South Florida is one of North America's last frontiers. Large-scale settlement began only in the first decades of the 20th century with the extension of the Florida railroads to Miami and the drainage of interior swampland for farms. Only a series of violent hurricanes in the late 1920s followed by the Great Depression could halt the frenzy of land speculation. The boom resumed in the 1950s, when the improvements in infrastructure brought by the US Army Corps of Engineers' massive Central & South Florida (C&SF) Project provided better flood protection for the cities and superior drainage for new farmlands.

From 1970 to 1990, the regional population nearly doubled, while employment and real income have grown much faster (see Table 2, lines 1-3). Although growth rates have slowed in the most recent decade and are expected to fall yet further, the population should reach 6 million by the year 2000 and almost 10 million by 2045. The population boom is reflected in the remarkably high number of multifamily housing starts built during the 1970s and the subsequent rise in the number of single-family units in later decades³ (lines 4.a-b). The state's forecast of a strong upward trend in both single and multi-family housing will have great implications for continuing land pressures in these counties.

The age distribution of the residents (lines 5a-c) indicates an inverting of the age pyramid by 2045. The share of the young (19 years and less), which was 29% of the population in 1970, is expected to fall to 20% by 2045. The oldest cohort (over 65 years), which accounted for 17.5% of the total in 1970, is expected to rise to 28%. Even these trends understate the importance of the older population due to the exclusion of seasonal residents ('snow birds') from the permanent population.

A summary of the distribution of non-farm employment (line 6) highlights the declining share of the combined construction-manufacturing-utilities sectors from 30.6% of the labour force in 1970 to 15.111/o by 2045, with the manufacturing sector declining the most. However, it is the rise of service employment (lines 6d-g), from 69.4% in 1970 to 80.5% in 1990 and to nearly 85% in 2045, that is determining the general character of the region.

In summary, South Florida's strong economy compounds its environmental problems: an expanding agricultural and commercial society serving a large tourist, retiree, and transient population. All this surrounds North America's only semitropical and largest remaining wetland, the Florida Everglades.

The conflict between the economic growth of South Florida and the Everglades is occurring on four fronts (see the right-hand map in Figure 1). In the northern and central stretches of the ecosystem, the growth of Orlando's suburbs, the cattle lands of the Kissimmee River Valley, the diking of Lake Okeechobee, and the drainage of the Everglades Agricultural Area (EAA) have all resulted in changing the quality and quantity of the downstream flow to the remaining Everglades. To the southwest of Lake Okeechobee, new areas are still being opened for cattle, citrus, and vegetable production. To the east, only the 100 mile-long protective levee restrains the expansion of the Miami-Ft. Lauderdale metropolitan areas into the Everglades, and three large water conservation areas, actually vast holding tanks, compartmentalize what once was the 'River of Grass'. To the south, Florida Bay and the Keys receive the nutrient-rich runoff from canals that drain the nearby eastern counties. The cities thrive as the Everglades die, despite the $1.1 billion already spent on restoration projects.⁴

__________________

274   R. Weisskoff 

Table 1. 
Some economic characteristics of the Greater Everglades (South Florida) ecosystem.

______________________

Missing Pieces in Ecosystem Restoration   275 

Table 1. 
(Continued)

Notes for Table 1: 

   Methods and Sources to Table 1: All data refer to 1997 unless otherwise noted.
   n.a. = not applicable; n.c. = data not comparable; ** = in cols. 4-6, refers to share of farmland in the smaller unit to farmland in the larger unit of col. heading. 
   All county, state, and some national-level data are from Bureau of Economic & Business Research (1998b), cited below as 'fsa' (Florida Statistical Abstract), unless otherwise noted. County data were summed for the region and compared to state and national totals. Other national data are from US Census Bureau (1998) cited below as 'us'. (Statistical Abstract of the US).

Section A: Line 1: fsa, t.8.01 & 8.03; line 2; Bureau of Economic & Business Research (1999). FL ranks 4th after CA, NY, & TX; line 3: fsa, t. 5.09; line 4: fsa, t. 5.48; us t. 756. US poverty rate for 1993 from fsa 1995, t. 25.01; line 5, fsa, t. 11.05; line 6: fsa, t. 2.36; line 7: fsa, t. 8.39 (refers to all freshwater withdrawals, including irrigation); line 8: fsa, t.8.16; us t. 4.02; line 9: fsa, t. 10.40; us t. 1159; lines 10-11: fsa, t. 13.21 & 13.41. Includes out of state registrations; lines 12-14: Office of Tourist Research (1998), table, p i, gives nos. and regional destinations; table, p. 39 gives no. and length of stay of auto and air visitors; line 15: fsa, t. 19.53. Includes 'parks and areas.'; line 16: fsa, t. 19.52; us t. 424; lines 17-18; fsa, t. 21.07; us t. 496.

Section B: All data from US Department of Agriculture (1999). the region is summed from county data given in the state files. The size distribution of land (line 4) is constructed by multiplying the midpoints of all the detailed closed intervals of land sizes by the number of farms in each interval. The number of acres in the open-ended interval was computed as a residual. The shares of the numbers of farms and their corresponding area were then computed and aggregated to three intervals. 

