIV/AIDS programs, measurable decrease
of drug and alcohol use, and demonstrated continued support from senior
management.
Executives need to be engaged and committed
Without commitment at the highest level, HIV/AIDS policy and
interventions will not be taken seriously. Every enterprise needs to have
executive involvement in the HIV/AIDS policy planning and programming. When
managers do not participate in these programs they provide a powerful signal
to employees, potential investors, and the general public. They show that
the company is not willing or able to address the types of socioeconomic
issues that will affect its bottom line and long-range effectiveness.
Programming should be sensitive to the considerations of the target
audience
Where appropriate, education programs should be held in the local
language.152 The timing of these programs should take into consideration the
employee population and their spouses/partners. If women are one of the
target audiences, programming should not take place during cooking hours, or
late in the evening when women are occupied caring for young children.
Furthermore, where information is provided by media presentations, there
should be interactive follow-up discussions between the presenters of the
program and the target audience. This helps insure that the issues are well
presented and are properly understood.
HIV testing and/or screening, if undertaken, should be within specified
guidelines
If employees are tested for HIV status, the results should remain
confidential. Counseling should precede any disclosure of HIV status. Those
employees who wish to know the results of HIV testing should be provided
with accurate information and, if necessary, counseling. Employers should
not require HIV screening as part of general work place physical
examinations or recruitment. When testing is offered, it should be
voluntary, informed, and confidential. Employers who insist on HIV testing
before recruitment should openly state the requirement in the job
advertisement.
4. Concluding Comments
My review of the literature on work place interventions in response to
HIV/AIDS has revealed a number of features. For a start, there has been an
encouraging rise in the scope and variety of work place interventions. As
more information is disseminated about the epidemic and its effect on the
labor force, and as more examples of successful activities emerge,
additional companies will begin to tackle the problems involved. It is
reassuring that among the earliest enterprises to adopt such programs are
mining and agricultural communities and trucking firms, where the incidence
of HIV/AIDS is the highest. As evident from epidemiological studies of HIV,
early intervention, especially among high-risk populations, is critical for
reducing the rate of infection.153
Data on work place interventions are limited and highly selective. Since
the collection of more accurate data about HIV incidence in response to work
place interventions would conflict with the need to maintain
confidentiality, this situation is unlikely to change dramatically. The lack
of data inhibits the type of analyses that can be undertaken and hence the
type of assistance that can be offered. Without longitudinal studies, the
effectiveness of the different types of assistance is also hard to
determine.
AIDS education and awareness programs are among the most common work
place interventions. With the assistance of non-government organizations,
these programs are also among the easiest to design and implement. AIDS
education does not necessarily correlate with the adoption of preventive
measures at the individual level. Even a little education, however, is far
better than none at all.
Notwithstanding the large amount of literature on work place
interventions, it needs to be understood that the predominant means of
gathering data has serious drawbacks. One-time interviews have specific
biases. These show up in discrepancies between the information provided by
the interviewee and the activities as they are implemented in practice.
Furthermore, in the absence of evidence from longitudinal studies, it is
difficult to conclude that the interventions are effective.
Such limitations have led to the emphasis on "best practice" profiles and
surveys. These provide only selective information since they give no
indication to the number or nature of work places that do not have
interventions. Furthermore, there is no way of understanding why such
interventions have not been adopted. The information gathered through these
surveys represents company actions at one point in time. It cannot be
assumed that the interventions are either ongoing or effective. Moreover,
the "best practices" may not be feasible under particular circumstances. In
the case of HIV/AIDS, even a partially effective practice is better than
none at all.
Finally, it is hard to judge what "best practices" imply under
circumstances when labor legislation and HIV/AIDS codes are non-existent or
not enforced. For whom are they "best practices" -- employees (both skilled
and unskilled)?; employers?; medical specialists?; and/or society as a
whole? Although the literature to date provides much valuable information,
it needs to be recognized that much relevant information on the effects of
HIV/AIDS and ways to ameliorate them remains unavailable.
Work place interventions in SSA can vary within one company according to
the status and skill level of the employee affected. These interventions
intersect on a number of levels: practice, cost, and ethics. Due to the
rising cost of dealing with HIV/AIDS, discriminatory practices are likely to
increase. By their actions, many companies have already determined that the
value of preventing new infections or prolonging lives is not worth the cost
for their less skilled employees since they are easy to replace. In these
cases, companies are more likely to use strategies such as outsourcing,
multiple hiring, and multi-skilling. By contrast, the same companies may
conclude that life-extending antiretroviral therapy is cost-effective for
highly skilled employees whose short term contributions are critical.154
It is the exception when companies to look beyond prevention or avoidance
strategies and consider more humane methods of dealing with the consequences
of HIV/AIDS in the work force. Little, for example, has been done to
address, or even think about, the larger psychological, demographic or
educational issues related to AIDS and the work place. Furthermore, there is
little indication that businesses are considering interventions based on
perceived changes in market conditions, declining levels of employee
motivation and morale, or counter-productive (opportunistic) behavior.
Business leaders have generally shown little willingness to confront the
broader implications of a working environment where employees' productive
lives are being dramatically foreshortened. Neither does it seem that they
have faced the reality of operating within economic systems under stress
when the very people who are needed to overcome the difficulties are also
under stress from declining income and disrupted family circumstances.
Another feature of the literature is the general absence of innovative or
imaginative long-range action to deal with the capacity problems associated
with the spread of HIV/AIDS. Capacity building efforts across Africa
continue to be undermined as the epidemic intensifies.155 Firms that have
routinely invested in expensive, long-term overseas and regional training
for their employees need to recast their efforts and replace them with
numerous opportunities for short-term and flexible on-the-job training. The
aim should be to enhance the supply of workers with the versatility and
skills needed to cover the gaps left by the workers who have died. To date,
the only interventions that begin to approach this type of thinking include
job understudies, multi-skilling, and succession training. The need for more
horizontal and vertical flexibility among employees will become increasingly
pressing.
At least two significant populations have been overlooked in the work
place interventions reviewed in this survey. The first is employees who are
retirement age and are HIV-negative. The second is female employees. Often
it is the older employees, or recently retired workers, who are adversely
affected -- financially, socially, and emotionally - by the spread of
HIV/AIDS. They could directly benefit from additional income to care for
their family members who are suffering. The benefit of encouraging later
retirement among this group, or re-hiring them, is an option that has been
given little attention. Businesses may want to reconsider existing
retirement ages by building in incentives for older workers to remain in the
work force, and to encourage recently retired, but still productive, workers
to return. Although some forward-looking work places have designed
"return-to-work" programs for employees suffering from AIDS, similar
programs should be considered for those who have retired.
Despite the increasing awareness of the inequities experienced by women
over the past decade or so, the concerns of women have been largely
neglected in the design of HIV/AIDS work place interventions. An obvious
example is that while male condoms are widely available, female condoms are
not. Women are at higher risk of HIV infection than men, and yet the social
and cultural factors which put them at greater risk have been neglected in
work place AIDS education and awareness programs. Violence against women,
and the relationship of this violence to HIV/AIDS, has received far too
little attention.156
None of the interventions outlined in this survey represents a response
that addresses long-term, skill-deepening requirements. The assumption was
that skilled workers can be obtained from "outside" the firm or
organization. This assumption is no longer valid. The prevalence of HIV/AIDS
is so high across Southern Africa that there is effectively no "outside."
Moreover, local employees who would normally succeed to skilled positions
tend to be contracting HIV earlier in their careers and are soon lost from
the work force.
All work place interventions that reduce HIV and counteract the effects
of AIDS need to be encouraged. Better data may indicate that they are far
more successful than the limited literature indicates. The data that are
available show that increasing numbers of companies have begun to respond to
the problems posed by the spread of HIV/AIDS. What is needed, however, is an
even greater willingness among managers and supervisors to confront the
implications of the losses being incurred when workers become debilitated
and productivity declines. The measures outlined in this review indicate the
types of changes that enterprises have made. The remaining challenge is to
devise ways of creating the conditions that will encourage these (and other
constructive) measures to be universally adopted.
Annex B
Framework for Thinking About Labor Force Issues in Response to HIV/AIDS
in Southern Africa
by
Donald R. Snodgrass
Labor force participation is high in Sub-Saharan Africa, even by
conventional measures which undercount the contributions of women. Moreover,
most labor force members are actively employed in some form of production,
rather than being unemployed. In the early 1990s, 68% of working-age men and
women were in the labor force.157 Of these, 94% were employed and only 6%
unemployed. Relatively few of the employed, however, held wage jobs. Ten
percent of the employed held wage jobs in the service sector, 6% worked for
wages in agriculture, and only 3% had wage employment in industry. A much
larger number -- 81% percent of the employed, or 76% of the labor force --
were self-employed or served as unpaid family workers (see table).
