ASSESSING ENVIRONMENTAL-ECONOMIC INTERACTIONS IN PUERTO RICO
This study of economic-environmental interactions focuses on quantifying two relations: land use change-food production capacity and manufacturing sector growth-toxic waste release. The manufacturing sector is an economic activity that extends the natural carrying capacity of the island. In this chapter, I investigate the relative impact of that sector as a employment generator and toxic waste releaser. The potential impact of growth in this sector is described through a projection of the two variables, employment by the manufacturing sector and toxic chemical releases.
In the second part of the chapter, I estimate changes in population supporting natural capacity based on food production. For these estimates, I depend in land use change projections presented in Chapter Four, and population growth estimates (Chapter Three).
This chapter tries to integrate the effect of production activities of the manufacturing sector on employment, energy use and toxic waste released. One objective is to determine if some of these factors might be limited to the ability of the manufacturing sector to extending carrying capacity on the island. This is done through the evaluation of the manufacturing sector impact on the sustainable character of the economy, based on the evaluation of its labor absorption potential and its potential environmental impact. Since job creation has been one of the major objectives of development policies and planning, this is considered a major social benefit of growth in the manufacturing sector while toxic waste released is considered a major social impact.
The relations discussed in this analysis can be identified in the system diagrams presented in Chapter Three (Figures 3.1, 3.2, and 3.3). I describe three relations among components of the system: land use change as it effects biological carrying capacity, manufacturing sector growth and employment, and manufacturing sector growth and change in toxic waste release. In this section I describe the methodology used in these projections. A more detailed description is offered in the individual sections that discuss the projection results.
The analysis describes relations and identifies patterns of change. I use ratios to project changes in the variables chosen as the system's carrying capacity indicators and limits and introduce variations as options or alternatives. The indicators are: employment, food production capacity, toxic waste released.
I project toxic waste release and job creation potential in the manufacturing sector assuming there is a continuation of the ratios between these and GNP. The model assumes that there is a direct relation employment and GNP, and calculates toxic chemical releases using ratios of releases per person employed. The model sets as an objective the satisfaction of full employment during the period of the projection and uses rates of growth in GNP representative of past patterns of growth. The rationale behind this part of the model is that development in the manufacturing sector is a major strategy of the Economic Development Administration for job creation. I assume that employment in this sector will increase linearly with GNP as well as toxic waste releases. It does not take into consideration toxic chemical releases reductions due to technological improvements, more stringent regulations, or environmental awareness. Although these are important variables to consider when describing future release, the analysis purports to evaluate the relation between the manufacturing sector composition, employment potential and toxic waste release as a way of improving our understanding of the potential impact and means of balancing these and the social benefits that can be described from this sector activities.
In the second part, I project changes in the biological carrying capacity or food production capacity of the island. For these projections I assume that land use change occurs in the rates described previously based on historical data. The calculations use three crops and combinations of them, rice, dry beans, and sweet potatoes. These are tropical crops that supply significant percents of the caloric intake in many of the tropical countries and also popular foods on the island. Although I do not describe fossil fuel derived energy inputs it is assumed that these crops require high input levels; production per hectare rates used are based on the highest estimated production rates for Puerto Rico.
Although the analyses was developed independently, these are all components of the overall system diagramed in Figure 5.3. The individual analyses are intended to describe relations for specific variables that form part of that system. However, its significance is clearer when interpreted as parts of a whole system.
EFFECTS ON TOXIC WASTE RELEASES AND EMPLOYMENT OF GROWTH IN THE MANUFACTURING SECTOR
The objective of this projection is to predict what would be the levels of employment and toxic waste release assuming the manufacturing sector maintains a similar composition than in the 1980s, as well as employment, and toxic release rates. The projection is made as a function of growth in GDP in manufacturing. I assume two annual growth rates.
THE RELATIONSHIP EMPLOYMENT-MANUFACTURING GROWTH
A main objective of development programs in Puerto Rico has been to reduce unemployment through industrialization of the manufacturing sector. Yet, despite significant growth in that sector unemployment continues to be a chronic problem. At the same time, high rates of toxic chemical release have a potential impact that could reduce the natural productive capacity of the system.