The Kuznets ratio of inequality is computed by summing the absolute differences of the percentage of the number of farms in each interval and the percentage of the corresponding area occupied by those farms. The Kuznets ratio ranges from zero (extreme equality) to 2 (extreme inequality).
Line 8 includes land in sugar cane for both seed and sugar production. Line 9 refers to land in vegetables harvested for sale. 

Section C: Lines 1-2: Species on the national list is available for each Florida county from Envirotools, Inc (1998). These species were compared to those appearing on the lists for the state sand the nation, which are posted on the website of the US Fish & Wildlife Service, www.fws.gov/ , dated 30 April; line 3; fsa, t. 8.01; line 4: fsa, T. 8.74. data for selected cities are taken to represent the regions: col. 1 (Miami); col. 2 (Jacksonville); col 3 (Average of Los Angeles, Atlanta, Chicago, New York). The 13 full counties included in the Greater Everglades region are Broward, Collier, Dade, Glades, Hendry, Highlands, Lee, Martin, Monroe, Okeechobee, Osceola, Palm Beach, and St. Lucie. 

_____________________

276   R. Weisskoff 

278   R. Weisskoff 

(b) Impact Analysis for Planning (IMPILAN) was first released by the Forest Service of the US Department of Agriculture in 1979 and was privatized in 1993 as the Minnesota IMPLAN Group (MIG). IMPLAN's flexible software and county-level databases permit researchers to construct and configure their own regional models. IMPLAN regionalizes the national input-output table computed by the BEA, adds more detail to agriculture, applies regional purchase coefficients (RPCs), and constructs a complete social accounting matrix (SAM) for 528 sectors and 21 economic and demographic variables. Regional impact multipliers may then be computed on the basis of the SAM accounts (see Minnesota IMPLAN Group, 1997).⁹

(c) Regional Economic Models, Inc (REMI.) provides a dynamic modelling capability that allows the user to simulate scenarios involving a much larger number of economic and demographic variables. 'The historical coefficients are based on time series, the modelling horizon extends 35 years into the future, and the blocks of variables include output, population, labour, market shares, and prices. Once the dynamic framework of REMI is collapsed and their closure rules are made uniform, then all three of the above modelling tools-RIMS, IMPLAN and REMI-are somewhat comparable (see Rickman & Schwer, 1995).

The present study uses IMPLAN to investigate the economy of 13 South Florida counties.

The complete SAM for a region, together with its data availability, appears in Figure 2. The SAM balancing equation can be written¹⁰ as follows.

x= X+y                                                                                                                                                                                             (1)

where x is a [m + n + h] x I vector of total outputs of m industries, n factors and h institutions. X is a square [m+n + h] matrix of inter-industry transactions, value-added, factor disbursements, inter-institutional transfers, and household spending, as shown in Figure 2. y is a vector of final demand that includes exports but excludes household expenditures.

Let A be the square coefficient matrix of total direct input requirements defined as A=Xx, such that X = Ax. Then the balancing equation is:

x - Ax + y                                                                                                                                                                                           (2)

______________

Missing Pieces in Ecosystem Restoration   281

Figure 2 
Social accounting framework, full format. 

_______________

For the SF models, around 60'Yo of the total impact on both output and employment is direct (bottom bracketed line, Panel A). Only for rice and sugar milling and car rentals do the indirect employment effects overshadow the direct (as indicated by '<' between columns 5-6). This is in contrast to the US model where direct impacts account, on the average, for 34% of total output and 41.5"/o of the employment generated (columns 9 and 13). In the national model ' three sectors show greater indirect than direct output impacts and five sectors show greater indirect than direct employment impacts (columns 9-10, 13 14). These compare to the sparseness of indirect linkages in the regional model (compare columns 1-2 with 9-10; 5 6 with 13-14).

In the SF model, most of the induced multipliers for the sectors are higher than the indirect ones except for 6 sectors in the case of output and 7 sectors in the case of employment (columns 2-3, 6-7). These contrast with the national model in which rice milling is the lone sector for which induced is less than indirect output (columns I 0- I 1). In the case of employment on the national level, induced impacts for all sectors are greater than the indirect ones (columns 14 15).

In the regional model, the sectors generate between $13.9 and $19.0 million in total output (column 4), but new employment ranges from 90 jobs in rice milling to almost 900 jobs for agricultural services (column 8). In general, the agricultural and service sectors generate the most employment, and manufacturing the least. The profiles of car rentals and airlines are more similar to manufacturing than to services, and commercial fishing is more similar to services than to agriculture.

It is striking that some of the direct employment multipliers (vegetables, sugar crops, commercial fishing, sugar mills, and car rentals) for SF are higher than for the US model (columns 5 and 13). This may be due to climatic factors, such as a longer growing season, or to regional specialization in a more labour-intensive branch of production, such as sugar cane in SF, in contrast to sugar beets, which dominates the national model The indirect labour impacts for SF are all lower than the US model except for the nearly identical coefficients for vegetables and greenhouses (columns 6 and 1,I). This might reflect the fuller development of agricultural suppliers, relative to the sparse linkages in manufacturing and services. For the national model, the induced impacts account on the average for 42'Y() of' output and 39.5'Yo of employment generated by $10 million of additional demand for these sectors (columns 11 and 1'5), in contrast to the 25.0% of output and 22.3% of employment induced in the regional model (columns 3 and 7).

The above observation are important for several reasons. First, country planning department tend to use RIMS multipliers, which are specific for the state but not the region and exclude the induced effects completely.  thus, they fail to measure at least a quarter of the employment and income impact of new projects.