Economic decline in many African countries has been associated with the
expansion of the informal sector. This implies that the share of wage
employment in total employment in most countries is even smaller now than it
was in the early 1990s.
Somewhat more than one-half of all formal sector workers are government
employees. Government employment (including all levels of civilian
employment and all parastatals but excluding the military) ranged from 40 to
75% of formal employment in several Sub-Saharan African countries back in
the 1970s.158 Although these employment statistics are incomplete, dated,
and approximate at best, the employment structure across Africa has changed
very slowly over time. Thus, it seems likely that 10% or so of all workers
and more than half of wage employment are in the public sector, with a
slightly smaller percentage in private wage employment.
Sector shares of income generation and production are very different from
those of employment. There are large differences in output per worker among
and within sectors between formal and informal modes of production. The
average wage worker has a productivity level several times as high as that
of the average non-wage worker. Accordingly, the shares of GNP produced by
wage workers are much larger than their employment shares. Precise estimates
are hard to come by, and data are becoming even scarcer as national
statistical offices lose staff to HIV/AIDS. Most likely, however, the formal
sector, which generates less than one-fifth of total employment, produces
one-half or more of GNP.
Even before HIV/AIDS became widespread, labor productivity in the region
was low and growing very slowly if not stagnant or negative. Output per
worker in most African economies was lower in the early 1990s than it had
been in 1980.159 In many cases, it was lower than it had been as far back as
1965. Labor force growth in 1965-93 was rapid, averaging about 3% per annum.
This growth rate was similar to that pertaining in the Middle East and North
Africa as well as in Latin America and the Caribbean. (Labor force growth
was slightly slower in South Asia and in East Asian and the Pacific.) Annual
GNP growth, however, averaged less than 3% in Sub-Saharan Africa, compared
to 4% or more in other regions. Thus, the low GNP growth in Sub-Saharan
Africa during this period was derived from increases in the numbers of
workers, not from growth in output per worker. In 1994-97, there was some
improvement in the region's economic growth but little acceleration in
employment creation.160
The HIV/AIDS epidemic hit Africa in the early 1980s. By 1999, 23.3
million adults and children were living with HIV or AIDS in SSA (UNAIDS
1999). This represented more than two-thirds of the global total of 33.8
million cases. The adult prevalence rate was estimated at 8%, by far the
highest in the world. In no other region was it as high as 2%. In some
countries of SSA, and certainly in certain socioeconomic groups, prevalence
is 15-25% or even higher.
HIV/AIDS in Africa is spread primarily through heterosexual contact.
Fifty-five percent of HIV-positive adults are women who tend to contract the
disease at younger ages than men. This is a manifestation of women's lack of
power in negotiating sexual contact and the poverty that induces young girls
to enter into "sugar daddy" relationships with older men. The epidemic has
severely affected all levels of society. The educated middle class is at
least proportionately impacted and may have been hit even harder than the
poor. The epidemic is still growing; new infections in 1999 were twice as
numerous as deaths. Although 13.7 million Africans have died already, more
than half of those currently infected (another 12-15 million) can be
expected to die in the coming decade. Life expectancy at birth in southern
Africa, which rose from 44 years in the early 1950s to 59 in the early
1990s, will probably fall back to 45 by 2010 because of AIDS.161 The
concentration of the epidemic on young adults has created millions of AIDS
orphans and will undoubtedly result in a drop in the growth rate of the
labor force.
Following infection with HIV, an individual typically has six years or so
before he or she becomes sick with AIDS. During this period, the infected
person may or may not know that the disease is present. Physical and mental
abilities are not impaired yet, but one's morale may begin to suffer. Once
symptoms of AIDS begin to appear, the patient becomes increasingly
incapacitated and able to work less effectively and for shorter periods of
time. Death commonly occurs within one year of the onset of AIDS.
Two points in this sequence have economic significance. First, when the
individual realizes that he or she has a disease that will prove fatal
within a few years, that person's time horizon shortens. Economic behavior
can be expected to emphasize the short run and pay less heed to a longer run
that the person will not live to see. Likely results are decreased work
effort and lower savings. The second important point, which appears much
later, comes when physical and mental capacity decline and continuous
medical care is required. At that point, productivity begins to diminish and
the question of how to finance medical care must be faced. The AIDS-infected
worker's contribution to productive activity falls as his/her strength ebbs
and absenteeism rises.
Economic losses incurred from HIV and AIDS are substantial but difficult
to quantify. The nature and amounts of loss vary considerably in type and
magnitude among sectors of the economy. In own-account agriculture and
informal sector activity, little impact would be expected until AIDS-related
illness strikes. After that, work previously done by adults who have now
become sick must be taken up by others, most often children and the elderly.
Alternatively, the work may simply be left undone. Time and energy must be
reallocated from other activities, at some opportunity cost. To the extent
that labor supply is lacking or productivity falls, the family's income
level will decline. Household expenditure may also be diverted from other
types of expenditure to medical care, but only to the extent that such care
is available, utilized by the family, and requires payments of fees. Most
likely at this level of society, relatively little medical treatment will be
utilized. After the death of a young adult parent, responsibility for
raising orphaned children usually falls on surviving relatives. Their family
size increases, reducing both per capita consumption and household savings.
Agriculture and the informal sector use far smaller amounts of human
capital than formal sector activities. Many African farmers are illiterate
and few people with post-primary education work in agriculture.
Nevertheless, productivity is likely to decline as experienced farmers are
replaced either by old people whose skills are outdated or by inexperienced
young people.
In wage employment, education and training are important means of raising
productivity. High rates of return to investment in education have been
calculated for African countries, although the validity of these estimates
recently has been questioned.162 Investment in education is highly sensitive
to life expectancy.163 Shorter life expectancy greatly reduces returns to
investment in education, particularly for the more expensive forms of
investment, such as higher education. With life expectancy declining because
of AIDS, investment in education becomes a far less attractive proposition
in SSA, both for individuals and governments along with other sponsoring
bodies. Yet skills are still needed to operate public and private
enterprises and institutions. The most important form of adjustment to
shortened life expectancy would be to invest smaller amounts in skill
formation for larger numbers of people. Ways must be found to keep
productive enterprises working after key employees become incapacitated or
die.
As shown by recent research on several firms in South Africa, the
epidemic is imposing significant costs of different kinds on modern sector
firms.164 Organizations that have provided generous health, retirement, and
death benefits are finding that they must cut back on such benefits to avoid
ruinous increases in their costs.
Faced with a rising death rate and shortened life expectancy among their
employees, formal sector enterprises and organizations are likely to cut
back on training and employee benefits. Large firms often pay a premium over
the market wage as a way of attracting high-quality employees and reducing
labor turnover. As a result, when they have job openings they are typically
flooded with applicants. These employers may also feel that paying a higher
wage increases productivity as employees are better fed and more highly
motivated. When HIV and AIDS appear as a factor that lowers productivity, it
may become impossible to maintain such a wage policy. Wages and benefits are
likely to decline with deleterious effects on consumption and savings. This
in turn may contribute to a downward spiral in which demand for a firm's
products falls and further contraction ensues.
Another likely reaction is that firms will attempt to substitute capital
for labor. After all, machines do not contract HIV/AIDS. But there are
serious limits to the feasibility of this reaction, which is not socially
desirable when there is high unemployment. As their profitability falls,
firms have fewer financial resources with which to buy machinery. If demand
for their products declines, they may be willing to reduce their capacity,
rather than maintaining or increasing it through capital investment.
Finally, operating machinery effectively requires skilled engineers and
technicians. Given the scarcity of such skills in most African countries and
high prevalence of HIV/AIDS at all levels of society, the attempt to escape
the epidemic's impact on the cost of production through capital
intensification may well prove futile.
New approaches are needed to deal with this complex problem. During the
period after an employee learns that s/he is HIV-positive but before s/he
becomes ill, the main problem is one of motivation. Ways must be found to
harmonize the motives of the employee with those of the employer to the
extent possible. Later on, when sickness occurs, there are two major
problems. The first is how to meet whatever responsibility the employer is
willing to accept for the maintenance and care of the sick employee. The
second problem is how to compensate for the sick worker's declining
contribution to production.
Education and training must be important elements in any solution because
formal sector employers must find ways to replace the skills lost to AIDS if
they are to maintain current production levels, let alone bring about
economic growth. Schooling provides a general base that has demonstrable
economic benefits. More specific skills are better learned in the workplace.
Even without the AIDS problem, employers have little incentive to provide
such training because labor turnover can make it difficult to realize the
benefits. AIDS heightens this problem by making it still less likely that
firms will be able to enjoy the benefits of staff training activities. Even
without AIDS, cost-sharing mechanisms such as training funds managed by the
government or industry groups are needed to encourage such training.
