CHAPTER THREE



MICAH M. CHEATHAM



PRIVATE SECTOR SOLUTIONS TO FERTILITY REDUCTION IN INDIA







The Demographic Transition

The demographic transition is the very core of family planning theory, 

tracing the transition of a given society across time, from very high 

birth and death rates to very low ones, such that the post-transition gap 

between births and deaths is less than or equal to its level prior to the 

transition.  Conceptually, the demographic transition begins in an 

environment in which households, in need of familial labor and heirs, 

must compensate for high levels of mortality by producing large numbers 

of children; as the provision of public health and sanitation 

technologies increases, mortality should greatly decrease, cutting death 

rates far below rates of birth.  At this point in the transition, high 

birth rates are no longer being offset by high rates of death, and 

population will boom; the perceived value of having children will 

eventually change, and individual reproductive behaviors will adapt, 

lowering birth rates until the difference between birth and death rates 

is similar to, if not lower than, that which existed prior to the transition.

        In the transition experienced by most industrialized nations, 

mortality fell slowly and haltingly, relying on innovation and invention 

to incrementally cut death rates; eventually, fertility followed this 

trend and, after two hundred years of transition (Demeny, 1989), Europe 

and the rest of the industrialized world eventually reached stable birth 

and death rates and a steady fertility rate at near replacement level.  

The experience of the developing world, however, has been quite 

different; not until after World War II, and the end of colonialism, did 

this region begin its transition.  Medical and public health technologies 

first discovered during the European transition had since been perfected, 

and were grafted into the transition of the developing world, cutting 

death rates to levels enjoyed by the industrialized world in a quarter of 

the time (U.S. Census Bureau, 1991).  Unfortunately, despite the infusion 

of large amounts of Western monetary and technological aid, birth rates 

have resisted most attempts at forced reduction, leaving a large gap 

between current birth and death rates.



        As a result, population in much of the developing world today is 

growing almost without bound.  As the sheer numbers of people within a 

given region reach and surpass levels that can be sustained by existing 

political, economic, or ecological systems, family planning becomes an 

issue of increasing importance.  Unfortunately, many fertility-reducing 

programs, both past and present, have generally been of limited 

effectiveness; although a number of programs have been able to cut 

fertility rates to some extent, most of the developing world continues to 

face frighteningly high growth rates and 

above-replacement-level-fertility.  One major limitation of past programs 

was a concentration on supply-side marketing; family planners assumed 

that there existed a latent demand for contraceptive technology and that 

fertility could be reduced by providing a supply of this technology.  

This is not to say that these programs have been failures; many have 

quite effectively reduced fertility, but in each case, birth rates have 

reached some lower threshold, beyond which they simply cannot pass.



Building a Theory on Fertility Reduction        

        Family planning practices that seek only to meet latent demand 

will always reach a limit to fertility reduction which they cannot 

breach; this lower limit is, in fact, indicative of a saturation of the 

natural (latent) demand market.  Clearly, any family planning program 

that seeks to overcome limits to fertility reduction must be based not on 

meeting an existing, limited demand for contraceptive technology, but on 

the expansion of demand markets.  Supply-side theories fail in family 

planning because they disregard the externalities that exist within a 

given households decision to reproduce:  The social costs of high 

population, and therefore high fertility, are extremely high; yet, so 

long as large families are desirable, or profitable, to the individual 

household, the (individual) opportunity costs of high fertility will 

remain significantly below the (aggregate) social costs, and households 

will continue to produce large numbers of children.  Thus, effective 

family planning programs can exist only where the household opportunity 

costs of high fertility are more reflective of its social costs.  

Significant reductions in fertility come about only where reduced 

fertility is perceived as beneficial to the individual household; in 

other words, family planning practices can only truly be accepted where 

they are seen as a desirable expansion of choice.