Second if planners use the current regional IMPLAN coefficients for forecasts

____________________

Missing Pieces in Ecosystem Restoration   289

(as we have in the next section), they would still be underestimating the ultimate impacts, as the region, through its growth, may capture a fuller set of linkages which are more characteristic of the national model.¹³ However, if national coefficients are 'borrowed' by a regional model to simulate the potential of soon-to-be-build industries, then the regional model overstates their impacts, since regional sectors rarely acquire the full linkages of the national sectors.

The impacts of $ 1 0 million of new demand on the 'composite scenarios' (Panel B of Table 3) are based on actual engineering data and on projects in progress that affect combinations of sectors (which are given in column 9). All the scenarios yield about the same output (column 4), except for the lower output generated by the Operation and Maintenance of the current Everglades Protection Project (EPP, line 2, Panel B), due to the higher proportion of direct wages in this programme. The tourist package (line 3, Panel B), which is based on several different statewide and regional visitor spending surveys, yields the highest employment of the four scenarios (column 8).

The consumption patterns of the three income classes (Panel C) suggest the contradiction between the slightly higher output generated by the $10 million spending by the lowest class (column 4) and the slightly higher employment generated by the same level of spending of the highest class (column 8). A growth pattern that results in higher incomes to the top classes, therefore, stimulates more jobs, while growth that results in greater spending of the lower classes generates greater output. In developing countries, this same effect, but to a more exaggerated degree, explains the short-term social stability associated with greater income inequality and the loss of jobs that results from populist attempts to redistribute income.

5. Results: Components of the Scenarios

In the following sections, we review the components and their combination into scenarios (see Table 4).

(a) Rotation of rice and sugarcane replaces the uninterrupted monoculture of sugarcane as currently practised, leads to higher sugar yields and reduced soil subsidence and pesticide use.¹⁴ In the fourth year, rice replaces sugar and provides a second or 'ratoon' cutting prior to the replanting of sugarcane in the fifth year.

(b) Organic rice could become the most profitable crop for the Everglades Agricultural Area (EAA) as soil subsidence lowers sugar yields, environmental standards regarding downstream runoff are tightened, and low priced foreign sugar is given a larger share of the US market.

(c) Water storage only in the HAA removes all agricultural activity and converts the sugarcane lands into reservoirs and marshes. The loss is calculated as the reduction of all current sugarcane production and milling.

(a) EPP: the ongoing Everglades Protection Projects, which consist of 14 projects, will be completed by 2005, with Operation and Maintenance (O&M) costs continuing into the future.

__________________

290   R. Welsskoff

Table 4. 
Summary of economic and employment impacts on So. Fla. scenarios

__________________________________

Missing Pieces in Ecosystem Restoration   291

 Table 4. 
(Continued)

__________________

292   R. Weisskoff

(b) Restudy, completed by the US Army Corps of Engineers, could be under construction by 2005 and end by 2025, with annual O&M costs continuing thereafter. The project includes 287 conventional water control structures (pumping stations, spillways, culvert, canals and levees) as well as several untried technologies, such as 345 aquifer storage-and-recovery wells (ASRS) and 47 miles of underground seepage barriers of 10 to 28 feet in depth. The impacts measured here result from construction expenditure only and exclude land acquisition costs.

(a) Auto and air tourism is based on continuing the four-year trend in expenditure of visitors who report SF as their primary destination.

(b) Everglades Region nature tourism assumes a doubling of the historical rate of increase in the number of visitors to the major national parks, wildlife preserves, and waterways in the region that was observed for 1995-96, but maintaining a fixed expenditure per visitor (US Army Corps of Engineers, 1998, Appendix E, pp. 209-213).¹⁵

The historical state-wide rate for all new construction activity was extended to the region for the entire period.

5.5. Consumption of New Residents

This reflects the expenditure of new retirees and in-migrants at twice their historic numbers in the region.

6. Results: The Scenarios

The scenarios are composed in order to sum up the effects of the missing pieces and compare them to the base projections of output and employment for the same years. The base projections, which appear in Table 2 above, exclude a 'tourists sector' per se. Rather, elements of tourism, such as spending on food, hotels, and amusements, are 'absorbed' into the spending patterns of local residents, making it virtually in endogenous sector. In our model, tourism is treated as an array of exogenous demands, despite the fact that these demands are influenced by the state's own heavy promotion in domestic and foreign markets. A restored Everglades, the flagship of the US environmentalist movement, will also add to the region as an important destination for eco-tourism and recreation.

A. 1. The 'Growth as Usual' scenario (Table 5, line A. 1) sums up the projects that are already underway and those trends that are likely to continue: the conclusion of the EPP, a continuation of the greater shares of the air but diminishing shares of auto visitors to the region, continuation of the overall construction trend, and high rates of in-migration.¹⁶ During the projected years, increases due to tourism and construction dominate the output and job impacts. The 'Growth as Usual' scenario results in an 0.86% increase in output and a 2.09% increase in employment over the total output and total employment, respectively, for 2045 (see Table 5,

__________________

Missing Pieces in Ecosystem Restoration   293

line B. 1, columns 5 and 1 0). But when compared to the annual increase in output and employment of the 2045 economy, then the 'missing pieces' output and employment are quite significant, i.e. 47.4% and 342.2% over the respective annual increases for the basic economy in 2045 (line C.1, columns 5 and 10). This highlights the degree to which the official projections understate the effects summed here. (See Figure 5 for a comparison of the annual employment changes in the base and the scenarios.)