Shortened life expectancy lowers the returns of training, so as HIV/AIDS
intensifies it is likely that greater government involvement and/or
inter-firm cooperation is needed. Promotion of basic education to provide
the labor force as a whole with skills that can be used in a wide range of
settings (e.g. computer literacy) is more important than ever. Ways must
also be found to encourage firms or industry groups to offer the kind of
training that provides large numbers of people with relevant, usable skills.
Annex C
Zambia Model Update: Life Expectancy, Investment, and Income Growth
by
Malcolm McPherson and Tzvetana Rakovski
1. Introduction
This annex reports econometric estimates of the impact of HIV/AIDS on
economic growth in Zambia. The results have been obtained from an existing
econometric model165 that we modified to highlight the influence of
HIV/AIDS. The original model explained income growth, government revenue,
agricultural output, inflation, the nominal exchange rate, import demand,
and mining output.
To enable us to focus on the impact of HIV/AIDS, we modified the
specification to include the effects of changes in life expectancy at birth,
the dependency ratio, and the rate of growth of the labor force. We have
also added the growth of investment to include an element that is
fundamental to the endogenous growth model discussed in the text. The
re-specified model has equations for income growth, investment, the exchange
rate, and inflation. Both theory and experience have shown that there is a
strong positive link between increasing life expectancy and economic
growth.166
Over the last decade, a strong link was evident between the spread of
HIV/AIDS and reduced life expectancy. In Zambia's case, this relationship
was confounded somewhat by the sharp decline in real per capita income and
reductions in economic welfare. Zambia experienced a significant increase in
infant mortality even before the rapid spread of HIV/AIDS.167 This adverse
trend implies that the results reported below will need to be interpreted in
the context of Zambia's acute economic problems.
Yet, even with this qualification, the model we use helps to unravel some
important macroeconomic relationships, particularly with respect to the
links between economic growth and changes in key social indexes that can be
directly tied to the spread of HIV/AIDS.
Though the simultaneous equation system we estimate is not a strict
application of the endogenous growth model discussed in the text, the system
includes the basic elements of such a model. As such the results shed light
on the non-linearity of the growth response associated with the spread of
HIV/AIDS.
2. Data: Description and Sources
For all variables, the data source is the World Bank Africa 2000 CD-ROM
database. The data cover the period from 1967 to 1998. Taking first
differences and lags into account, there are 30 observations.
Most of the variables have been defined in descriptions of the broader
model.168 Three new variables have been added -- life expectancy at birth,
the dependency ratio, and the growth of the labor force. Life expectancy is
measured as the average number of years that a person is expected to live
given prevailing age-specific rates of mortality. Because data have not been
reported for every year over the period we are analyzing, the missing
observations have been derived through interpolation. The dependency ratio
is defined as the number of dependents divided by the working-age
population. Missing observations in this series were also obtained by
interpolation.
Real income and investment are measured in billions of kwacha in constant
prices. The exchange rate is measured as kwacha per US dollar. An increase
in the exchange rate represents a devaluation of the kwacha. Domestic
inflation is measured by the change in the consumer price index. Its base is
1995. The change in US producer price index (PPI), also with a base of 1995,
is used as the index of foreign inflation. The real exchange rate is
calculated as the nominal exchange rate multiplied by the ratio of the US
PPI to the Zambian CPI. Foreign aid is taken as the U.S. dollar amounts
reported as Official Development Assistance.
3. The Model
The model has four equations that explain the growth in real income, the
growth of investment, the rate of change of the exchange rate, and the rate
of inflation. Each equation is structured in the following ways.
The growth equation reflects the conventional determinants of long run
economic expansion on the right hand side -- capital accumulation and the
growth of labor force. We have used the change in real investment as a proxy
for capital accumulation. The coefficients on investment and the labor force
are expected to be positive. The change in life expectancy has been included
in this equation as an index of human capital that directly reflects the
impact of HIV/AIDS. Its coefficient should be positive as well.
Because of Zambia's high degree of aid dependence, we have included
foreign aid in the growth equation. Productively used foreign aid should
enhance a nation's productive capacity, thereby promoting growth. Thus, the
estimated coefficient should be positive. We have also included the real
exchange rate. Though this variable is not directly considered a determinant
of real income growth, a large and growing body of literature suggests that
it should be included.169 Systematic changes (typically increases) in the
real exchange rate (the price of tradables to non-tradables) provide
tangible evidence of effective economic reform. Recent research on why
African countries have been marginalized in world trade and exchange
revealed that the principal reason is the lack of growth rather than the
lack of trade.170 Furthermore, the research is showing that African
countries have not been growing because of the gross distortions in their
principal relative prices, of which the real exchange rate is the most
important. The coefficient estimate on the real exchange rate is expected to
be positive.
The equation explaining the growth of investment includes the growth of
real income, the growth of the labor force, the change in foreign aid, the
dependency ratio, and the change in the real exchange rate. The increase in
real income should raise the rate of investment. The growth of the labor
force has two separate effects on the growth of investment. An increase in
the labor force raises the output of investment goods. However, as the labor
force expands, producers can only maintain capital per worker if they raise
investment commensurately.
The coefficient on foreign aid is expected to be positive. In principle,
the majority of foreign assistance is meant to expand productive capacity.
For example, in its 1994 World Development Report, the World Bank
highlighted the importance for economic growth of investing in
infrastructure. A major focus of foreign aid over the last five decades has
been such investment. Yet, there could also be a tenuous link between
foreign aid and income growth. External support may support investment
activities, but it also frees up domestic resources that may not be used
productively. That, of course, has been a major problem in Zambia where for
many years, foreign assistance allowed the government to divert its own
resources to food subsidies and supporting an over-blown and inefficient
public sector.
The equation for the nominal exchange rate has a straightforward
specification. It incorporates the idea of purchasing power parity and
includes both domestic and foreign inflation as regressors. It also includes
real income to measure the impact of overall economic activity on the
foreign exchange rate.
The inflation equation is derived from the demand for money. The growth
of real income, interpreted as a measure of real supply, is expected to have
a negative coefficient. Since rapid changes in the money raise prices, the
coefficient on the growth of the money supply should be positive. The
exchange rate links the equation to the rest of the system. Lagged inflation
is a measure of the speed of price adjustment over time.
4. The Results
The estimated coefficients have been derived using three-stage least
squares (3SLS). This generalized instrumental variable technique provides
consistent estimates in the presence of lagged dependent variables,
endogenous regressors, and error terms that depart from the (normally
assumed) white noise disturbances (i.e., independently and identically
distributed error terms). Table 1 has the results. The equations are
arranged by column. The first column contains the growth equation, the
second the investment equation, and so on.
In the growth equation, investment has the correct sign but is not
significant at the standard levels. The growth of the labor force and the
change in the real exchange rate have negative coefficients, although
neither is statistically significant. Life expectancy shows the expected
positive effect on growth.171 The change in foreign aid has a highly
statistically significant negative effect on income growth. The highly
significant coefficient on income growth is not surprising given Zambia's
history. There has been little structural change over the three decades
considered and virtually no growth. That is, the current period change in
real income departs little from past changes.
In the investment equation, the coefficient estimates for the labor force
and dependency ratio have the correct signs (positive and negative
respectively) and both are significant at 5 percent. An important result is
that neither foreign aid nor the growth in real income appear to have
influenced the trend in investment. The change in the real exchange rate was
negative and statistically significant. This is contrary to expectations. A
depreciation of the real exchange rate should stimulate investment. The
outcome could be the result of several contradictory factors. The cost of
imported capital rises as the exchange rate depreciates. Real activity is
stimulated as the price of tradables and nontradables are realigned. While
foreign aid adds to investible resources it also tends to overvalue the real
exchange rate. Finally, the coefficient on lagged investment is positive and
statistically significant. One implication is that both producers and
consumers have been responding to the incentives of investment in a
structured and adaptive way.
In the exchange rate equation, the coefficients of domestic and foreign
inflation have the correct signs and are statistically significant. They
are, however, statistically different from the values of 1 and -1,
respectively, indicating that the exchange rate deviates from the path
consistent with purchasing power parity. The negative and insignificant
impact of the real income growth is an anomaly. The expansion of real income
should lead to a depreciation of the exchange rate as domestic demand spills
over to tradables.