        The increasing prevalence of fertility reduction theories of the 

type outlined above amongst demographers and family planners (Coale; 

Knodel, 1984; Cleland, 1987) has generated a growing body of 

policy-related variables that effect fertility rates; one variable with a 

pronounced effect on fertility is womens education.  As women become 

increasingly educated, they begin to take on more of the skills required 

for wage-labor; the wages associated with female employment, because they 

would be lost to women rearing large numbers of children, provide a 

disincentive to high fertility by increasing the opportunity costs 

associated with large families.  The scatter-point diagram on the 

following page seems to support this hypothesis, showing a clear negative 

correlation between fertility rates (1990) and the percentage of females 

aged twenty-five or older who have completed primary school (1989) over a 

wide range of countries. 

        Although the conceptual link between womens education and 

fertility makes intuitive sense and seems to be supported by empirical 

evidence, it reveals only one part of a more complex relationship.  

Womens education, in this case, acts as a proxy for the somewhat vague 

concept of the value of women; specifically, the average level of 

education attained by females within a particular society should act as a 

flag, indicating the degree to which women participate in that society, 

including the value that is placed on their labor.  The level of 

acceptance of female education within a society should be correlated with 

that societys acceptance of female employment in the formal sector; in 

other words, womens ability to develop their own human capital, in the 

form of education, must be linked to the market demand for female human 

capital.  Under these conditions, female education acts as a proximate 

measure of a given societys norms regarding the full participation of 

women in formal sector employment; this is part of what is meant by the 

value of women.  The link between market demand for female labor and 

fertility provides the same conclusion as the link between female 

education and fertility, only in a more direct way:  The possibility of 

losing the potential to earn a wage increases the opportunity costs 

associated with large families, and thus has a dampening effect on fertility.



Fertility Reduction in India

        The theorization laid out above is heavily biased towards solving 

the population problems of the developing world.  This concentration is 

characteristic of most theories in family planning and is by no means 

accidental; the population problems experienced in this part of the world 

are more severe than those affecting the industrialized nations, as are 

the consequences of ignoring them.  In so saying, the key to 

understanding the difficulties of this region is in remembering that 

these problems are not unique to the nations of the developing world so 

much as they are characteristic of the stage of the demographic 

transition which most of these nations have reached.

        The nation of India is suffering the effects of being trapped in 

the median stages of the demographic transition perhaps more than any 

other country in the developing world.  India currently enjoys a death 

rate of 10.07 deaths per 1,000 population (CIA World Factbook, 1995), 

lower than that of 

both Europe and Russia (U.S. Census Bureau, 1991); yet, its birth rate 

remains unnecessarily high, at 27.78 births per 1,000 population (CIA, 

1995).  In a nation of almost 950 million people (CIA, 1995), a birth 

rate-death rate gap of over 17 births per 1,000 population produces a 

population increase of over 16 million in excess of replacement levels.  

A combination of strong political support on the national level for 

family planning programs and extensive Western aid has reduced Indias 

fertility rate to 3.4 births per woman (CIA, 1995), an enviably low level 

by developing world standards; unfortunately, in a nation as large as 

India, the population pressure of even this low a fertility rate is 

incredible.  The age structure of the Indian population adds an air of 

urgency to efforts to reduce fertility; thirty-five percent of the 

population is under fifteen years of age, and fertility rates must be 

reduced before this cohort reaches reproductive age.

        Unfortunately, there is little evidence to suggest that fertility 

rates in India will fall in the near future.  With national family 

planning programs dating back to 1951, no other government in the world 

has placed as much of an emphasis on reducing family size than Indias; 

however, these programs have generally failed to see the results that 

they could have.  Government family planning programs have ranged from 

offering incentives to couples for contraceptive use to forced male 

sterilization, yet none of these has brought about significant reductions 

in fertility, largely because the low social status of women has been 

consistently ignored.  Indian households bear a strong preference for 

male children; the son in the traditional Indian household not only 

serves as heir and continuation of the family line, but also provides 

labor for the family and, according to Hindu tradition, must perform the 

parents sacred burial rites.  A daughter, on the other hand, represents a 

net economic liability because of the financial pressures that her dowry 

places on her parents.  The dichotomous relationship between the sexes 

creates a situation in which women are denied the property rights and 

control over economic resources that would increase their social value, 

relative to men.