A.2. The 'New Projects' scenario adds the rice rotation, the full Everglades Restoration, and additional regional nature tourism (lines la, 2b and 3b from Table 4). The first of these leads to a net loss of 2800 jobs per year throughout the period. The second will add 7000 jobs per year at the height of construction, tapering off to 3000 jobs per year for maintenance and operation. The increase in nature tourism would add another 3700 jobs by the end of the period. The sum of the elements in the 'New Projects' scenario will make relatively small contributions to the annual increases in output and jobs by 2045 (line C.2).

A.3. The 'Full Speed Ahead!' scenario sums the two scenarios above. By 2045, annual increases in output and employment will be 49.5% and 358.1%, respectively, above the forecast annual increase (line C.3).

A.4. The 'Naturalist Moderation' scenario would substitute rice for sugarcane, a more labour-intensive crop, and reduce the levels of the activities that make up Growth as Usual and New Projects. This scenario generates an annual increase in output and employment of 20% and 140%, respectively, above the annual increases for 2045 (line C.4).

A.5. Naturalist Moderation minus the EAA cuts out all sugarcane growing and milling in the EAA, further reducing the output and employment below the levels of Naturalist Moderation.

7. Conclusions and Caveats

(1) The annual increases in output and employment due to the inclusion of pieces hitherto 'missing' from the basic forecast are relatively small magnitudes compared with the large base. Even with higher growth rates of tourism, construction, immigration and a sugar/rice rotation in addition to the full restudy (the 'Full Speed Ahead!' scenario), the annual increase will be $5.8 billion in output and 90 092 new jobs by 2045. The former adds only 0.9%, and the latter 2.18% to their respective totals (base 1) for 2045.

(2) The annual increase in output and employment are very large numbers when compared to the annual increase in the base, especially in the later years. The 'Full Speed Ahead!' scenario generates 49.5% more output and 358% more employment than the annual increase generated by the base in 2045. This finding has important implications for those 'lumpy' public investments, such as water and sewage treatment plants for which the annual increase is critical in determining the need for new capacity, as well for the annual spending levels on education, police, and other services.

__________________

294   R. Weisskoff

Table 5.
Summary of scenarios

____________________

Missing Pieces in Ecosystem Restoration   295

Table 5.
(Continued)

Methods for Table 5:

Panel A:   annual impacts refer to full direct-indirect-induced impacts of scenarios in each of the benchmark   
                years at column head. 
Base 1:     Refers to a hybrid BEBR-BEA forecast for each year estimated by applying growth rates to the     
                most recent published base year estimate. Base year output is for 1995 ($225 billion in 1995 
                dollars) from IMPLAN data for the region. Base year total employment is computed for the region 
                from Bureau of of Economic & Business Research (1998a) for 2000 and 2010. Growth rates for 
                output and employment for 2000-2010 from BEBR (1998a) and for 2010-2015, 2015-2025 and 
                2025-2045 from BEA (1999) were applied to those output and employment base years to 
                compute 'base 1' forecasts for the later years. 
Panel B:   Share of the annual missing piece impacts to Base 1.
Base 2:     Refers to the annual increase in output and jobs generated by the Base 1 economy, as calculated  
                by applying the appropriate summary growth rate from Table 2 to Base 1 for each of the 
                benchmark years.
Panel C: Ratio of Panel A to Base 2. 
________________________

296   R. Weisskoff

Figure 5. 
Annual employment generated by base and scenarios

_______________________

Missing Pieces in Ecosystem Restoration   297

(3) The findings of our model do not lend direct support to the hypothesis that the new demands created by the 'missing pieces' will add significant stress to the ecosystem. This may be due to the shortcomings of our static model or to an understatement of the magnitude of the missing pieces. Nor does our analysis pinpoint those geographic locations or those sectors which may become the pressure points on the natural system.

It is not the contribution of new projects to new output or employment that should concern us, but rather how the new projects contribute to the growth of the basic economy. The $7.8 billion for Everglades restoration may permit the basic economy to grow to its forecast levels since the economy is already encountering severe ecological constraints. In the absence of restoration, the present growth path may be jeopardized. It is this frontier-the interaction of economy and ecology-that is excluded from both the economic modelling reported here and landscape modelling which underlies restoration planning.

(4) Several other reservations should be noted regarding the impact models applied here. The fixed technical coefficients, consumption patterns, and employment ratios should realistically be permitted to change over the years, along with the degree of inter-sectoral linkages and the local content of consumption. Regional linkages might actually be weakened with improved transport to other regions, greater international integration, and increased regional specialization. The IMPLAN model captures none of the dynamic effects of investment, prices, wages, and population. Local household consumption should be reduced once the cost of the complete restoration is finally assigned to the regional, state, and federal governments. Finally, the entire region, which has here been modelled as a single ecosystem, could be usefully broken into several discrete regional economies that correspond to the four major watersheds within the greater Everglades area. All the above qualifications constitute our future research agenda.