In the inflation equation, the real income has an unexpected positive
sign. This result is robust to a variety of model specifications. A
potential explanation is that over the last three decades the Zambian
economy has been severely distorted and unbalanced. Thus, we should not
expect what are 'equilibrium conditions' under such conditions. From a
policy perspective, the implication is that the distortions and imbalances
should be selectively removed. This would restore the correct negative
relationship between inflation and real supply. The growth of the money
supply has the expected positive effect. As expected, exchange rate
depreciation raises the rate of inflation. The coefficient is significant at
10 percent. The lagged inflation term shows that the annual rate of
adjustment to changing prices has been sluggish.
For comparison, we present the results from a single equation estimation
(ordinary least squares, OLS, estimation with robust standard errors) of the
growth equation of the system.
In this instance, the growth of investment has the expected positive
coefficient and is statistically significant. The change in the real
exchange rate is negative (as it was earlier) but is now statistically
significant. The other estimates are similar to those obtained from the
system of equations.
It is worth noting that in both sets of results the estimated regression
coefficient on the change in life expectancy at birth is almost identical.
This is a robust and important result. It shows that when systemic effects
are taken into account and when they are removed, the responsiveness of
economic growth to the change in life expectancy does not change. With a
coefficient of 0.68, economic growth is significantly responsive to changes
in life expectancy. Thus, a one percent reduction in average life expectancy
(half a year on average when life expectancy is 50 years) reduces the
average growth rate by .68 percent. This is a major source of income loss.
Since HIV/AIDS has been a major factor responsible for the sharp reduction
in life expectancy at birth over the last decade and a half,172 this
supports the contention in the main text that the impact of HIV/AIDS on
economic growth is non-linear. The capacity for economic growth in Zambia
has been increasingly undermined by the rapid spread of HIV/AIDS.
Concluding Comments
The results from our model are highly suggestive. The sharp decline in
life expectancy during the latter part of the 1990s in Zambia directly
reduced the rate of growth. The change in the dependency ratio also produced
a significant reduction in the growth rate of investment. This, in turn, fed
back to further reduce the growth of income. Notwithstanding the
deterioration of other social indicators, the precipitous drop in life
expectancy in Zambia can be attributed, in large part, to the HIV/AIDS
epidemic. The implication is that HIV/AIDS has significantly reduced
Zambia's rate of growth.
The results also show that shifts in the dependency ratio have had
important macroeconomic effects. As noted in the text, there are many
potential explanations for the relationships being measured in the model.
One of these, reflected in the dependency ratio, is the systematic reduction
in the number of workers who are supporting the number of consumers. Seen
from an overlapping generations perspective, we would expect growth to
decline. The above model is consistent with that outcome. Because investment
directly enhances growth, the inference is that the change in the dependency
ratio has indirectly reduced growth.
Since the model does not directly connect HIV/AIDS to declining life
expectancy, the model does not show conclusively that HIV/AIDS has reduced
the rate of growth. Yet, to the extent that there is a connection between
HIV/AIDS and falling life expectancy at birth and a rising dependency ratio,
the model provides evidence that HIV/AIDS has been undercutting and will
continue to undercut the capacity for economic growth in Zambia. For the
future, the policy challenge is to ensure that the negative impact on
economic growth of HIV/AIDS is minimized.
References
AIDSCAP (1995) "Private Sector AIDS Policy: African Workplace Profiles."
Washington, D.C.: Family Health International.
AIDSCAP (1996) AIDS in Kenya Socioeconomic Impact and Policy Implications
(Senior editors: S. Forsythe and B. Rau) USAID AIDSCAP/Family Health
International
Addo, S.K. (1998) "Trickle up micro grants and positive living with
HIV/AIDS" International Conference on AIDS, abstract 24291
Arrow, K.J. (1974) the limits of organization New York: W.W. Norton & Co.
Barro, R.J. (1997) Macroeconomics Cambridge: The MIT Press
Barro, R.J. (1999) Determinants of Economic Growth A Cross-Country
Empirical Analysis Cambridge: The MIT Press
Beer, C., A. Rose and K. Tout (1988) "AIDS - The Grandmother's Burden"
Bennell, P. (1996) "Rates of Return to Education: Does the Conventional
Pattern Prevail in Sub-Saharan Africa?" World Development Vol. 24, No. 1
(January): 186-199.
Bollinger, L. and J. Stover (1999) "The Economic Impact of AIDS in
Zambia." Glastonbury, CN: The Futures Group International, September
Calamitsis, E., A. Basu and D. Ghura (1999) "Adjustment and Growth in
Sub-Saharan Africa" Working Paper of the International Monetary Fund
Carroll. J. (2000) "Stopping Africa's AIDS nightmare" The Boston Globe
Tuesday, 18th January
CDC (1998) "AIDS in the Workplace Survey is a Wake-up Call for American
Businesses." Vancouver. [Available at: www.cdc.gov/nchstp/hiv_aids/media]
Cohen, B. and J. Trussell (eds.) (1996) Preventing and Mitigating AIDS in
Sub-Saharan Africa: Research and Data Priorities for the Social and
Behavioral Sciences, Washington D.C.: National Academy Press
Cohen, D. (1997) "The HIV Epidemic and Sustainable Human Development"
UNDP HIV and Development Programme Issues Paper no. 29
Cohen, D. (1998) "Poverty and HIV/AIDS in sub-Saharan Africa" UNDP HIV
and Development Programme
Collins, J. (1997) "Leveraging Private Sector Support for HIV/AIDS
Prevention: Opportunities and Obstacles. A Report on Zimbabwe and Brazil."
Washington, D.C.: Family Health International/AIDSCAP.
Commonwealth Secretariat (2000) HIV/AIDS in the Commonwealth 2000/01
Kensigton Publications Ltd.
Engels, F. (1845, 1958) The Condition of the Working Class in England
(Translated by W.O. Henderson and W.H. Chaloner) Stanford: Stanford
University Press
Essex, Myron (1999) "The New AIDS Epidemic," Harvard Magazine
September/October
Fleming, A.F. et al. eds. (1988) The Global Impact of AIDS New York: Alan
R. Liss Inc.
FHI=Family Health International, see
http://resevoir.fhi.org/en/gen/corpreport/cr25yrs.html
Forsythe, S. and B. Rau, eds.(1996) see AIDSCAP op. Cit.
Ghura, D. and M.T. Hadjimichael (1996) "Growth in Sub-Saharan Africa" IMF
Staff Papers, Vol. 43, No. 3, September: 605-634.
Gillis, M., D. Perkins, M. Roemer and D. Snodgrass (1996) Economics of
Development 4th Edition New York: W.W. Norton & Company.
Green, J. (1998) "Employers Learn to Live with AIDS." HR Magazine
February.
Grindle, M.S. (ed.) (1997) "The Good Government Imperative: Human
Resources, Organizations and Institutions" Ch. 1 in (ed.) M.S. Grindle
Getting Good Government: Capacity Building in the Public Sectors of
Developing Countries Cambridge MA: Harvard University Press for the Harvard
Institute for International Development
Grossman, G.M. and E. Helpmann (1995) "Technology and Trade" in (eds.)
G.M.
Grossman and K. Rogoff Handbook of International Economics vol.III
Amsterdam: North Holland
Hall, J. (1999) "Row Shatters AIDS Complacency." World, 19th July
Hayek. F. (1945) "The possibility of social organization" American
Economic ReviewÖ
Hill, C. and M.F. McPherson (1999) "Economic Growth and Development in
Zambia, Further Thoughts on the Way Forward" Report prepared for the Embassy
of Sweden, Lukasa, December
Hirschleifer, J. (1970) Investment, interest, and capital Englewood
Cliffs: Prentice-Hall
Hoover, D.A. and M.F. McPherson (1999) "Capacity Building in the Ministry
of Finance, Zambia" HIID Development Discussion Papers no.704, June
Hoover, D. and M. McPherson (2000). "HIV/AIDS in the Commonwealth:
Capacity Building Programmes - Facing the Reality of HIV/AIDS." London:
Commonwealth Secretariat (forthcoming July 2000)
ING Barings (2000) "The Economic Impact of AIDS in South Africa A Dark
Cloud on the Horizon" April
International Labor Office. (1998) World Employment Report 1998-99:
Employability in the World Economy: How Training Matters. Geneva:
International Labor Office.
International Labour Organization (1995) The Impact of HIV/AIDS on the
Productive Labour Force in Uganda, East Africa Multidisciplinary Advisory
Team (EAMATDMT) International Labour Organization, Addis Ababa. Working
Paper # 4.
Johnson, D.Gale (2000) "Population, Food, and Knowledge" The American
Economic Review vol.90, no.1, pp.1-14
Kambou, G., S. Devarajan, and M. Over (1992) "The Economic Impact of AIDS
in an African Country: Simulations with a Computable General Equilibrium
Model of Cameroon" Journal of African Economies vol.1, no.1, pp. 109-130
Kremer, M. (1993) "Population Growth and Technological Change: One
Million B.C to 1990" Quarterly Journal of Economics vol.108, no.3, August,
pp.681-716
Lindauer, David. (1981) "Public Sector Employment in Africa: Facts and
Concepts." Studies in Employment and Rural Development No. 68. Washington,
D.C.: World Bank.