A Free-Market Solution

        In a society where womens roles are marginalized, the value of 

their labor tends to be underestimated.  Where this is the case, the 

private sector has the ability to capitalize on this undervaluation by 

hiring women at a lower wage than that which is offered to men, thus 

capturing increased returns to production; the difference between actual 

female wages and the potential wage bill, had males been employed, 

generates a fiscal surplus that can be invested into capital expansion of 

the industry.  Now, if it is true that increasing the employment 

opportunities available to women has a significant negative effect on 

fertility, this implies that private industry has the opportunity to 

affect positive social change in the course of normal profit-seeking 

operations.  The situation at hand is beneficial to both society and 

industry alike.

        Where a firm hires women to capture the lower wages associated 

with female labor, that firm can experience lower costs of production 

without reducing the market price of its goods, in this way increasing 

its returns to labor.  As other firms recognize the scale effects of 

increasing their female work force, they will hire increasingly more 

females, boosting the rate of growth of the female labor force.  The 

female labor force will grow increasingly until the environment of 

heightened demand begins to nudge the female wage upward; at this point, 

the growth rate of the female labor force will begin to decline.  The 

growth rate of female participation in formal labor will incrementally 

decline as the female average wage increases; eventually, the female wage 

will rise to a point where employers are indifferent between the sexes, 

and the female participation rate will stabilize.

        The increasing-then-decreasing rates of growth story is typical 

of models of innovations that lead to lowered costs of production; if the 

increased use of womens labor can be characterized as an  innovation, the 

growth of the female labor force can be mapped out in another way, one 

which is complimentary to that which is outlined above.  When a firm in a 

competitive market captures a lower wage bill due to the employment of an 

increased number of women, that firm will lower the asking price for its 

particular good, in an effort to expand its share of the market for its 

product.  Once several firms in an industry expand their female labor 

force, they will be in direct competition with each other to capture the 

greatest savings to production that this innovation can offer.  As more 

firms increase their female work force, the practice of hiring women will 

become more commonplace, and the original innovators will not see 

significantly greater returns to labor over their direct competitors.  At 

this point, these firms will begin to look to other sources of economic 

growth, and the growth rate of the female labor force will begin to 

decrease.  Once female-to-male employment ratios are equalized across the 

industry, the individual firm will have nothing more to gain from 

employing more women, and the industrys female labor force will stabilize.

        Systems that experience growth rates that at first increase and 

then decrease to a steady state equilibrium of no growth can be 

replicated with the following generalized equation:



                 dY/dt = kY * (q - Y)/q (partial derivative)



Here, Y is the percentage of the national labor force that is female, q 

is an exogenously-generated upper boundary to growth, t is time, and k is 

the constant of proportionality.  The variable k is referred to as the 

constant of proportionality because the derivative of Y at any point can 

be explained, at least partially, as some proportion k of the original 

function Y.  That the partial derivative of Y is equal to some form of 

itself implies that this is an exponential function:  de^t/dt = e^t * dt, 

and above, dY = kY * dt * (q - Y)/q.  The second part of this equation, 

(q - Y)/q, is a forcing term, introduced to limit the growth of the 

exponential in this equation; this equation is attempting to predict the 

growth of a variable that is measured in percentages, and so we must 

necessarily impose an upper limit of one hundred percent, if not some 

even lower value.  The forcing term (q - Y)/q dampens growth by 

interacting with Y; where Y is a small number, the term is very close to 

one, but as Y increases, the forcing term approaches zero.  In short, 

this equation states that the female percentage of the labor force would 

grow exponentially over time (kY), if it were not for the existence of 

some exogenous limit to growth ({q - Y}/q); instead, it grows 

exponentially only up to a point, after which growth slows until the 

function stabilizes at a fixed value.



We can see that the equation laid out above is just another 

interpretation of the Verhulst equation:  

                                     

                       Y'/Y = (k/q) * (q - Y),                       -- 1



where Y'is another form of notation for the first derivative of Y with 

respect to t.