This analysis looks at the question of how much growth
in the manufacturing sector would be necessary to absorb the labor surplus,
and what would be the impact of that growth as indexed by toxic waste released.
Estimates of toxic waste generation are calculated from the projection of growth in the work force. I estimated the volume of toxic waste that would be released by production in the manufacturing sector assuming that it maintains a similar composition, projecting growth, labor demand and applying the ratio toxic waste per employee.
To estimate labor demand by the manufacturing sector, I extended the population model used to project growth (Chapter Three). This model uses population data to project growth of the working age population (16 or older), and labor force that would be available at three participation rates (Appendix IV). Participation rates represent the highest, average, and lowest rates of participation observed during the last twenty years.
The model calculates labor demand by the manufacturing, agricultural, and "other sectors" using past trends of employment, and ratios to production growth. It uses labor force projected for three participation rates based on population growth forecasting. "Other sectors" refers to sectors other than manufacturing, or agriculture, including service, commerce, and the public sectors. I projected estimates to year 2015 using 1980 as base year and regression equations of empirical data.
I calculate labor demand using a regression of persons employ by sector and GDP as an aggregated indicator of production. The regression uses data for years 1970-1988. Labor demand using regression of employment and GDP was not made for agriculture because there is an inverse relation between these variables. This inverse relation can be explained as a function of other factors such as increasing energy inputs and capital investments, and change in economic structure.
I projected labor demand as a function of two GDP growth rates (in 1954 dollars) for each of the sectors represented. For the manufacturing sector these are: 6.99 percent, the average annual growth rate for the 13 year period between 1975 and 1988, and 4.45 percent, the average annual growth rate for the last 9 years (1980-88). For the activities grouped under other sectors, the GDP annual growth rates are: 3.03 percent (average for 1975-1988), and 2.15 percent (average for 1980-88).
The projected labor demand of the manufacturing and other sectors, current employment trends, and two GDP growth rates, are used to calculate the work force absorption potential by these. They are calculated as a percent of total labor force projected for three participation scenarios.
The carrying capacity idea can be applied with a model such as this by imposing limits that, if chosen as policy options, would constrain change based on social objectives. It assumes that the objectives established and represented as limits, would enhance the long term sustainable potential of the system and recognizes a conflicting relation between these goals and other social objectives.
This analysis is structured based on current relations between the productive sectors and labor. It helps to improve our understanding of the system and options rather than providing factual information about expected trends.
A summary of the projections of employment and toxic
waste release based on employment appear in Table 5.1 (Program code and
output in Appendix IV). Projections to year 2015 of labor demands show
that if current trends of decreasing employment in the agricultural sector
continue, labor demand by agriculture would be reduced to 2.39 percent
of the total work force. This is a reduction of 1.97 percent of work force,
between 1980 and 2015. Employment by manufacturing, if current trends continue,
would increase by approximately 1 percent for the same period (Figure 5.1).
Activities grouped under the other sectors category, the labor demand category
of fastest growth, would increase its demand in 18.4 percent of the total
working force (Figure 5.2). These estimates are based on projections that
assume the lowest participation rate based on population growth estimates.
TABLE 5.1 PROJECTIONS OF EMPLOYMENT AND TOXIC WASTE RELEASE FOR THE MANUFACTURING SECTOR BASED ON ASSUMPTIONS OF GROWTH
Sources: EPA 1991; PRPB 1988b; Appendix IV
a as percent of total employment
b in pounds
FIGURE 5.1 PROJECTION OF PERCENT OF WORK FORCE ABSORBED BY MANUFACTURING
Source: Appendix IV
FIGURE 5.2 PROJECTION OF PERCENT OF WORK FORCE ABSORBED BY "OTHER SECTORS"
Source: Appendix IV
FIGURE 5.3 PROJECTION OF TOTAL LABOR DEMAND BY MANUFACTURING AND OTHER SECTORS UNDER DIFFERENT ASSUMPTIONS
Source: Appendix IV
Projections of labor demand caused by an increase in GDP at the higher growth rates (6.99 and 3.03 percent in manufacturing and "other sectors" respectively), show that at the lowest participation rate, the economy would reach full employment by 1996 (Figure 5.3). That would require an increase in employment opportunities in the manufacturing sector equivalent to 3.7 percent of its total employment in 1980 (i.e. 57,646 positions). At the highest rate of growth under which estimates were calculated for the other sectors category, these would have to provide 238,413 new employment opportunities by year 1996. For these estimates I assumed the lowest participation rates of the ones projected based on population growth estimates.