Lindauer, David. (1995) Labor and the Growth Crisis in Sub-Saharan
Africa. Regional Perspectives on World Development Report 1995. Washington,
D.C.: World Bank.
Loewenson, R. et al. (1999) Best Practices: Company Actions on HIV/AIDS
in Southern Africa, UNAIDS Intercountry Team for Eastern and Southern Africa
Pretoria, February.
Mankiw, N. Gregory (1997) Macroeconomics 3rd Edition New York: Worth
Publishers
McPherson, M.F. and C.F. Zinnes (1992) "Institutional Weakness, Social
Norms, and Economic Retrogression" HIID Development Discussion Paper no.
223, May
McPherson, M.F. and T Rakovski (1998) "Exchange Rates and Economic Growth
in Kenya: An Econometric Analysis" HIID Development Discussion Paper #651.
McPherson, M.F. and T. Rakovski (1999) "A Small Econometric Model of the
Zambian Economy" HIID Development Discussion Paper #672.
McPherson, M.F. and T. Rakovski (1999a) "Financial Deepening and
Investment in Africa: Evidence from Bostwana and Mauritius" HIID Development
Discussion Paper No. 727, October
McPherson, M.F. and T. Rakovski (2000) "Trade and Growth in Sub-Saharan
Africa: Empirical Evidence" revised draft, HIID, Cambridge
Michael, K. (1999) "Best practices: a review of company activity on
HIV/AIDS in South Africa." AIDS Analysis Africa, 10(3) October/November
Michael, K. (1999a) "Unbelievable: AIDS Reporting in the Business Press."
AIDS Analysis Africa, 10(4) December 1999/January 2000
Myrdal, G. (1957) Rich Lands and Poor, the Road to World Prosperity New
York: Harper, chapters 2-5
National AIDS Fund.(not dated) "Business Case: HIV:AIDS Awareness."
Online at www.aidsfund.org/business.htm.
National Business Initiative (1999) National Business Initiative Annual
Report 1998/99 Colorpress, Johannesburg
Ng, H. (2000) "AIDS in the Media." Harvard AIDS Review, Fall 1999/Winter,
Okello, D. and K. Owino (2000) "Report of Conference on Restarting and
Sustaining Growth in Kenya Held at Safari Park Nairobi, 26th January, 2000"
draft report
Quattek, K. (1999) "The Demographic Impact of AIDS on the South African
Economy" ING Barings, Johannesburg, 17th December
Perkins, D.H. (1994) "There are at least three models of Asian
Development" World Development vol.22, no.4, pp.651-665
Pissarides, C.A. (1997) "Learning by Trading and the Returns to Human
Capital in Developing Countries" World Bank Economic Review, vol.11, no.1,
January
Ram, Rati and T. W. Schultz (1979) "Life Span, Health, Savings, and
Productivity."< Cultural and Development Economic> 27: pp.399-421.
Rodrik, D. (1998) "Trade Policy and Economic Performance in Sub-Saharan
Africa" NBER Working Paper No.6562, May.
Romer, D. (1996) Advanced Macroeconomics New York: The McGraw-Hill
Companies
Rosen, S., J.L. Simon, D.M. Thea, and J.R. Vincent. (2000) "Care and
Treatment to Extend the Working Lives of HIV-Positive Employees: Calculating
the Benefits to Business." South African Journal of Science 96(6) pp.300-34.
Rosen, S., J.R. Vincent, J.L. Simon, G. Singh, D.M. Thea (2000) "A model
for assessing the costs of workforce HIV/AIDS." Forthcoming in the extended
abstract volumes of the XIII International AIDS Conference, Durban, South
Africa: 9-14 July, 2000.
Rugalema, G., with S. Weigang and J. Mbwika.(1999) "HIV/AIDS and the
Commercial Agricultural Sector of Kenya", UNDP and FAO.
Simon, J.L., S. Rosen, A. Whiteside, J. R. Vincent, and D.M. Thea. (2000)
"The Response of African Businesses to HIV/AIDS." HIV/AIDS in the
Commonwealth 2000/01 London: Kensington Publications.
Smart, R. (1999) "HIV/AIDS in the Workplace: Principles, Planing, Policy,
Programmes and Project Participation." AIDS Analysis Africa 10(1) June/July
Smith, A. (1776) in Cannan edition (1937) The Wealth of Nations New York:
The Modern Library.
Sunday Times (South Africa) (1997) "How AIDS will hurt SA where it
matters most" April 5th
Tarantola, D. (1988) "Global Strategy for the Prevention and Controls of
AIDS" in The Global Impact of AIDS New York: Alan R. Liss, Inc., pp.207-214
UNAIDS (1998) "HIV/AIDS and the workplace: forging innovative business
responses" Technical update, July
UNAIDS (1999) AIDS Epidemic Update, UNAIDS Joint United Nations Programme
on HIV/AIDS, Geneva: UNAIDS/WHO, December
Wehrwein, P. (1999) "The Economic Impact of AIDS in Africa" Harvard AIDS
Institute, Fall 1999/Winter 2000
Whitelaw, K. (2000) "AIDS in the Classroom." U.S. News, February 14th,
[available at: www.usnew.com/usnews/issue/000214/aids]
World Bank (1988) World Development Indicators Washington, D.C.: Oxford
University Press.
World Bank (1999) Confronting AIDS: Public Priorities in a Global
Epidemic, World Bank Policy Research Report New York: Oxford University
Press
World Bank (1999a) Intensifying Action Against HIV/AIDS in Africa:
Responding to a Development Crisis Washington D.C.: The World Bank Africa
Region
World Bank (2000) Can Africa Inherit the 21st Century Washington, D.C.:
Oxford University Press.
World Bank (1993) World Development Report: Investing in Health
Washington D.C.: Oxford University Press
Zambia, Ministry of Health (1997) Zambia Demographic and Health Survey
1996, Preliminary Report Lusaka: Central Statistical Office and Ministry of
Health
Zambia, Government of (1999) HIV/AIDS in Zambia: Background Projections
Impacts Interventions, Ministry of Health, Central Board of Health, Lusaka,
September.
Endnotes
1 "AIDS: A Crisis in Development" Lecture to celebrate the 25th
anniversary of the Harvard Institute for International Development, Askwith
Hall, Harvard University, 9th October 1999. There are many examples. Kambou,
Devarajan and Over (1992) reported a WHO estimate of the cumulative deaths
from AIDS in Africa by 2000 as 5.5 million (p.112). The cumulative total, in
fact, will exceed 15 million.
2 This is evident from the earliest models from Kambou et al. (1992) to
ING Barings (2000) both of which (and others) have been cited below.
3 Although he did not develop a formal model of the process, Cohen (1997)
summarized the nature of the problem when he noted:
There are many reasons to believe that the effects of HIV will be to
reduce total savings, and in so far as these decline there will be less
investment, lower incomes, a slower rate of GNP growth, and possibly a lower
level of GNP (p.3).
Yet, even Cohen, whose work on the impact of HIV/AIDS has been most
perceptive and far-sighted was tempted to offer a guess at the ultimate
impact of the disease. In a subsequent article (Cohen 1997), he stated:
The evidence from high prevalence countries in Africa who are
experiencing more mature epidemics is that growth rates of GDP may be
reduced by 0.5 to 1.0% per annum due to the epidemic. (p.5)
4 This does not imply that nothing is being done. Some of the responses
are clearly unsustainable. One experiment to improve morale in Uganda
involved providing a subsidy equivalent to U.S.$100 those taking part and
then observing their response (Addo 1998). With the number of people in
Africa with HIV/AIDS approaching 30 million, such an experiment would cost
$3 billion. That amount of resources for this one aspect of dealing with
HIV/AIDS is simply not available. Moreover, one would have to question the
effectiveness of such one-off payments.
5 This point was echoed by Cohen (1999) whose extensive work on HIV/AIDS
over several years led him to conclude:
It needs to be stressed at the outset that much of the applied research
on socio-economic causes and consequences of the HIV epidemic in sub-Saharan
Africa has yet to be done (p.1).
He continued:
Ösuch research on both the causes and consequences of the epidemic needs
to be timely -- Ö -- but are generally everywhere under-recognized.