We can approach a general form solution of the equation by integrating 

equation 1.  We begin by multiplying through by Y



                  dY/dt = Y * [k - {(k/q) * Y}],                     -- 2



and then separating variables



                  dY/(Y * [k - {(k/q) * Y}] = dt.                    -- 3



In order to simplify the integration process, we want to separate the 

left hand side of equation 3 into the form:



           1/(Y * [k - {(k/q) * Y}] = A/Y + B/{k - (k/q) * Y},



where A and B are as yet unknown.  To do this, we multiply both sides of 

the above equation by Y * [k - {(k/q) * Y}], to get:



               (A * k) - [A * (k/q) * Y] + (B * Y) = 1.



Now, where Y = 0, A = (1/k), and where Y = 1, [(1/k) * k] - [(1/k) * 

(k/q)] + B = 1, which gives us the solution B = (1/q).  Thus,



       1/(Y * [k - {(k/q) * Y}] = (1/k)/Y + (1/q)/{k - (k/q) * Y}.



(This can be checked by multiplying the equation through by Y * [k - 

{(k/q) * Y}], which produces 1 - (Y/q) + (Y/q) = 1.)



We now have an equation of the form



              [(1/k)/Y + (1/q)/{k - (k/q)Y}] * dY = dt.         -- 4



Integrating with respect to t gives us



         (1/k) * ln Y - (1/k) * ln[k - (k/q) * Y] = t + C,      -- 5



where C is the constant of integration.



Multiplying both sides by k and consolidating the natural logs produces 

the following solution:



           ln[Y/(k - {(k/q) * Y})] = k * (t + C),               -- 6



which simplifies to...



   Y = [k * e^(k * {t + C})]/[1 + {(k/q) * e^(k * {t + C})}]    -- 7



This is the generalized form of the solution to the Verhulst equation 

above; it makes explicit the relationship between time and growth for 

this model.  

        Equation number seven is extremely useful in describing the 

growth of the female labor force in India, with some caveats; the most 

important of these considerations is that time is used as a proxy for 

attitudes towards womens labor.  The hypothesis, as stated, is highly 

time-dependent; womens labor is used increasingly over time as more of 

the individual players in the labor market recognize its value.  The use 

of time as a proxy for this phenomenon may be slightly misleading, but 

the malleability of time as a variable makes it a satisfactory measure of 

what is actually a rather nebulous concept.  Time is a valid proximate of 

any incremental change because it is measured in ordinal numbers; that 

is, each number in a time series measures only its location within a 

sequence, and does not have any particular value attached to it.  Thus, 

time will continue to be used in further transformations of the above 

equation, with the implicit assumption that it is a proxy for the 

changing intensity of use of womens labor.

        The generalized solution form makes explicit the relationship 

between Y and t, q, and the two constants k and C; however, in order to 

generate solutions for specific data, k and C must be replaced with more 

definite values.  Conventionally, the constant of integration, C, is used 

only in indefinite integration; where at least one explicit relationship 

between the variables is known, this is definite integration, and C is 

assumed to be equal to zero.  If we use the most recent value for female 

labor force participation available from the World Resources Database and 

assume a specific upper limit to growth, we have values for Y and t, and 

q, respectively, and can assume C out of the equation, generating a new 

generalized form:



           Y = [k * e^(k * t)]/[1 + {(k/q) * e^(k * t)}].         -- 8



Unfortunately, the argument of definite integration is not, by itself, 

the most robust justification for assuming away C; the transition from 

equations 4 to 5 is actually one of indefinite integration.  This is 

because our explicit relationship between variables uses observed data 

from 1990; however, there is no reason to believe that any behavioral 

change occurred in 1990 to institute the growth that this equation 

describes.  Our 1990 data does not constitute any initial conditions, but 

instead acts as a baseline for speculation as to the general 

applicability of this model.  

