Report
Documenting the Economic Impacts of New York City’s Investment in Water Supply: A Data Visualization Foundation for Economic Analysis
Abstract
The New York City (NYC) Water Supply System, consisting of the Catskill/Delaware and Croton Water systems in upstate New York, provides one billion gallons of potable water to New York City’s 8.5 million residents every day. With an investment approaching $3 billion from 1993 to 2019, NYC Department of Environmental Protection (DEP) is interested in understanding the economic impacts of source water protection. This project develops an interactive, web-based visualization tool for exploring trends in and relationships between DEP’s Land Acquisition Program and economic metrics including annual Real Average Salary, Average Employment and number of Establishments in NYC’s watersheds. A literature review in the field of ecosystem service valuation was additionally developed with key interest from DEP.
1 Introduction
1.1 Motivation
City agencies often lack capacity to analyze the impacts of their investments and to present their findings in a manner that can inform decision making. In collaboration with the New York City Department of Environmental Protection (NYC DEP), this project aims to supplement previous analysis efforts by producing a methodology to visualize the economic performance of the region impacted by NYC’s source water protection programs.
The cost and sustainability of supplying reliably clean water is a concern for cities globally. NYC DEP has had considerable success in this area, supplying one billion gallons of unfiltered, potable water a day for nine million residents of the city and surrounding areas. This water comes mainly from north of the city in the Catskill Mountains West of the Hudson River (WOH) and secondarily from counties East of the Hudson River (EOH) (Figure 1).
NYC began pulling its water from the EOH region in 1842 and expanded to the WOH in 1905 [1]. During these periods NYC used eminent domain to purchase property and protect against sources of contamination making this land permanently unavailable for commercial development. Watershed communities have balked that this practice limits potential sources of employment for workers, farming practices, and stifles sources of taxation to support infrastructure and schools. In 1997, Mayor Rudy Giuliani signed a Memorandum of Agreement (MOA) ensuring the enactment and continuation of watershed operations, research, and protection of the Southeastern part of New York State, where the watersheds are located [2]. One key policy is the Land Acquisition Program (LAP) beginning from 1997 from which DEP has purchased surrounding lands. The MOA also established Catskills Watershed Corporation to ensure that local communities have a consulting role over the water system and to administer investment of City funds into infrastructure repair and the development of recreational and tourist industries in the watershed [3]. Since 1997, DEP has also invested close to $3 billion through its watershed programs, prompting interest in measuring the economic value these actions have brought to the local and wider New York communities.
1.2 Literature Review
An area of interest to the DEP was exploring published research methods used to determine ecosystem service valuation. The DEP investments over the years fostered a healthy watershed, a natural ecosystem that collects and filters precipitation through soil, streams, and into reservoirs [2]. In this way, the DEP has provided an “ecosystem service” to the communities that consume its water. Ecosystem service valuation is a developing area of environmental economics. As no two watersheds are alike, neither in source, landscape, amount of water generated, or size of the population it serves, identifying viable valuation methods from the literature provides a guiding framework to measure economic impacts as a result of the NYC watershed.
Ecosystem service valuations include the Hedonic Pricing Method, where the value of the natural resource is examined through nearby property values, Market Based, which approximates the value of environmental goods bought and sold on the market, and Replacement Cost Method, which calculates the amount of money it takes to replace an ecosystem service with a human made equivalent, to name a few [4].
Despite the suspected viability of methods researched and excitement of the DEP to ideate from these methods, the relevant data and metrics to conduct these valuations were not available during this project or of DEP interest to explore at this time. A separate, more robust literature review was requested by the partner sponsor which can be found as Appendix D of this report. This has been shared with the DEP and met with positive feedback as a resource on a variety of valuation methods.
Considering the limitations, and after discussions with DEP, the project instead turned its focus to the job and wage trends within the watershed counties. DEP provided unpublished, internal reports that considered economic data in the watershed counties. This gave the team guidance and inspiration to pursue economic data in this project.