In the absence of the restoration, the continuing growth of the regional economy faces severe constraints due to limited land and water. Current state and federal policies promote urban sprawl and agricultural expansion. Tax exemptions, wage subsidies, low impact fees and low utility rates are all packaged to attract new manufacturing, commercial and tourist facilities and to develop new housing. Generally low energy rates encourage the construction of larger air-conditioned homes, and low water rates promote green lawns all year round.

Is all this growth desirable? Slower population growth need not lead to reduced consumption. Fewer but larger houses and yet greener lawns add pressure on land and water, just as increasing vehicle size implies more road congestion and air pollution. In the absence of extremely aggressive policies of conservation, the present trend is to encourage greater resource use in the urban areas. In the countryside, the water-intensive and land-extensive technologies borrowed from temperate agriculture for the region's crops, dairies, and fish ponds may prove highly inappropriate in the fragile, semi-tropical Everglades ecosystem.

Either path-higher levels of activity with current resource intensities implicit in our static projections, or higher levels of population with higher intensities-will accelerate the ecological pressures. Absent in the growth modelling are the costs the region faces as a consequence of these pressures. The 'restoration' efforts, designed by engineers from outside the region, need to be complemented by efforts at 'smart' consumption, designed and imposed by tax-payers inside the region, as

__________________

298   R. Weisskoff

the broad population comes to recognize the true social costs of living within a fragile ecosystem.

Hurricanes and rising sea levels may be Nature's way of imposing her own limits on growth. Hurricane Andrew caused $425 billion in damages in just a few days in August 1992, and powerful hurricanes in 1926 and 1928 shut down South Florida land speculation for two decades. New hurricanes could put a stop to the current building frenzy-or shift it northward to the less densely populated counties. Rising sea levels may also be part of a violent rather than a gradual process, with major storms effecting changes that irrevocably alter the shoreline.¹⁷ If the present trend continues' much of South Florida could be under water by the middle of the 21 St Century.

The economy may exert brakes of its own, for the South Florida engines of growth, such as tourism and construction, are sensitive to the national business cycle. The opening of Cuba and access to its sugarcane lands could result in the reduction of Everglades agriculture, especially as the peat soils of the EAA subside and productivity falls in much of the area. These factors aside, it may be only the redesign of South Florida's urban sprawl and changes in farm practices that jointly enable the Everglades to survive into the 21st Century.

Notes

1.   See Wier (1998) on Denmark; Ostblom (1998) on Sweden; Lange (1998) on Indonesia; Zhang (I 998) on China and Antweiler (I 996) for pollution caused by the trading patterns of 164 countries.

2.   See US Army Corps of Engineers (1998, Appendix E).

3.   The number of housing starts has been averaged for each decade due to the wide annual fluctuations. Also note that South Florida has accounted for more than half of the state's multifamily starts since the 1970s, a share considerably greater than the region's population share. See Table 2, lines 7a and 7d. I.

4.   See McC]uney (1969), Douglas (1988), Derr (1989) and Watercourse and South Florida Water Management District (I 996) for history and background. See South Florida Ecosystem Restoration Task Force Working Group (1998) for detailed expenditures to date.

5.   It is instructive to compare the two C&SF Project proposals (US Army Corps of Engineers, 1948, 1998). See also www.evergladesplan.org

6.   Litigation began with the seminal 1988 case, U.S. v. South Florida Water Management District, et al., in the US District, So. Dist. of Fl., Case No. 88-1886-CIV-HOEVELER, which cited the sugar growers as the primary polluters of the Everglades. A complete collection of Everglades litigation is housed at the University of Miami Law School and can be viewed at http://www.law.miami.edu/library/everglades.

7.   See Lissakers (1991).

8.   See Governor's Commission for a Sustainable South Florida (1995).

9.   IMPLAN was extended to Puerto Rico by Ruiz et al. (1994).

10.  The notation follows Alward & Lindall (1996).

11.  Holland & Wyeth (1993) compare 1-0 and SAM multipliers. The IMPLAN Newsletter, January 1999, p. 2 (available on the website, www.IMPLAN.org) compares multipliers computed from the full SAM with those computed from the 1-0 framework with value-added and consumption collapsed into single vectors.

12.   See comments in Rose et al. (1988, p. 10), on the simultaneity of the research by Weisskoff (1976) and Miyazawa (1976). See Weisskoff & Wolff (1975, 1977) for method. See Weisskoff (1985, ch. 15) and Hewings & Jensen (1986) for literature reviews, and Rose & Beaumont (1989) and Jackson (1998) for recent regional applications. For recent applications, see Weisskoff (1994) on Nicaragua, and Ruiz & Wolff (1996) on Puerto Rico.

13.  See Rey (1997).

14.  On the compatibility of profitable agriculture and improved environment, see Glaz (1995). On subsidence, see Shih et al. (1997).

__________________

Missing Pieces in Ecosystem Restoration   299

15.  The historical rate of increase in nature tourism (3.4%) is also similar to the average annual growth rate of the number of visitors (3.61%) to all the state parks in the region from 1993-97.

16.  This scenario is extremely conservative, as it excludes the growth of other leading sectors, such as greenhouses and commercial fishing. The recent introduction of 'best management practices' (BMPS) in sugar growing has led to dramatic declines in nutrient runoff and a surprising rise in sugarcane output, suggestive of the potential impact of changing cropping patterns and practices on the environment.