        Nineteen-ninety data is used as a baseline value only because it 

is the most recent available; this by no means implies that the process 

described above commenced at this date.  We perform indefinite 

integration between equations 4 and 5 because the data we are using is 

clearly not the start of an historical example of the above process; were 

it actual historic data, it would be referred to as initial conditions, 

and definite integration could be performed.  Instead, we use baseline 

values, plugging them into the variables in question to generate 

hypothetical scenarios as a means of testing the robustness of our model; 

we use indefinite integration to preserve the generalizability of the 

solution equation, allowing us to test the results of a variety of 

baseline assumptions.  We are now placed in a position where we have 

specific data, yet we have performed indefinite integration and, as a 

result, we have the constant C in our equation.  Yet, precisely because 

we do have specific data, we can still assume away C; we can do this 

because, in the following analysis, we treat each set of baseline values 

as if it were historical data, and we know that C is not included in the 

definite integration of historical values.

        Baseline values play their most important role in the 

determination of definite values for the constant of proportionality, k.  

The World Resources Database provides the following baseline data:  

Females accounted for 48.318% of the population, yet only 25.191% of the 

labor force of India in 1990.  Thus, Y = 0.25191 and, because 1990 is the 

baseline year in our iteration, time t = 0.  If we assume that, once a 

demand increase has been generated, the percentage of women in the labor 

force will expand up to but not beyond their proportional representation 

in society, then we have set our upper limit to growth at q = 0.48318.  

We now have the tools necessary to generate an estimate of k; we begin 

with the general form solution,



          Y = [k * e^(k * t)]/[1 + {(k/q) * e^(k * t)}]              -- 8



and include our baseline values, to generate the following:



      0.25191 = [k * e^(k * 0)]/[1 + {(k/0.48318) * e^(k * 0)}].     -- 9



Replacing e^(k * 0) with 1 and multiplying the right hand side of the 

equation by (0.48318/0.48318) produces...



            0.25191 = (0.48318 * k)/(0.48318 + k),



which provides us with a k-value of approximately 0.52630.  Thus, the 

general form of our solution for this set of assumptions is:



  Y = [0.5263 * e^(0.5263 * t)]/[1 + {(0.5263/q) * e^(0.5263 * t)}]. -- 10



        The equation now provides an explicit relationship between the 

independent variable t, the policy variable q, and the dependent variable 

Y, subject to the baseline assumption that Y(0) = 0.25191, q = 0.48318.  

The following graph details the convergence of this equation on the upper 

boundary of Y = 0.48318.  The accompanying table clarifies the growth 

pattern of the function, highlighting the fact that it infinitely 

approaches 0.48318 without actually reaching that value.  





  Year    1990    1991    1992    1993    1994    1995    1996    1997    1998    1999
%       25.10   31.19   36.41   40.40   43.20   45.04   46.20   46.92   47.35   47.61
L.F.       4       3       0       2       0       3       8       5       9       9
2000    2001    2002    2003    2004    2005    2006    2007    2008    2009    2010
47.77   47.86   47.92   47.95   47.97   47.98   47.99   47.99   47.99   47.99   47.99
   4       6       1       3       2       4       0       4       7       8       9