1.3 Research Question
Past reports analyzing and communicating economic impacts of NYC’s watershed investments were lengthy, static and limiting in informing policy decisions. A visualization approach is thus favored, focusing on examining: How has land acquisition around the watershed, and associated changes in land uses, correlated with economic trends in the EOH and WOH watershed counties? This paper hypothesizes that strong, conclusive correlations will be observed in industries most dependent on types of land acquired under the LAP and in counties with the highest area overlap with the watershed.
This project has produced a web-based, interactive data visualization site displaying county level temporal trends relating to lands acquired under DEP’s LAP, employment and wage for each major industry sector and the correlations between the two. This tool will serve as an internal resource for DEP.
2 Data
The final analysis studied data from DEP’s LAP and the annual employment and wage trends of each major industry across the eight watershed counties. In the case of WOH, land use data was further added to complement the LAP data. All data was compared at the annual and county level, from years 2000 to 2018. The following sections discuss the characteristics and limitations of the three selected datasets.
While qualitative and quantitative data on agricultural productivity and land use impacts on flooding have been additionally surveyed within this project, they were excluded from the final analysis due to the lack of or sparsity of data. Appendix A details an assessment of these topical areas and describes the data quality of all datasets considered for this study.
2.1 DEP Land Acquisition Program
Purchasing lands and regulating development within and in proximity to the watershed helps protect the quality of the watershed environment, and contributes to the local economy when the land acquired is later open to public recreation access. A shapefile of plot level data of the lands DEP owns and has acquired for the purposes of watershed protection was utilized. The dataset includes acquisition date, property area in acres and the acquisition method such as whether it is pre-owned or acquired through conservation easements. However, the dataset lacked critical information about how much the properties were acquired for, or how much each property contributes in taxes each year. Due to data confidentiality reasons, plot level tax information was not released by DEP.
From 1997 to early 2020, more than 142,000 acres of land were acquired in WOH and more than 11,000 acres in EOH. Majority of those were in Delaware (84,000 acres), Green (26,000 acres) and Ulster (21,000 acres) counties.
2.2 Land Use Land Cover
Plot level land use shapefiles for the WOH area availed by DEP were also included in the final analysis. Land use information helps distinguish the functional values of each acquired land, such as whether the land is suitable for recreation or agricultural activity. As such, land use data was spatially joined with the LAP dataset. Detailed land use codes were then aggregated to produce four main categories: Forest Land (84.02% of total acres), Agricultural Land (13.7%), Urban/Built Up Land (1.3%) and Others (0.8%).
2.3 Employment and Wage
There has been interest from DEP to identify how watershed policies may have shaped and improved job opportunities and wages for the watershed counties. The NYS Department of Labor’s Quarterly Census of Employment and Wages (QCEW), aggregated at the annual and county levels from 2000 to 2018, provides data on annual average wages, number of establishments and average employment numbers for various industries specified under the 2012 North American Industry Classification System (NAICS). Inflation adjustment of wage data to 2018 levels was subsequently performed by using the US Bureau of Labor Statistics’s Consumer Price Index for All Urban Consumers of the New York-Newark-Jersey City regions.
One of the main limitations involved is that the finest spatial resolution available is at the county level. This resolution is too coarse given that the watershed covers a different percentage area of each county and impacts of watershed investments should vary depending on the level of investment proportional to the watershed area. For instance, the WOH watershed covers 54% of Delaware county’s area but covers only 7% of Sullivan county. As such, it may be expected to observe Delaware county be more sensitive to DEP’s investment as compared to Sullivan county. While census tract level data is available in the American Community Survey, those data exist only at either three or five year level estimates available from 2005-2013 and 2010-2018 respectively. As such, the more spatially and temporally robust QCEW dataset was chosen instead.
3 Methodology
A data visualization approach was taken to uncover the relationships between DEP’s LAP and the watershed’s employment and wage characteristics. Visualizations were designed and then explored for three areas: LAP trends (Section 3.1), economic metric trends (Section 3.2) and Pearson correlation scores between LAP and economic metrics (Section 3.3). Section 3.4 describes the limitations of the method while the technology stack utilized is documented in Appendix B: Technology Stack.