17.  See Wanless et al. (I 994).

References

Alward, G. & Tindall, G. (1996) Deriving SAM multiplier models using IMPLAN. Paper presented at the 1996 National Users Conference, Minneapolis, MN, 15-17 August.

A,ntweiler, W (1996) The pollution terms of trade, Economic Systems Research, 10, pp. 361-65.

Derr, M. (I 989) Some Kind of Paradise: A Chronicle of Man and the Land in Florida (New York, William Morrow).

Douglas, M. S. (I 988) The Everglades, River of Grass (Sarasota, FL, Pineapple Press).

Duchin, F. & Lange, G. M. (I 994) The Future of the Environment.- Ecological Economics and Technological Change (New York, Oxford University Press).

Glaz, B. (I 99 5) Research seeking agricultural and ecological benefits in the Everglades, Journal of Soil and Water Conservation, 50, pp. 609-612.

Governor's Commission for a Sustainable South Florida (I 995) Initial Report (Coral Gables, FL).

Hewings, G. J. D. & Jensen, R. C. (1986) Regional interregional and multiregional input-output analysis, in: P. Nijkamp (ed.) Handbook of Regional and Urban Economics, Vol. I (Amsterdam, North Holland), pp. 295-355.

Holland, D. & Wyeth, P. (I 993) SAM Multipliers: Their Decomposition, Interpretation and Relationship to Input-Output Multipliers, Research Bulletin XB 1 027 (Pullman, WA: College of Agriculture and Home Economics Research Center, University of Washington).

Jackson, R. W. (1998) Regionalizing national commodity-by-industry accounts, Economic Systems Research, 10, pp. 223-238.

Lange, G.-M. (1998) Applying an integrated natural resource accounts and input-output model to development planning in Indonesia, Economic Systems Research, 10, pp. 113-134.

Leontief, W., Carter, A. P. & Petri, P. A. (I 977) The Future of the World Economy (New York, Oxford University Press).

Light, S. S. & Dineen, J. W. (1994) Water control in the Everglades: A historical perspective, in: S. M. Davis & J. C. Ogden, Everglades: The Ecosystem and Its Restoration (Boca Raton, FL, St Lucie Press). Lissakers, K. (1991) Banks, Borrowers, and the Establishment.- a Revisionist Account of the International Debt Crisis (New York, Basic Books).

McCluney, W. R. (ed) (1969) The Environmental Destruction of South Florida (Coral Gables, FL,

University of Miami Press).

Minnesota IMPLAN Group, Inc. (1997) IMPLAN System (1995 data and Vets. 1.1 software). 1940 So. Greely St, Suite 101, Stillwater, MN. 55082. http://www.implan.com.

Miyazawa, K. (1976) Input-Output Analysis and the Structure of Income Distribution (Berlin, SpringerVerlag).

Ostblom, G. (1998) The environmental outcome of emissions-intensive economic growth: a critical look at official growth projections for Sweden up to the year 2000, Economic Systems Research, 10, pp. 19-29.

Rey, S. J. (I 997) Coefficient change in embedded econometric and input-output models at the regional level, Economic Systems Research, 9, pp. 307-329.

Rickman, D. S. & Schwer, R. K. (1995) A comparison of the multipliers of IMPLAN, REMI, and RIMS 11: Benchmarking ready-made models for comparison, Annals of Regional Science, 29, pp.363-374.

Rose, A., Stevens, B. & Davis, G. (1988) Natural Resource Policy and Income Distribution (Baltimore, Johns Hopkins University).

Rose, A. & Beaumont, P. (1989) Interrelational income distribution multipliers for the US economy, in: R. E. Miller, K. R. Polenske & A. Z. Rose, Frontiers of Input-Output Analysis (New York, Oxford University), pp. 134-147.

__________________

300   R. Weisskoff

Ruiz, A. L., Weisskoff, R., Alward, G., Siverts, E., Hussain, A. & Maki, W. (I 994) Puerto Rico IMPLAN System: Model and Database Construction & Application (Atlanta, GA: US Forest Service)

Ruiz, A. L. & Wolff, E. N. (1996) Productivity growth, import leakage and employment growth in Puerto Rico, 1976-87, Economic Systems Research, 8, pp. 391-413.

Shih, S. F., Glaz, B. & Barnes, R. E. Jr. (1997) Subsidence Lines Revisited in the Everglades Agricultural Area, 1997, Technical Bulletin 902 (Gainesville, FL, Agricultural Experiment Station, University of Florida).

South Florida Ecosystem Restoration Task Force Working Group (1998) Integrated Financial Plan 1998: South Florida Ecosystem Project Activities (Miami).

US Army Corps of Engineers (1948) Comprehensive Report on Central and Southern Florida for Flood Control and Other Purposes, reprinted from 80th Congress, 2nd session, House Doc. 643 (Washington DC: US Government Printing Office).

US Army Corps of Engineers (1998) Central and South Florida Project, Comprehensive Review Study. Draft Feasibility Report & Environmental Impact Statement (Jacksonville, FL).

US Department of Commerce, Bureau of Economic Analysis (1992) Regional Multipliers: A User

Handbook for the Regional Input-output Modeling System (RIMSII) (Washington, DC: US Government Printing Office).

Wanless, H. R., Parkinson, R. W & Tedesco, L. P. (1994) Sea level control on stability of everglades wetlands, in: S. M. Davis & J. C. Ogden, Everglades: The Ecosystem and Its Restoration (Boca Raton, FL, St Lucie Press).