Specific examples of the organizational effects of HIV/AIDS are provided
by Hoover and McPherson (1999, 2000) in Zambia and by Okello and Owino
(2000) for Kenya. The latter authors noted:
In the Ministry of Agriculture alone some preliminary estimates, though
not scientific, suggest that nearly 50% of our workers who have died over
the last five years have died of causes which could be related to HIV/AIDS
6 For an exposition of the view that Africa was "on the move" see Madavo
and Sarbib (1997), Calamitsis (1996), and Camdessus (1997). McPherson and
Goldsmith (1998) provide a different perspective.
7 Recognizing the caveats attached to the numbers the prevalence of HIV
in some countries is high. For instance, Essex (1999) pointed out:
[In Botswana]Ö and in Zimbabwe, 35 percent to 40 percent of all pregnant
women are infected now; in some towns and villages, 50 percent are infected;
and in young adults, aged 20 to 30 years, 45 percent to 50 percent are
infected nationwide.
A report on Cable News Network on 29th April 2000 focused on a fishing
community in western Kenya (Koma Bay) where HIV prevalence among adults is
around 70 percent. For Kenya as a whole the prevalence is 15 percent of the
adult population.
8 As noted in the text, current approaches to the disease are both
important and well intentioned. They do not, however, address the questions
of replacement of high-level skills, the stabilization of key institutions
(such as central banks, revenue departments, ministries of finance and legal
affairs, and departments of public administration). In Annex A, Hoover
suggests mobilizing retired workers. This paper adds the suggestion that
donor agencies should consider major increases in direct technical
assistance combined with sharp cuts in the development agendas of all
African governments.
9 The World Bank has produced numerous studies, of which Confronting
AIDS: Public Priorities in a Global Epidemic (World Bank 1999) is only one
example. International conferences stimulate further outpourings. The recent
(July 2000) stimulated a special edition of the Science (23rd June, 2000).
It also prompted the Commonwealth Secretariat to sponsor approximately 70
studies, a selection of which were compiled in Commonwealth Secretariat
(2000). Two of the authors of this essay contributed to that volume
(McPherson and Hoover 2000). Under CAER II, USAID has sponsored a number of
studies, of which the present study is one.
10 Over et al. 1988; Kambou, Devarajan, and Over 1992.
11 The most recent example is the World Bank Policy Research Report
"Confronting AIDS: Public Priorities in a Global Epidemic", Oxford
University Press, 1999.
12 Moore (1999) examines these costs from a company's perspective.
13 Ironically, one study actually suggests a positive economic impact --
that national per capita income may rise in some cases due to the steep
decline in population ("Aids and Development" Indicator SA, 1998, vol.15,
no.3, pp.57-58). Work done at the Botswana Institute for Development Policy
Analysis has reached a similar conclusion. The key industries (diamond
mining, beef exports, and tourism) are not particularly vulnerable to the
loss of labor. (Personal communication with BIDPA researchers in March
1999).
14 BBC website "AIDS statistics likely to be conservative" 14th
September, 1999.
15 Hoover and McPherson 2000.
16 Loewenson et al. 1999.
17 Dr. Samuel Johnson is well known for observing that nothing
concentrates the mind as the prospect of being hanged in the morning. The
implication, however, is that if the person were not hanged, they would have
a "full life" ahead of them. That does not hold for those who are
HIV-positive. The question then arises how will people react who know they
have no prospect of a "full life"? What actions are required of others to
help them remain "focused" and productive?
18 Annex A has a number of references. See also Simon et al. 2000 and
Rosen et al. 2000a,b.
19 The reason is the oft-noted logical fallacy of composition. Within any
society, a few individuals can change their behavior without having much
impact on the majority. Large numbers cannot do the same. As increasing
numbers of employers shift the costs to families and the rest of society,
the pressures on existing health facilities will increase. Resources will be
diverted from both private and public sources to begin meeting the
additional demand. This diversion of resources will reduce real effective
demand and lower the public resources available to repair infrastructure and
provide other public services. The outcome will be lower demand for the
output and higher operating costs for all firms (including those that
originally began shifting the costs of HIV-positive workers).
20 Kambou, Devarajan, and Over 1992.
21 Essex (1999) described the spread of the HIV-1C sub-type as
constituting a "new" epidemic in Africa.
22 Giving the keynote speech at the January 2000 conference "Restarting
and Sustaining Growth in Kenya" Professor S. Migot Adholla, Permanent
Secretary of the Ministry of Agriculture and Rural Development, Government
of Kenya, stated:
Over 1 million Kenyans have died of AIDS related illness and we must
clearly wake up to the fact that some of the growth projection that we are
making today may not be achievable.
23 This critique can be applied broadly to all with/without studies that
are derived from some "base-line" setting. For example, ING Barings has
supported some recent work on the impact of HIV/AIDS on the South African
economy (Quattek 1999, ING Barings 2000). At best, both are examples of
potential trends in the course of HIV/AIDS in South Africa if one assumes
that the epidemic has no further impact on the structure of the economy from
this point on. One of our main arguments in this paper is that particular
assumption is invalid.
24 This is not a criticism of the actual computations made by Kambou et
al. Based on the information they had at the time, Cameroon would experience
a loss of 30,000 workers. This did not represent a major structural shift.
In this respect, it is worth noting that the HIV/AIDS epidemic, though
devastating, has only been of major economic significance for a decade or
so. The WHO launched its Global Programme on AIDS in 1st February 1987
(Tarantola 1988). By 1988, the generally accepted estimates of the number
infected world-wide was 60 thousand. One-third of these were in the United
States where 25,000 people had died from the disease (Fleming 1988).
25 Sunday Times, 5th April, 1997 A more recent ING Barings report (April
2000), noted below, suggested that the loss of growth in 2006 when the
HIV/AIDS epidemic was expected to peak in South Africa would reduce average
growth by .3 to .4 percent per year. This estimate was reported in the
coverage by Le Monde of the Durban conference on HIV/AIDS in July 2000.
26 Evidence of this point is widespread and recent (Boston Globe October
1999 "AIDS and the African: The Official Silence"; Swarns 2000; Carroll
2000; Cable News Network 29th April 2000 report on Koma Bay in Kenya.) Many
earlier references could be cited as well (Forsythe and Rau 1996; World Bank
1999). The CNN report on AIDS in Kenya on 29th April 2000, noted that
President Moi only "went public" about HIV/AIDS in 1999.
27 The prevalence of HIV/AIDS in the population 15-49 in 1997 was 12.9
percent (World Bank 1999, Statistical Appendix Table 4). The ING Barings
(2000) study gives the prevalence as 16 percent of the adult population.
Recent work by a team led by Jon Simon at Harvard University in South Africa
indicates that the prevalence is significantly higher than this (personal
communication with Simon).
28 Clear exceptions to this were President Sir Dawda Jawara of The Gambia
and President Abdou Diouf of Senegal. Both leaders were convinced during the
early stages of the epidemic (the former by researchers at the Medical
Research Council in Fajara, the latter by researchers from the Pasteur
Institute in Dakar) that public policy had to address the problem. As early
as 1985, The Gambia had a public awareness campaign to combat the spread of
HIV. Senegal's campaign drew in the media and local pop artists to advertise
the importance of safe sex. The success of both efforts has been reflected
in rates of HIV/AIDS that are orders of magnitude lower that what is being
experienced across Southern Africa.
29 This, however, is not a new problem. In The Conditions of the Working
Class in Manchester 1845 Frederick Engels introduced the concept of "social
murder" (Engels tr. Henderson and Chaloner, p.32). Engels was particularly
exercised about social policies that generated and sustained such degrees of
abject poverty that it resulted in significant numbers of death through
starvation and disease. This idea clearly carries over to the policies
(implicit and explicit) which prevented African leaders, especially in
Southern Africa, from addressing the problem of AIDS much earlier.
30 World Bank 1999, pp.44 ff, Ch.6.
31 HIV/AIDS in Zambia (Ministry of Health, Lusaka), September 1999,
pp.37,57. In this case, lack of knowledge can be no defense. In a
demographic and health survey commenced in August 1995, survey workers found
that there was almost complete knowledge of AIDS and its impact. Of men
survey, knowledge of AIDS exceeded 97.8 percent in every category. For women
it was above 98.1 percent in every category including those who had no
formal education (Zambia, Ministry of Health 1997, pp. 22-23).
32 This did not prevent The Boston Globe excoriating Chiluba for his
absence from the conference in its series "Aids and the African" 12th
October 1999.
33 The author was recently in Kenya (April 2000) and had the opportunity
to review the government's "Interim Poverty Reduction Strategy Paper
2000-2003" (Government of Kenya 2000). It was apparent from discussions with
government officials that the principal motivation for formulating the
poverty reduction strategy was to induce the donors to begin providing aid.
On its own terms the poverty reduction strategy was far too ambitious given
the capacity of the GOK.