FIGURE 5.4 PROJECTION OF MANUFACTURING SECTOR LABOR DEMAND AS A PERCENT OF WORK FORCE ASSUMING TWO RATES OF GDP GROWTH
Source: Appendix IV
FIGURE 5.5 PROJECTION OF OTHER SECTORS LABOR DEMAND AS A PERCENT OF WORK FORCE USING TWO RATES OF GDP GROWTH
Source: Appendix IV
The range of employment at different rates of growth between the two used is illustrated by the area between the two curves for the low and high rates of growth for each category, manufacturing (Figure 5.4), and other sectors (Figure 5.5).
At the lower rates of GDP growth used second scenario for labor demand projection based on GDP growth, i.e. 4.45 and 2.15 percent for manufacturing and other sectors, full employment by the two sectors is reached by year 2006 (Figure 5.5). That would require an increase of 62,237 employment opportunities in the manufacturing sector and 320,329 for the other sectors.
Employment opportunities in the manufacturing sector increased, from 1975 to 1988 with several ups and downs, in 37,000 employees (28 percent increase). These years can be described as a period of intensive growth in the export oriented manufacturing sector. During that period, GDP in this sector increased 331 percent (136 percent in 1954 dollars).
Manufacturing And The Environmental Limits
A question posed by these projections is; given the current structure and composition of PR's manufacturing sector, what would be the environmental impact of growth in that growth in manufacturing activity?
I estimated toxic chemical releases by the manufacturing
activities to be used as an indicator of impact. These estimates use ratios
of toxic waste released per employee calculated with empirical data for
year 1987 (EPA 1988). I assume that a similar manufacturing sector composition
and structure holds during the projection period, the conditions under
which the labor demand estimates were projected. Estimates are calculated
in two ways: using average toxic waste/employment ratio for the manufacturing
sector as a whole, and using ratios and employment percentages for individual
major manufacturing activities (Table 5.2). In both cases I used 1987 base
TABLE 5.2 TOXIC WASTE RELEASE ESTIMATES FOR THE MANUFACTURING SECTOR BASED ON EMPLOYMENT PROJECTION TO YEAR 2015
Source: Appendix IV.
a Growth in employment calculated from the regression equation of employment on the manufacturing sector and time.
b Percent of work force that would be employed by manufacturing. Assumes low participation rates.
c Estimates based in projections of total employment by manufacturing
d Estimates assume same employment distribution for the manufacturing sector, and toxic waste/employment ratios for major manufacturing sectors (SIC 20, 23, 28, 29, 30, 35, 36, 38, and an aggregate of others) as those of 1988.
In the estimates of employment based on current trends there is an 8.7 percent increase in manufacturing sector employment by 2000. Increase in employment by 2015 would be 21.3 percent. Increase in percentage of toxic waste release, calculated using individual ratios for manufacturing activities, is 34.9 percent by 2000, and 50.4 percent by the year 2015 ("Based on Sector Employment" column on Table 5.2).
In the 6.99 percent GDP growth scenario, employment by manufacturing grows by the year 2000, to 50.3 percent of its 1988 level, and 184.1 percent by the year 2015. Here, toxic waste released by the sector increases to 86.4 percent and 252.4 percent respectively.
Under the more realistic figure of 4.45 percent GDP
annual growth, manufacturing employment increases by the year 2000 by 21.2
percent from the 1988 employment, and 64.6 percent by 2015. Under that
assumption, the estimate of toxic chemical release is 50.3 percent higher
than the 1988 figures by the year 2000 and 104.2 percent higher by 2015
FIGURE 5.6 TOXIC WASTE RELEASE ESTIMATES AND PROJECTIONS, 1940-2015
Sources: Appendix IV, EPA (1988, 1989, 1990d), PRPB (1988b)
Projection of toxic chemical releases by manufacturing
illustrates the high rates of release that would occur under the conditions
described if the economy depends on this sector's growth as a strategy
to decrease unemployment and increasing investments on the island. This
could be significantly reduced by a strong commitment to protective policies.