3.1 LAP Trends
One visualization important to contextualizing the LAP data was the cumulative number of acres acquired over time within the watershed counties (Figure 2). For example, in 2015, DEP acquired three times as much land in terms of acreage in Delaware county than in Greene. This approach reveals which counties may be more or less impacted by DEP actions.
Figure 2: Cumulative Number of Acres Acquired under LAP
For WOH counties, further land use information was visualized for the lands acquired (Figure 3). This visualization revealed that the majority of lands acquired since LAP began 1997 is characterized as either Forest or Agricultural Land. This would suggest that impacts to industries heavily dependent on Forest or Agricultural Land will be more significant than those reliant on Urban/Built Up space.
3.2 Economic Metric Trends
Further exploration via visualization was also conducted for four economic metrics: Real Average Salary, Real Total Salary, Establishments, and Average Employment trends. These trends are visualized as a stacked bar chart for each county, with each bar representing each of the main 2012 NAICS industry sectors. For instance, Figure 4 illustrates how Ulster’s economy has grown towards 2007, shrank over the next five years before growing again. These changes correspond to those of the Finance and Insurance (sector 52) and Administrative and Waste Services (sector 56) sectors. However, despite these changes, Health Care and Social Assistance (sector 62) has consistently increased while Manufacturing (sector 31) has decreased.
In general, mixed trends were observed across the four metrics and across the eight counties. Sample plots and brief analysis are included in Appendix B: Economic Metric Trends.
3.3 Pearson Correlation
To determine whether DEP’s policies and land acquisition have had positive and significant impacts on local communities, the Pearson correlation method was adopted. At the county and annual level, the Pearson correlation is calculated between each available economic metric for each NAICS industry, and the number of acres DEP has acquired for that specified use types. These results are then visualized in an interactive heatmap, as shown below (Figure 5). county. For WOH counties, further the number of acres is further broken down to different land
Pearson correlation values range from -1 to 1, with -1 implying strong negative correlations and 1 being strongly positive. A histogram plot of the values within the displayed heatmap show that the majority of the correlations between real average salaries and LAP are weakly negative, between the values -0.4 and -0.1 (Figure 6).
Finally, to observe the distribution of correlation values within each county, the values of each metric were sorted in descending correlation and the counties were arranged in ascending percent county area covered by the watershed (Figure 7).
As observed above, a balance of positive and negative correlations are observed in Average Employment whereas Real Average Salary is dominated by weak or strong negative correlations. Strong positive correlations above 0.7 are only observed within Real Total Salary and Average Employment whereas strong negative correlations are observed in all metrics except Real Average Salary. Ironically, Delaware, whose county area overlaps greatest with the WOH watershed, appears to experience the weak correlations across all metrics and industries. This is especially when compared to Putnam, its EOH counterpart and Ulster, the third largest WOH county in terms of watershed coverage.
3.4 Limitations
Due to a lack of robust investment data, there were limitations in the methodology since the percentage of LAP’s budget being allocated to each county was unavailable. As a result, inference on the scale of program expenditure per county is instead accounted for by the percentage of county area covered by the watershed. Additionally, the exploration of economic impacts required defining economic indicators more broadly and in line with changes in employment and industries across time.
4 Results
The visualization tool is designed to be flexible in order to allow DEP to explore data. The following sections demonstrate examples of how the tool may be used to uncover insights.
4.1 Trends in Correlation Values
Overall, minimal trends were observed for correlation values across the counties and industries. For example, Figure 5 shows that correlations in Real Average Salaries for the Agriculture, Forestry, Fishing and Hunting sector is positive in WOH counties Delaware (correlation value +0.22), Greene (+0.16), Schoharie (+0.54) and EOH counties Putnam (+0.15) and Westchester (+0.04), but negatively correlated in counties Sullivan (-0.31), Ulster (-0.69) and Dutchess (-0.13). Furthermore, of these eight counties, only Schoharie and Ulster possess moderately strong correlations of magnitude above 0.5. This fails to support the hypothesis that the industries most dependent on agricultural and forest lands acquired under LAP are uniformly and strongly affected by DEP’s policies.