        The upper limit to this equation was arbitrarily set at 0.48318,  the percentage of the total population of India that was female in 1990;  however, this percentage has been growing steadily, although marginally,  over the past twenty years.  This implies that our equation may be more  reflective of the growth pattern it attempts to describe if the upper  limit q of our equation were to be included as an increasing function of  t instead of as an arbitrarily determined constant.  In fact, the  regression of this data over the years 1970 to 1990 produces the function  Y = (0.000026 * t) + 0.4823, which grows at so slow a rate that it can be  legitimately approximated by any constant function at or about Y = 0.4823.         Although it seems to be appropriate to estimate the upper  boundary to growth with a constant function based on current estimates of  female representation in the general population, the thinking behind this  conclusion may be ignoring some rather serious considerations.  India has  a long cultural history of son-preference between children, which may  bias contraceptive and child-rearing techniques; although these practices  have become less prevalent in modern India, the proportion of females in  Indian society in 1990 was still only 48.23%, less than the world rate of  49.68% and significantly below the 51.54% faced by the industrialized  world (WRD, 1994).  It may not be unrealistic to assume that some  residual son-preference continues to exist within Indian culture, and  that this is maintaining the proportion of females in society at an  artificially low level; this sort of consideration is linked to the above  model because the prevalence of son-preference amongst households is  directly related to the value of women within a society.  Now, where an  increase in womens employment is brought about in a climate of  son-preference, it seems safe to assume that the increased potential for  future wage labor that this places upon a female child may compensate for  any cultural predilection towards sons; in this way, the growth of the  percentage of females in the labor force may increase the percentage of  females in the general population, thus expanding the upper limit q used  in the predictive model.         Any activity that reduces artificial constraints placed upon the  size of the female population within a society is inarguably producing a  public good; however, any points regarding the effects of this  demographic change on the model may be moot.  As the graph shows, our  predictive function converges with its upper limit in only ten iterations  (iterations here are denoted as years because this is the unit of measure  that seems to make sense, although other units of time could be used); it  seems unlikely that behavioral patterns that are as deeply ingrained as  son-preference could be changed in as short a time as ten years.   Further, the delayed effects of changes in reproductive behaviors on the  labor pool ensure that any increase in the female population will occur  long after the demand boom has passed, such that female employment rates  are unlikely to be effected by an increase in the female population at  this stage.           A more fundamental criticism of the model suggests that there is  no reason to believe that the growth of the female labor force will be at  all limited by the proportion of females in the general population; it is  more likely that the female labor force will overshoot the proposed limit  of percentage female representation in society.  If an individual firm,  in the course of expanding its labor force, recognizes lower costs to  production by employing women instead of men, there is little to prevent  that firm from further reducing costs by replacing as many currently  employed men as possible with women.  The only hindrance to this sort of  mass replacement (if we can momentarily assume away contradictory  cultural norms) would be a limited supply of sufficiently-educated women,  and this constraint should weaken significantly once individual  households begin to recognize the newly increased value of their daughters.         It is difficult to determine the extent of the predicted excess  growth of the female labor force; it is easy to believe that employers  themselves will generally be male and will not be replacing themselves,  and so the female labor participation rate could never reach one hundred  percent.  In fact, most upper- and middle-management positions would be  closed to women for this same reason; the most physically strenuous jobs,  too, would almost certainly be denied to women.  This implies that some  significant percentage of employment positions within the economy exist  which are simply unavailable to women; that is, the demand for female  labor associated with this percentage is inelastically fixed at zero.  In  1990, male representation in the work force exceeded male representation  in society by approximately 23%; the fact that men, in a male-dominated  society, are over-represented in the labor force by twenty-three percent  seems to make this figure an appropriate estimate of the extreme-case  upper limit to growth of the womens labor force.  This assumption implies  a new upper boundary value of q = (0.48318 + 0.23) = 0.71318, which  generates a generalized solution form of   Y = [0.38948 * e^(0.38948 * t)]/[1 + {(0.38948/q) * e^(0.38948 * t)}],  -- 11 the graph of which follows.
Year    1970    1975    1980    1985    1990    1995    2000    2005    2010    2015
%       29.65   28.46   27.18   26.20   25.19   56.54   68.75   70.94   71.26   71.31
L.F.       0       0       1       6       1       8       6        1      4       0
2020    2025    2030
                                                         