The manufacturing sector has achieved in the last
years very high annual growth rates of up to 6.99 percent. Continued growth
at these rates would increase employment opportunities. If the island economy
achieves high rates of growth in both manufacturing and the "other" sectors,
full employment could be reached by the end of the 1990s (Figure 5.7).
This assumes no increase in labor productivity.
FIGURE 5.7 HISTORICAL DATA AND PROJECTION OF EMPLOYMENT UNDER ASSUMPTIONS OF HIGH AND LOW GROWTH RATES AND PROJECTION BASED ON CURRENT GROWTH PATTERNS
Source: Appendix IV
These rapid growth rates in the manufacturing sector has allowed the system to extend its carrying capacity from the natural limits imposed by its resources in comparison to the high population density. The inflow of capital in the form of investments in the manufacturing sectors and the resultant increase in economic productivity have provided the means for this extension.
However, the relation established between employment and toxic chemical releases by the manufacturing sector shows that if it grows at rates similar to past growth patterns, and maintains the same release per employee ratios, there would be a significant increase in environmental impact. An important fact to remember is that most of the releases reported, i.e. 30 percent of total, occurred in the fugitive emissions category. Although, in this study the release of toxic chemicals by the manufacturing sector has not been related to a specific impact, these projections serve to draw a picture of the potential environmental impact of continued growth in an export-oriented manufacturing with a similar composition that it had in the 1980s.
The comparison of present trends and projections with historical estimates of toxic chemical release shows an alarming trend in terms of the potential impact of these releases on resources (Figure 5.6). This analysis points to the need of more stringent implementation of policies and other mechanisms to reduce release of toxic chemicals into the environment.
This analysis shows that if the current structure and composition of the economy continues, growth of the manufacturing sector would increase toxic waste releases at a faster rate than the employment opportunities. If PR's government policies promote continuous growth in the more polluting sectors identified above, it must develop the mechanisms to protect vital resources and ecosystems from the releases of waste.
The impact of sector composition is illustrated clearly by the difference between the toxic chemical release calculated for the manufacturing sector or based in individual (each sector's) activities. There is an increase of about 25 percent the total toxic chemical releases calculated when individual rates per sector activity are used.
Efforts in industrialization policy and planning should be directed to increasing growth in the identified less polluting manufacturing activities, agriculture and other sectors of the economy. Emphasis should be placed on economic activities that can satisfy local demand for goods to substitute imports.
This assessment has been developed using empirical
data from the PR experience. However, the scenario might be transposed
to other developing nations with similar manufacturing sector composition.
The recently developed Caribbean Basin Initiative Programs are designed
based on incentives similar to those on which the Puerto Rico development
programs are based. These countries must create the mechanisms to protect
their resources from the impact of export oriented manufacturing activities
in the more polluting sectors.
EFFECT OF CHANGES IN LAND USE IN PUERTO RICO'S POTENTIAL POPULATION SUPPORTING CAPACITY
Self sufficiency in meeting of basic needs such as food is one of the major indicators of sustainability in development. In this section I estimate potential population supporting capacity according to food production based on land use changes and population growth. The objective of this projection is to calculate the effect of the change on past land uses on the food production potential. Food production potential can be used as an indicator of the system's ability to assimilate natural energies although it is not the only source of primary product.
I project the change in farmland area using available data (SCS 1982, USDC 1987). I estimated change in food production capacity assuming the average production rates reported for three combinations of crops. This estimates, together with my population growth estimates, were used in the projection of the capacity of remaining farmland to satisfy the caloric demand by the population.
I chose three crops based on popularly perceived preferences in food consumption on the island: rice, dry beans, and sweet potatoes. Production for these three crops was calculated using farmland estimates based on four assumptions. In the first land use change estimates, I assume there is no change (no reduction) in farmland after 1987. The second scenario assumes there is an annual reduction in farmland equal to the average annual rate in farmland from 1950-1987, i.e. 9,699 ha. per year. The third farmland change rates assumes an annual reduction of 6,119 ha., the average annual farmland reduction between 1982 and 1987. The fourth assumption is an annual reduction of 3,360 ha. per year, the lowest annual rate of change observed between 1950 and 1987. The second, third and fourth scenario rates of change are the same than those use in Chapter Four calculations of land use change. Farmland area for base year 1980 is 443,204 ha (Soil Conservation Service, 1982). The Department of Natural Resources Land Use Inventory of 1972 reported 489,545 ha of agricultural land (54.29 percent of total land).