34 World Bank 1999, Ch.6.
35 This issue is discussed in more detail in McPherson (1998, 1999).
36 McPherson and Zinnes 1991, 1992.
37 Kuznets 1966; Chenery and Syrquin 1975; Syrquin and Chenery 1989.
38 From the first empirical studies undertaken by Colin Clark, the
decline in the relative share in GDP of agriculture had been taken as both a
symbol and index of "development." Agricultural economists (e.g. Schultz
1957) were concerned when the income share dropped much more rapidly than
the share of the work force. In developed countries this posed a 'farm
problem', where "too much" labor was retained in agriculture. In developing
countries, the problem was typically a 'food problem' where, despite large
amounts of labor in agriculture, productivity was so low that output was
inadequate to support improved supplies of food and fiber. In Zambia's case,
this had not occurred because of a decline in the rate of growth, but
through a collapse in the growth process. The per capita contribution of
agriculture to national income had fallen more slowly than per capita income
in the non-agricultural sector.
39 Though even with drought, there have been many 'endogenous' responses
such as macro water harvesting and reforestation in Morocco and Burkina
Faso, tidal irrigation in several West African countries, the switch to more
drought resistant crops in Mali, Senegal, and Sudan, and the abandonment of
drought-sensitive activities such as maize in The Gambia and cattle grazing
in Mauritania that represent an internal adaptation to changing weather
pattern. Glantz (1989) also pointed to internal changes in response to
drought such as the build up of buffer stocks and the inclusion of more
weather-induced fluctuations in public planning.
40 This is equivalent to a rise in the rate of discount attached to
future benefits.
41 There is now considerable evidence showing that the rate of spread of
HIV falls significantly if other sexually transmitted diseases are treated.
42 Hoover and McPherson 2000.
43 The basic references are Romer (1986) and Lucas (1988). Expositions
may be found Romer (1996, Ch.3), Mankiw (1997, Ch.4), Barro (1998, Ch.11).
44 Arrow 1962.
45 Hyami and Ruttan 1971.
46 Kremer 1993; Johnson 2000. Johnson argued that since the supply of
natural resources on the Earth has been given, the only 'real' source of
economic growth and development since mankind evolved has been our
collective increase in knowledge-based means of manipulating that given
supply or both fixed and reproducible resources.
47 Pissarides 1997. The essence of his argument is that, because of the
size of the world market relative to domestic market opportunities,
entrepreneurs in developing countries experience a quantum shift in their
efficiency through the education they obtain by having to compete and trade.
Because of inter-industry linkages the spillover effects raise productivity
in the non-tradable sector.
48 Viewed in terms of the "Australian model" of internal and external
equilibrium (Gillis et al. 1996, Ch.20), a highly distorted economy is well
within its production possibility frontier (linking tradables and
non-tradables). Economic reform can produce sharp increases in output as the
economy moves "outward" towards the frontier.
49 Adam Smith 1776, Cannon edition 1937, Ch.2.
50 Myrdal 1957, Chs. 2-5.
51 This has been well documented in a broad-based study entitled
"Restarting and Sustaining Growth and Development in Africa" conducted under
the USAID-funded Equity and Growth through Economic Research (EAGER)
project. The basic approach taken in that research is described in
Duesenberry, Goldsmith, and McPherson (1999). An important conclusion of the
study derived from wide-ranging comparative analysis and several African
case studies is that reform cannot be sustained if African countries
continue to overload their development agendas. The degree to which the
spread of HIV/AIDS has undercut "state capacity" simply reinforces that
point.
52 Taken together these two countries comprise more than half the GDP of
SSA. They have been excluded because they have received negligible amounts
of foreign aid.
53 Elsewhere, McPherson argues that one of the reasons African countries
have declined so dramatically is because of the large, agenda-overwhelming
flows of foreign aid. This view is not widely accepted especially among
members of the foreign assistance community where foreign aid has been seen
as having many positive effects if only African countries would adopt the
appropriate policies (World Bank 1998). My argument is that many African
countries have been discouraged from adopting appropriate policies because
aid has been so large. The issues are debated in McPherson and Gray (2000).
54 Johnson 1964.
55 The minimal fallback is United States (or some other large country)
treasury bills and bonds. Investors willing to take more risk can purchase
stock index funds, bond funds, or some other highly liquid instrument.
56 An accessible source is Hirschliefer (1970). See also Mankiw (1997,
Ch.17).
57 Mankiw 1997, pp. 420-425.
58 Although the point is not emphasized here, the spread of HIV/AIDS
reduces the willingness of surplus holders (whether individuals or
organizations) to lend. The reduction in expected working lives reduces the
prospects that borrowed resources will be repaid. In this way, the spread of
HIV/AIDS undermines established (or, in developing countries, emerging)
credit markets.
59 The two studies on Zambia, cited earlier, suggest that at the present
time there are already 500,000 AIDS-related orphans in Zambia out of a
population of around 10 million. By the year 2014, that total is expected to
increase to 1 million out of a projected population of 12 million (loc.
cit., p. 35).
60 To repeat a point made elsewhere, one of the major challenges that has
not been adequately addressed is dealing with the effects of HIV/AIDS has
been to devise novel approaches to training. Adapting new technology in ways
that raises productivity as the supply of skilled workers declines is one
such challenge.
61 The modest reversal in the growth of per capita real incomes,
emphasized by Madavo and Sarbib (1997), resulted from improvements in
weather and some of the spillover effect of economic liberalization. Those
gains, however, have not been sustained (World Economic Outlook Spring 2000,
Statistical Appendix, Table 5; World Development Indicators CD-ROM 2000).
62 Schultz 1957, Ch.13.
63 Makiw 1997, pp. 134-135.
64 Whether workers pay for the training through lower initial wages (as
Becker and others have argued GET Economist reference), is not crucial.
According to efficiency wage theory, the wage adjusted for training costs
will be higher than earnings in comparable activities.
65 In the absence of universal HIV testing, employers face a major
'signaling' problem. The outcome, well known since Akelof's work on "lemons"
(i.e., defective products), is that employees will have an incentive to pay
at rates reflecting average productivity. This will over-value workers with
HIV/AIDS and undervalue those without. The outcome will be a reduction in
morale and efficiency.
66 Hill and McPherson 1999.
67 Denison 1962; Leveson 1967.
68 Collier 1995; World Bank 1995; Yeats et al. 1996, 1997 .
69 Rodrik 1998; McPherson and Rakovski 1999 .
70 CNN international ran a report on AIDS in Kenya on 29th April 29,
2000. They reported that although the rate of infection nation wide was 15
percent, the incidence of HIV among the adult population in the fishing
village of Koba Bay approached 70 percent. The death rate in that community
has doubled in the last year.
71 The additional burdens falling on grand parents, particularly
grandmothers, were noted early in the epidemic (Beer 1988, pp171-174). Those
burdens have become especially heavy in agriculture.
72 Hoover and McPherson 1999, 2000. During a visit to Lusaka in January
2000, McPherson discussed the problem of HIV/AIDS with senior ministry of
finance officials. None wished to be quoted but they agreed that the
epidemic had seriously disrupted the work of the ministry of finance, the
central bank, and many other parts of the government with which the ministry
of finance had to deal.
73 Arrow 1974, p.33.
74 Arrow 1974, p.33 This point has earlier echoes in the work of Hayek
(1945) and Arrow (1952). Hayek pointed out that for the price mechanism to
function efficiently there has to be a pooling of information among a wide
range of participants. Arrow asked that in a competitive market where
everyone takes prices as given, who makes a decision on price? Both
observations are fundamental to the operation of modern economies. The
spread of HIV/AIDS has been so disrupt of efficiency and growth precisely
because it has undermined the basis of established collective action.
75 Arrow 1974, p.59.
76 We have used the same approach adopted in McPherson and Rakovski
(1999).
77 McPherson and Rakovski (1998, 1999a, b) draw on the model developed by
Khan and Knight (1991) to estimate small simultaneous systems for Kenya,
Zambia, and Ethiopia. McPherson and Rakovski (2000) have also estimated a
simultaneous equations model for 39 countries over the period 1970 to 1998.
HIV/AIDS related variables could be inserted in this system to measure the
impact of the epidemic on growth.
78 This conclusion is evident in Hill and McPherson (1999).
79 This point is developed at length in the forthcoming volume by
Duesenberry, Goldsmith, Gray and McPherson entitled Restarting and
Sustaining Growth and Development in Africa. This volume contains a
collection of studies (some of which have been published over the last two
years) that have been completed under the USAID-funded Equity and Growth
through Economic Research (EAGER) project.
80 My colleague, Professor Dwight Perkins, has been making this point for
many years. Asian countries he has noted (Perkins 1994) did not attempt to
leapfrog the development process by cluttering up their development agendas.