Watercourse and South Florida Water Management District (1996) Discover a Watershe & The Everglades (Bozeman, Mont., Montana State University).

Weisskoff, R. (1976) Income distribution and export promotion in Puerto Rico, in: K. Polenske & J. Skolka (Eds) Advances in Input-Output Analysis (Cambridge, MA, Ballinger).

Weisskoff, R. (I 985) Factories and Food Stamps: The Puerto Rico Model of Development (Baltimore, Johns Hopkins University).

Weisskoff, R. (1994) The dilemma of export revival: Nicaraguan agriculture at a turning point, in: I.

T. de Alonso (ed.) Central America: Trade, Industrialization, and Integration in the Twentieth Century Central America (Westport, CN, Praeger).

Weisskoff, R. & Wolff, E. N. (1975) Development and trade dependence: the case of Puerto Rico, 1948-1963, Review of Economics and Statistics, 57, pp. 470-477.

Weisskoff, R. & Wolff, E. N. (1977) Linkages and leakages: industrial tracking in an enclave economy, Economic Development and Cultural Change, 25, pp. 607-628.

Wier, M. (1998) Sources of Changes in Emissions from Energy: A Structural Decomposition Analysis, Economic, Systems Research, 10, pp. 99-112.

Zhang, Z. (1998) Macro-economic and sectoral effects of carbon taxes: a general equilibrium analysis for China, Economic Systems Research, 10, pp. 135-159.

Data Sources (for Tables 1 and 2 and Appendix A)

See  Table 1,  Table 2,  Appendix A

Alvarez, J., Lynne, G. D., Spreen, T H. & Solove, R. A. (1994) The economic importance of the EAA and water quality management, in: A. B. Bottcher & F. T. Izuno, Everglades Agricultural Area (EAA): Water, Soil, Crop, and Environmental Management (Gainesville, FL, University Press), pp. 194-223.

Bureau of Economic and Business Research (BEBR) (1998a) The Florida Long-Term Economic Forecast 1998. Vol. 1: State & MSAS; Vol. 2: State & Counties (Gainesville, FL, BEBR).

Bureau of Economic and Business Research (BEBR) (I 998b) 1998 Florida Statistical Abstract (Gainesville, FL, BEBR).

Bureau of Economic and Business Research (BEBR) (I 999) Long-term population forecasts by county, 2000-2020, Population Studies (Gainesville, FL, BEBR).

Burns & McDonnell, Inc. (1994) Everglades Protection Project, Palm Beach County, Fla. (Kansas City, MO).

English, D. & Kriesel, W. (1996) Linking the Economy and Environment of Florida KeysIFlorida Bay: Economic Contribution of Recreating Visitors to the Florida KeYSIKey West (Washington, DC, National Oceanic and Atmospheric Administration).

EnviroTools, Inc. (1998) Threatened and endangered species software (TESS), CD-ROM, 5200 NW 43 Street, Suite 102-314, Gainesville, FL 32606.

Izuno, F. T, Rice, R. W. & Capone, L. T. (1999) Best management practices to enable the coexistence of agriculture and the Everglades environment, HortScience, 34, pp. 26-33.

__________________

Missing Pieces in Ecosystem Restoration   301

Office of Tourism Research (I 996-98) Florida Visitor Study (various years) (Tallahassee, FL: Florida
Department of Commerce).

Olson, D. & Lindall, S. (1996) IMPLAN Professional Software, Analysis, and data Guide (Stillwater, MN).

Snyder, G. H., Burdine, H. W, Crockett, J. R., Gascho, G. J., Harrison, D. S., Kidder, G., Mishoe, J.

W, Myhre, D. L., Pate, F. M. & Shih, S. F. (I 978) Water Table Management for Organic Soil Conservation and Crop Production in the Florida Everglades Bulletin 801 (Gainesville, FL, Agricultural Experimental Station, University of Florida).

Strategy Research Corporation (1997-99) Visitor Profile and Tourism Impact:- Greater Miami and the Beaches, Annual Report (various years) (Miami).

US Census Bureau (1998) Statistical Abstract of the United States: 1998 (Washington, DC).

US Department of Agriculture (I 999) Census of Agriculture 1997, on web site: www.nass.usda.gov/ census/census97/volume 1/

US Department of Agriculture, Economic Research Service (1997) Rice: Situation and Outlook Yearbook (Washington, DC).

US Department of Agriculture, Economic Research Service (I 998) Sugar and Sweetener. Situation and Outlook Report (Washington, DC).

US Department of Commerce, Bureau of Economic Analysis (BEA) (1999) State Projections of Employment, Earnings, and Product, http://www.bea.doc.gov/gsp/gspdata/.

Visitor Services Project (1989, 1996) Everglades National Park Visitor Study (Moscow, Idaho, Cooperative Park Studies Unit, University of Idaho).

Appendix A: Data Sources and the Scenarios

The SAM Framework and the IMPLAN Database (for Tables 3 - 5).  