71.31   71.31   71.31
   7       8       8                                                        





Again, a table is provided for ease of reference.         Growth of female representation in the labor force is staggering  here, and probably not particularly realistic.  Lack of availability of  data requires that time be expressed in units of five years, which may  exaggerate the graphical representation of  the function; however, this  does not negate the fact that the percentage of females in the labor  force more than doubles between 1990 and 1995.  In order to make this  data at all useful, the ordinal nature of time must be taken advantage  of; if the difference between two years can be seen only as a difference  of a few places in a generalized time series, then the model can begin to  take on more meaning.  Where 1990 is viewed as simply the one thousand,  nine hundred, ninetieth iteration of the series, then the span of a unit  of time increases in its variability, and can take on values larger than  a single year.  Although this assumption increases the applicability of  the model, it does so at the expense of its reliability; more  importantly, it fails to address more fundamental ethical questions  associated with female employment overshooting female representation in  society. Policy Implications         Sustainable economic growth occurs where new jobs are created;  yet within this model, the primary engine for growth of female  representation in the labor force is female replacement of males in  existing jobs.  Where females simply replace males at a lower wage, the  individual firms operating costs fall, but social costs may actually  rise; even if each displaced male worker were replaced by a female in the  same household, the overall average household income will fall.  Only  where male replacement frees up capital for expansion of the labor force  can replacement dampen the social costs it incurs; unfortunately, this  case increases the number of households earning a wage income while  reducing the wage earned.  The true social benefits of replacement growth  come from the reduction in birth rates that this generates; where female  employment rises, fertility should fall, and, under any conditions,  reductions in fertility will cut overall costs to society.         It is quite likely that the population effects of female  replacement may more than compensate for the effects of male  unemployment.  In 1990, the labor force growth rate only marginally  exceeded the rate of population growth, while a mere five percent of  working-age Indians were employed in the formal sector (26 million person  labor force, 503 million persons between the ages of 15 and 65, WRD); in  a slow-growth, low employment economy, the sort of restructuring of the  labor market that female replacement requires may incur only a minor  social cost.  If this is the case, the benefits of fertility reduction  should easily exceed the costs of male unemployment, leading to clear and  direct benefits to society, at least in the short run.  Long run  forecasting is more difficult because the long-term effects of male  unemployment at the expense of females is unclear.  The direct effects of  increased male unemployment in an environment of extremely high  unemployment may, indeed, be minor, as discussed above; however, the  effective gender transition of the households primary wage-earner, from  male to female, may produce significant disruptive forces in the  pervading culture.           It should be clear that any increase in the value of women,  especially one as significant as this, is acting in opposition to current  trends and, therefore, must necessarily act as either the cause or the  product of a significant cultural disruption, the effects of which could  never be predicted in their entirety.  While this may seem to be a rather  melodramatic conclusion to make, the uncertainty involved in the decision  to consciously change a culture must be taken into full account; the  risks involved in such endeavors must be clearly outweighed by the  benefits.  In so saying, it does appear as if Indias population growth  may warrant such strong actions:  the negative effects on society,  culture, and the environment associated with large population growth  rates cumulatively represent the greatest problem India has had to face  to date.  Sacrificing an indeterminate, and possibly quite large, amount  of culture for the preservation of society as a whole is hardly an ideal  solution; it is not a first- or second-best solution, but an nth-best  solution, the optimal solution subject to a large number, n, of constraints.         One such constraint that has been excluded from discussion to  this point is the fundamental lack of employment opportunities across  both sexes.  Where only five percent of the working-age population is  engaged in non-agricultural employment, it is conceivable that the sheer  numbers of unemployed men would bid the male wage down to such a point  that a significantly lower female wage would be infinitesimal; if the  female wage were as low as this suggests, it would actually provide a  disincentive for a given woman to leave the home, as her earned wages  could not compensate for the domestic work she would be sacrificing in  order to enter into formal wage employment.  However, in a country that  is 70% rural (WRD, 1995), yet concentrates a majority of its educational  opportunities in urban areas, a majority of the 95% unemployed are  actually agricultural laborers; any residual un- or under-employment may  be composed mostly of rural immigrants to industrial centers who lack the  education, and therefore the job skills, required for increasingly  technical industrial jobs.  This implies what we already know to be true  about all labor markets: there is no universal wage; instead, wages are  graded across positions, based on the skills required to perform a  particular job and the responsibilities entailed in it.  