I calculated production rates for a combination of rice and sweet potatoes, rice and dry beans, and rice and a combination of the last two, based on production per hectare reported in literature (Guadalupe 1980, Llorens et al. 1978, Moscoso 1955, Chandler et al. 1977, Colegio de Ciencias Agrícolas 1984, Espinet et al 1977). Rice was maintained as one of the crops in all projections because of its predominant position in the Puerto Rican diet. However, I assume that only 30,261 ha of land are planted in rice following recommendations of the rice growing manual for the island, Cultivo intensivo y perspectivas del arroz (Chandler et al. 1977). I estimated the production for rice on that area leaving the difference for one or a combination of the other two crops. I calculated total kilocalories produced by crop and combinations based on nutritional information for each food group. Total kilocalories is then used to estimate the number of people whose caloric requirements could be met. This is used here as an estimated indicator of the food production based natural carrying capacity of the island. Results are presented as tons of food, kilocalories and the percent of total population projected for the corresponding year.
Table 5.3 summarizes the results of the population supporting capacity estimates based on farmland change projections. These projections show that not only the reduction in farmland area as an important determinant factor in the food (caloric) production capacity, but also in the combination of crops chosen. In this study, however, crops were chosen for their caloric value,1 and their cultural significance rather than for nutritional reasons. Therefore, these results should be interpreted as an indication of production capacity more than as a analysis of the absolute potential to satisfy nutritional needs.
The "no change" section on Table 5.3 presents the change in carrying capacity due to population growth assuming farmland remains the same as in 1987. That projection assumes that, with adequate inputs, average production levels can be maintained over the time projected. The major variables in this scenario are the crops and population growth.
People supporting capacity based on caloric content is drastically reduced when crops are rice and dry beans as opposed to sweet potatoes.
TABLE 5.3 SUMMARY OF POTENTIAL FOR SUPPORTING PEOPLE ESTIMATES IN PERCENT OF TOTAL PROJECTED POPULATION FOR SELECTED YEARS
a Crops represent combinations of the following: rice (R), sweet potatoes
(SP), and dry beans (DB).
Source: Appendix V
This projection reflects the reduction in food production capacity by reductions in farmland area and population growth. The crop combination chosen prove to be more significant than the reduction caused by land use change. The difference is due to the difference in production rates by crop. The sharp disproportionate reduction in the last period of the first scenario is caused partially by the limitation of hectares left to rice growing. Sweet potatoes produces around 23,919 pounds per hectare while dry beans produces an average 3,683 pounds per hectare.
When the "no change" in land use scenario is compared
with the farmland area reduction scenarios we observe that in the case
of sweet potatoes, around 15 percent of the reduction in people supporting
capacity is due to population growth. Therefore, in the other scenarios
we can assume that the difference in carrying capacity is due to the reduction
in farmland. The reduction in carrying capacity due to food production
potential equals 53, 33, and 18 percent of the total population for the
three scenarios 1, 2, and 3 (Figure 5.8).
FIGURE 5.8 PEOPLE SUPPORTING CAPACITY AS A FUNCTION OF FARMLAND AREA: FOUR SCENARIOS
Differences in production by crop were more significant in determining the ability of existing farmland to influence the people supporting capacity than land use and population growth. Except for the case where the higher rates of change in land use, differences between sweet potatoes and dry beans in production were higher than those due to the reduction in farmland available. This points to the significance of crop type in food production self sufficiency for the island. However, this should not be overemphasized in an open economy exposed to external and international markets where preferences are influenced by a variety of different factors, and since there is a real reduction caused by changes in farmland area.