See:  Table 3, Table 4, Table 5

General. The IMPLAN database provides a consistent format for 21 economic and demographic variables and 528 industrial sectors for all the US counties by reconciling data from different sources. County data for the variables are controlled by state totals, and the state totals are controlled to the national totals. The regional input-output tables are adjusted by applying margins, deflators and regional purchase coefficients (R.PCs) to the national make and use tables made by the Bureau of Economic Analysis (BEA) for benchmark years and adjusted to the IMPLAN sectors. IMPLAN's SAM is based primarily on establishment data by place of work, and is reconciled through institutional transfers with expenditure and household income data reported by household place of residence.

Employment refers to full and part-time jobs for wage and salary employees and for the self-employed. Data are reconciled from three sources. The Labor Department's ES202 file for unemployment insurance gives annual averages of numbers and payrolls of the covered wage and salary workers at the 4-digit level for the counties. These are adjusted for non-disclosure by the Commerce Department's County Business Patterns (CBP) which provides 1st Quarter data on the number of wage and salary workers and the self-employed. BEA's Regional Economic Information System (REIS) provides control totals of the number of self-employed and proprietors on the 2-digit level. The three sources altogether are used to achieve consistency at the county, state and national levels.

Value-added consists of four categories: (1) labour income includes compensation, benefits, and non-cash payments; (2) proprietor's income for the self-employed, private business owners, and professionals; (3) other property-type income (OPTI), which includes income from interest, royalties, dividends, profits, and property income and distributed corporate profits; (4) indirect business taxes (IBT), which includes excise and sales taxes paid by individuals to businesses. National control totals for value-added are from the National Income and Product Account (NIPA), published in the Survey of Current Business. Data from REIS for

__________________

302   R. Weisskoff

the national level employment compensation and proprietor's income for 20-digit national industries are distributed to the 528 IMPLAN sectors. County totals for OPTI and IBT are developed from state employment ratios per sector and adjusted to control totals.

Output for groups of sectors and adjusted from various annual surveys, censuses, and accounts, and then state and country outputs for some sectors are derived with the use of national output, value-added, and county workforce shares.

Final demand is estimated nationally for six broad categories and then distributed to states and counties. Household consumption by sector and income class is estimated from the Consumption Expenditure Survey (CES) in purchaser prices, adjusted for margins and distributed to the 528 sectors according to the IMPLAN/BLS bridge. Totals are controlled with NIPA. Four government sectors (Federal nonmilitary and military and state non-education and education) are derived from procurement and sales data and from annual surveys. Inventory purchases, gross private capital formation, and foreign exports are from annual surveys.

Inter-institutional transfers are estimated by reconciling household income, based on the BEA's REIS CA 35 table constructed from establishment data, with household income and expenditure from the BLS's Consumer Expenditure Survey (CES), which is residence and household based. Federal and state taxes are from annual surveys and the CES. Imports, exports, and capital accounts are used for balancing.

Input-output tables: The national 'make' of commodities by industries is computed from the BEA benchmark tables, updated by price indices, and stored. The national use matrix for the 537 BEA sectors are re-aggregated and split into the 528 IMPLAN sectors as coefficients. Value-added, final demand, totals for industrial and commodity outlays are added and the matrix is re-balanced.

Regional purchase coefficients are used to construct the regional input-output tables. These are based on the 1977 Multiregional 1-0 Accounts (MRIOA). RPC's for shippable commodities are estimated for each region and industry, as functions of (a) the share of foreign imports of the commodity to regional production, and (b) the share of domestic imports to regional production. The former is set for each region to be the national share. The latter is estimated as a function of (1) total regional compensation in the industry, (2) the ratio of local to national employment in the industry, (3) the ratio of the regional to national share of employment in the industry, and (4) the relative size of the region. Errors in the system of RPC's are due to the heterogeneity of products within each sector and cross-hauling due to consumer preferences for commodities produced elsewhere despite a surplus of local production.

For detailed procedures and sources, see Olson & Lindall (1966), Book 3, 'Database Guide'.

Scenarios and Income Patterns (following the rows of Table 4.  See Table 4)

(1) Agricultural data: details on sugarcane and rice are found in Snyder et al. (1978), Alvarez et al. (1994), US Department of Agriculture (1997, 1998), and Izuno et al. (1999).

(2) Everglades Restoration: Line items for the detailed construction costs and also operation and maintenance costs for each of the projects within the Everglades Protection Project are specified in the appendices of Burns & McDonnell (1994). Details on the Restudy are in US Army Corps (1998),

__________________

Missing Pieces in Ecosystem Restoration   303

Appendix E. I 1, computed by the present author on the basis of seven prototypical structures.

(3) Tourist expenditures are constructed from the Office of Tourist Research (1996-98) for statewide spending patterns, Strategy Research Corps (1997-99) for urban Miami, English & Kriesel (1996) for detailed regional recreation in the Keys, and Visitor Services Project (1989, 1996) for Everglades National Park visitor spending. The growth of spending for auto and air visitors were averaged for 1994-98 for each mode and then deflated by the travel price index for the state. Total spending level was computed from per-day expenditures and the total number of person-nights reported for each mode and for the regional destinations by quarter. Almost half the air visitors but only 13-16% of auto visitors reported South Florida as their regional destination.

(4) The 'new development package was estimated on the basis of the 1996 sector profiles for new construction, given in Bureau of Economic & Business Research (1998b).

(5) Expenditure patterns are given by IMPLAN for three levels of spending over the full range of BLS commodities. New residents are assumed to be attracted at twice the present rate of 43 000 middle-income households per year.