So far, that message has been missed. It will continue to be missed while
the World Bank continues to hustle its "comprehensive development framework"
and the IMF persists with its "poverty reduction and growth" strategies.
Both seriously misinterpret the capacity of African governments to implement
the programs they adopt despite their formal commitments to do so.
81 A critical lesson that African countries (except Botswana) have not
yet learned from the Asian experience is the over-riding importance of a
dynamic, expanding agricultural sector. Although the point has been made in
many different ways over long periods (World Bank 1981, 1990; Lele 1981;
Eicher and Baker 1982; Mellor 1998), African governments have persistently
failed to grasp its implications. Those are that there can be no long-term
sustained development in Africa while food security remains an issue.
82 Two recent examples are the advanced copy of the World Development
Report 2000/01 that was released on the World Bank's web site and the latest
report (number six since 1981) on Africa. This former is a collection of
impressions and observations lacking a coherent core. Ostensibly aimed at
the reducing poverty, the narrative skirts around the principal requirement,
namely that if poor countries want to become rich they have to promote and
sustain economic growth whatever the temptations and inducements may be to
focus on other objectives. The latter study asks whether "Africa can inherit
the 21st century?" (World Bank 2000). The uncertain tone is a long way from
that of the first study in the series, the so-called Berg Report entitled
"Accelerated Development in Sub-Saharan Africa: An Agenda for Action.".
83 The main problem with the IMF in this regard is that its staff has no
fundamental commitment to reducing poverty. They are bankers whose strength
is that they focus about financial accounts and the prudent use of money and
credit.
84 Available on www.pub.whitehouse.gov/uri-res/Ö./1993/1/21/1.text.1.
85 The Economist 13th May, 2000.
86 Quick Print, Botswana, for example, keeps no medical records.
87 AIDSCAP, p.29.
88 AIDSCAP, p.14.
89 Carroll (2000) quoted Kofi Annan, Secretary-General of the United
Nations, that there is a need to "smash the wall of silence and stigma
surrounding" HIV/AIDS. The Boston Globe (October 1999) devoted one of its
four sections on HIV/AIDS in Africa to the problems created by "official
silence" about the disease.
90 James Hall (World) reports the outrage expressed by Swazi workers at
Fridge Master, a large manufacturing concern, when management publicly
announced company statistics on AIDS and HIV infections. He also reports
that the "Swaziland Federation of Labour said that AIDS talk would deter
foreign direct investment." The threat to investment is also raised by Ng
(Harvard Aids Review).
91 "Private Sector AIDS Policy: African Workplace Profiles", Washington,
D.C., Family Health International, 1995., pp. 8-12.
92 ILO, chapter.5.1.
93 National Business Initiative Annual Report, 1998/99, p.1.
94 See the 1997/98 and 1998/99 Annual Reports. A similar situation is
found with Zimbabwe's Partners for Growth, a network of chief executives of
the country's leading companies, dedicated to promoting economic growth and
prosperity. The organization's director has noted that the topic of HIV/AIDS
has never been raised.
95 Essex , 1999.
96 For example, Essex noted:
[In Botswana]Ö and in Zimbabwe, 35 percent to 40 percent of all pregnant
women are infected now; in some towns and villages, 50 percent are infected;
and in young adults, aged 20 to 30 years, 45 percent to 50 percent are
infected nationwide.
97 This characterization of companies' responses to HIV/AIDS is found in
Simon, etc., p.7.
98 Collins, p.10, and AIDSCAP, p.20.
99 Collins, p.13.
100 Collins, p.10.
101 Collins, p.8.
102 UNAIDS, p.3.
103 Collins, p.8, 10, 18.
104 AIDSCAP, pp. 71 and 74.
105 Rugalema, p.48.
106 World Bank (1999, p 40.) raises the issue of powerlessness of women
which hinders them from taking precautions (such as insisting their husbands
use condoms) to prevent infection.
107 Dezign Incorporated (Deadly Dezigns), Zimbabwe.
108 The booklet is entitled"AIDS: Toward a Greater Understanding"..
109 Examples include: the Botswana Meat Commission (AIDSCAP, p.16-17);
Rio Tinto, Zimbabwe (AIDSCAP, p.73); and Eastern Highlands Plantations,
Zimbabwe (AIDSCAP, p.69).
110 UNAIDS/WHO, p.18.
111 Collins, p.17.
112 Harmony Gold Mining offers prophylactic treatment of sexually
transmitted diseases to commercial sex workers in the mining community.
(Michael, 1999, p.5).
113 Sanachem, South Africa.
114 Loewenson, et al., p.27.
115 Bollinger, p.11.
116 Collins page 9 made special note of Triangle Ltd., and the Commercial
Farmers' Union in Zimbabwe.
117 UNAIDS/WHO, p.19.
118 Rugalema, et al, p.44.
119 ILO, p.3.4.2.
120 Michael, Karen. "Unbelievable: AIDS Reporting in the Business Press,"
p.7.
121 ILO, p. 3.4.2.
122 ILO, p.3.4.2.
123 Collins, p.7, and Rugalema, et al, p.43.
124 ILO, p. 3.2.1.
125 AIDSCAP, p.21.
126 Rugalema, et al, p.43.
127 Collins p.12.
128 Collins, p.13.
129 Simon, p.8.
130 Collins, p.12.
131 Collins, p.12.
132 Rugalema, et al, p.44.
133 One of the reasons given for making this change was to support the
formation of a black entrepreneurial class. However, it is clearly no
coincidence that truck drivers are one of the highest-risk groups for AIDS.
134 UNAIDS, p.18, and Rugalema, et al., p. 44.
135 ILO, p.3.4.6, and Collins, p.7.
136 Collins, p.13.
137 Collins, p.13.
138 Sunday Times, (SA).
139 Delta Corporation, Zimbabwe (Loewensen, p.69), and Sunday Times,
(SA).
140 Whitelaw, "AIDS in the Classroom.".
141 UNAIDS/WHO, p.18.
142 For example, at Delta Corporation, First National Bank of Botswana
Ltd., and Kgalagadi Breweries (Pty) Ltd., Loewensen, pp.70, 75, 79.
143 See Loewensen, pp. 65-80.
144 For example, Kgalagadi Breweries (Pty) Ltd., Loewensen, p.79.
145 Barclays Bank in Zambia has a special, generous termination package
that is negotiated with employees who decide they can no longer contribute
to the work of the firm.
146 UNAIDS/WHO, p.19.
147 See examples in Loewensen, pp.51-62, 105-6.
148 Collins, p.14.
149 Botswana Meat Commission, AIDSCAP, p.17.
150 The SADC Summit in 1997 endorsed a Code on HIV/AIDS and Employment,
to guide employers, employees and governments on the most effective,
economically sustainable and humane ways to respond to HIV/AIDS in the
workplace. Many SADC countries also have national codes or guidelines on
HIV/AIDS and employment.
151 Smart, p.5.
152 Although this is an important point for companies that employ
numerous illiterate workers, it has only been mentioned in the case of
Africa Beverage Co., Southern Africa (AIDSCAP, p.50).
153 World Bank Policy Research Report, pp.77-85.
154 Simon, p.8, and ILO, 3.4.2.
155 Hoover and McPherson 2000.
156 This is true even in the literature. Note especially the two World
Bank references and the Kenyan study by Forsythe and Rau, published by
Family Health International, AIDSCAP, and USAID.
157 Lindauer 1995.
158 Lindauer 1981.
159 Lindauer 1995.
160 ILO 1998.
161 UNAIDS 1999.
162 Bennell 1996.
163 Ram and Schultz 1979.
164 Rosen, Vincent et al. 2000.
165 See McPherson and Rakovski (1999). That model focused on the key
elements in the financial programming framework used by the International
Monetary Fund. It was based on the model presented in Khan and Knight
(1982).
166 World Development Report 1993, Ch. 1.; Gillis et al. 1996, Ch.11.
167 Indeed, Zambia was one of the few countries in the world where infant
mortality and under-5 mortality had increased during the 1980s and 1990s
(World Development Indicators 1998, Table 2.17, p.106).
168 McPherson and Rakovski 1999.
169 Gillis et al. 1996: Chapter 20; Ghura and Hadjimichael 1996;
Calamitsis, Basu and Ghura 1999; McPherson and Rakovski 2000.
170 Rodrik 1998; McPherson and Rakovski 1999a, 2000.
171 Missing from our analysis is any test for reverse causality. This
would require an equation for life expectancy. This aspect can be examined
in subsequent research.
172 In Zambia, life expectancy at birth peaked at 51.2 years in 1982. It
declined to 49.6 years in 1987. By 1992, it had fallen to 48.4 years. By
1998, it was 43 years.