Decrease in potential population supporting capacity caused by land use change proved to be more significant than the reduction due to projected population increase. On the other hand there is a reduction in carrying capacity for the more productive crop combination of from 33 to 68 percent depending on farmland change rate. Therefore, in the lower rate of farmland area reduction, change in farmland is at least as significant a factor in reducing people supporting capacity as is population growth. In all the other cases farmland reduction is a larger factor in effecting reductions in carrying capacity than population growth.
The carrying capacity rates are only indicative of the people supporting capacity based on food production potential on the island. In this study farmland is not necessarily equal to total agricultural land. Agricultural land might be more extensive than farmland if we consider that land under other uses such as forested land could be transformed to farmland. However, the 1972 land use inventory by the Department of Natural Resources reports 489,545 ha or 54.3%, as agricultural land. It also describes 292,930 or 32.5 percent of total land as forest. There is a modest difference between that which is reported in 1972 as agricultural land and what is reported for 1980 as farmland by the SCS (1982), 443,204 ha. (90 percent of the Department of Natural Resources' 1972 figure).
In this projections I assumed uniform land productivity rates in farmland. As explained in Chapter Four, productivity varies significantly throughout the island. The analysis also assumed high energy inputs, and maximum productivity, therefore projections should be interpreted as maximum potential based on caloric production ability. Since data for farmland (and defined by photo interpretation2 ) does not necessarily include all land with agricultural potential. We might assume that productive agricultural land might have a greater extent than that indicated by land use statistics which, in part, would counterbalance drops in productivity.
The interpretation of these results must consider patterns of agricultural land change in the last 10-20 years. There has been significant substitution of the traditional extensive crop, sugarcane, by other horticultural crops such as fruits and vegetables. Sugarcane occupied most of the flat agricultural coastal lands. It is on that land where most of the new extensive residential development and other permanent development occurred. Therefore, most of the land use change on the coastal plains is to permanent development and cannot be returned to agricultural uses. Most of the change of agricultural land to other uses, non-developed, has been to forest. This shift has occurred mostly on central mountainous agricultural land, where productivity might be lower.
Estimates based on farmland availability and population show that until the mid-1970s farmland availability was sufficient to satisfy the region's population caloric demand. During the mid-1970s farmland area decreased and the population grew at rates that caused a reduction, during this decade, of about 30 percent of the population. Food production capacity estimates here assume high energy inputs in the production process.
A policy to substitute imports with local production probably represents a higher cost than imported food because of high labor costs. High land costs due to intensive urbanization and high population density demands policies directed to the control of the degradation and degradation of land with agricultural potential.
The Census of Agriculture (USDC 1987) 1987 figure for farmland was 358,893.84 ha. This figure lies between the 1990 farmland projection for the highest and the medium reduction rates scenarios. Based on this projection that represents a supporting capacity of about 60 percent of the population projected for that year (Figure 5.8). This reduction represents a lower people caloric demand satisfaction ability than the projection to 2015 if farmland reduction had been stopped at 1980 levels, that is 62 percent (a reduction of about 7 percent due to projected population growth).
These results show that future natural capacity to satisfy food needs and reduce toxic chemical release into the environment on the island depend on policies designed to achieve this as an objective of long term development. Protecting agricultural productivity, policies must emphasize reducing the transformation of agricultural land to other uses.
The projections presented show that land use change in Puerto Rico is a more significant factor in reducing natural food production capacity than is population growth. Actual agricultural production is at significantly lower levels than its maximum capacity, for farmland under agricultural production is estimated at 14 percent of its potential. Therefore, to describe the economic and social significance of a shift in emphasis of public policies toward the development of a greater participation of agriculture, additional analysis must be made. However, we can conclude that based on observed patterns of labor demand and value added in agriculture, greater participation of this activity in the economy would have a positive effect in terms of employment and a lower impact on ecosystems that current major economic activities.
I consider both the change in farmland area and the release of toxic chemical substances major factors that determine the environmental sustainability of the Puerto Rican system. Changes in these variables are difficult to relate quantitatively; however the pattern of reduction in farmland area can be related chronologically to growth in the manufacturing sector due to emphasis in industrialization strategies by the Economic Development Administration, and substitution of local production by imports.
1 This follows other studies where potential population supporting capacity is calculated like Higgins et al. (1982).
2 DNR 1972 land use
inventory was based on photo interpretation.