This model is  meant to consider female entrance into higher-skill, and therefore higher  wage, employment, where a womans wage can be less than that of a man  without being menial; these are the types of job that have an implicit  value attached to them and, by association, can raise the perceived value  of women within Indian society.         The policy implications of this sort of theorization are heavily  dependent upon the amount and the type of economic growth that is  stimulated by female replacement.  Where the implications to the firm of  female replacement are extra-normal profits, the government should be  concerned with how and where those profits are invested; the government  has the capability to lower the interest rate through the central bank  and to provide tax incentives for capital expansion, both of which are  powerful investment-stimulating tools.  More importantly, the government  is the only player in the market that has the power to confront  traditional biases against women in employment; it must take the leading  role in the stimulation of demand for womens labor that should eventually  generate the process described in the model.  The government has the  ability to make female labor more attractive to firms by increasing  female education opportunities, providing logistical and management  training and support to firms transitioning towards a greater female/male  ratio amongst their work forces, and continuing its long-standing  tradition of encouraging female participation in all levels of government.         It seems clear that this program could not succeed if the  government were to focus on either the education of women or the  generation of demand for their labor alone; both must be given equal  attention, and a direct link between the two must be made at every stage  of the process.  Although a number of approaches to this problem exist,  the simplest and most direct may be the best option in the earliest  stages; one example would be to increase the degree to which funding of  public schools is based on female representation in classes and to  provide tax credits to individual firms based on the income taxes  collected from their female employees.  These types of policies are  suggested not only because they reduce the costs to the central  government that are generally imposed by recruitment campaigns, but more  importantly, because they assign recruitment responsibilities in a more  efficient manner; it seems likely that local public school administrators  and teachers and the owners and operators of local industry would be more  familiar and respected in their communities, and therefore better able to  affect social change there, than unknown bureaucrats from economically,  if not geographically, distant New Delhi.  Under its New Economic Plan,  the Indian central government has very successfully turned over many of  its powers to more efficient private-sector groups; by concentrating more  normative taxes on firms, the government can continue in this policy by  privatizing away governmental responsibilities where it can be done at  less cost or with increased efficiency in the private sector.         The one policy implication that should not be garnered from this  material is that a laissez-faire, free market approach to the economy  will provide solutions to social problems.  This paper outlines a partial  solution to Indias population problems that works in the private sector  and probably only on a regional level, but only with governmental  support.  Without significant government involvement, cultural  predilections will continue to overrule profit-seeking tendencies, and  the role of women in the labor force will always be inappropriately  discounted.  Without continued government funding of its own and other,  private family planning programs, the supply of contraceptive technology  could not meet the increase in demand that would come about from the  fertility-reducing effects of female replacement.  Most importantly, it  is the role of government of the largest democracy in the world to  regulate both the use of womens labor in its economy and the reduction of  fertility in its society to ensure that its citizens are protected and  that its policies create positive and lasting social change and economic  development. REFERENCES Chiang, Alpha C.  Fundamental Methods of Mathematical Economics.  London:  McGraw-Hill Book Co., 1984. Cleland, John and Chris Wilson.  Demand Theories of the Fertility  Transition: An Iconoclastic View.  Population Studies 41, 1987.  (Publisher unknown) Coale, Ansley J.  The History of Human Population.  Scientific American.  (Date and publisher unknown) Concept Paper:  Commercial Contraceptive Marketing.  USAID/New Delhi:  1995. (Unpublished) Conly, Shanti R. and Sharon L. Camp.  Indias Family Planning Challenge:   From Rhetoric to Action.  Country Study Series #2, 1992.  Washington,  D.C.: Population Action International. Demeny, Paul.  World Population Growth and Prospects. Working Papers, No.  4, 1989.  Washington, D.C.: The Population Council. Knodel, John, et. al.  Fertility Transition in Thailand: A Qualitative  Analysis.  Population and Development Review 10, 1984. (Publisher unknown) Lewis, Maureen A. and Genevieve Kenney.  The Private Sector and Family  Planning in Developing Countries.  The Urban Institute International  Activities, V. 3686-01, (July) 1988.  Washington, D.C.: The Urban Institute. Mason, Karen Oppenheim.  The Impact of Womens Social Position on  Fertility in Developing Countries.  Sociological Forum, V. 2, # 4. (Date  and publisher unknown) Simon, Carl P. and Lawrence Blume.  Mathematics for Economists.  New  York: W.W. Norton & Co., 1994. Watkins, Susan.  Conclusion.  in The Decline of Fertility in Western  Europe, ed. by Susan Cotts Watkins and Ansley J. Coale.  New Jersey:  Princeton University Press, 1986. World Population Profile: 1991.  Washington, D.C.: U.S. Census Bureau. World Resources Database, 1994.  Baltimore, Maryland, USA:  World  Resources Institute. 1995 CIA World Factbook.  Washington, D.C.:US Central Intelligence Agency