The two sectors Finance and Insurance, and Real Estate and Rental and Leasing appear to be the most consistent negative correlations across the counties. Sullivan and Ulster also appear to comprise mainly of negative correlations as compared to mixed results for other counties.
The lack of clear trends within the correlation values point towards the difficulty of measuring the economic impact of a specific policy. DEP can benefit from complementing exploratory data analysis within the watershed counties with high granularity data the agency already possesses and comparing the watersheds development to that of the region.
4.2 Ulster County
Ulster presents an interesting case as it produces most of the strongest positive and negative correlations across Real Average Salary, Average Employment and Establishments despite having a moderate watershed coverage of 31%. Evident from Figure 7 above, Ulster had the highest number of strong negative correlations in Real Average Salary, with nine of its 14 industries possessing a correlation value smaller than -0.4. Similarly, five of its industries had a value of smaller than -0.5 in Establishments. Negative correlations arise from the increase in employment and establishments while the rate of land acquisition has slowed over the 20 year period (Figure 8). Table 1 further summarizes the industries responsible for Ulster’s highest and lowest correlation values.
Figure 8: Number of acres acquired in Ulster
Table 1: Highest and lowest correlation values in Ulster county
Plots of Table 1, detailed with land use type, were generated as a means to dive deeper contextually, comparing Ulster county’s stronger correlations against county reports and news articles on specific industries in order to understand how growth and declines is reflected.
Reports from Ulster County municipalities describe that there is growth in the Agriculture, Forestry, Fishing and Hunting Industry, reflected in the bold red line of Figure 9 [5]. The strong negative correlation in the data is reflected in the plot by the growing discrepancy of real average salary despite low amount of acquired lands. This was also true of official county reporting and trends in both the Arts, Entertainment, and Recreation [7] as well as Professional and Technical Services Industries, see Figure 14 and 15 in Appendix C: Results [5, 6]. From the data, it is conclusive that the growth in these industries is likely not attributed to DEP actions.
The Information and Manufacturing Industries, both in the top ten private industries for Ulster County, reflects a strong positive correlation with land acquisition despite the industry decline over the years [5, 6]. Average employment in the Information Industry (Figure 10) has the strongest positive correlation at .74, and it’s noteworthy to observe how differences in correlation strength extract more precise observations compared to Manufacturing Establishments at a .67 correlation (Appendix C: Figure 16). It’s visually apparent that employment in the Information Industry between 2002-2004 is closely related to All Lands Acquired, but more specifically Forest Land, Urban/Built-up Land, and Other land types, while evidently negatively correlated with acquisition of Agricultural Land (2010-2012). This level of detail is informative to the DEP and sector’s understanding employment performance in relation to DEP actions.
Despite this clarity being true for Figure 10, similar conclusions for strong correlations cannot be made for Figures 9, 14, 15, & 16. Therefore, even with a closer look at the land types of acquired lands and overlaying it with contextual information, the results fail to support the hypothesis.
Figure 10: Ulster County, Average Employment of the Information Industry correlated acquired land use types
5 Conclusions
This project provides a flexible method for DEP to view the correlations of land acquisition and forest cover with economic activity in the EOH and WOH watershed counties. We believe that this tool will help DEP communicate better internally and will spur further thinking about new ways to visualize and communicate the findings of research and analysis. The project also provides a strong literature review to allow the partner sponsor to further investigate the potential for ecosystem service valuation.
DEP could expand these visualization methods in subsequent projects to consider more granular data. A key hurdle, however, will be limitations on what data can be made publicly available. DEP appears to have a deep resource of data, but also maintains firm restrictions on data access, which will need to be considered before initiating further visualization projects. As was highlighted above, the marginal correlation of land acquisition with economic indicators appears to be highest in counties with lower rates of protected lands. This sensitivity may be an interesting area for DEP to explore further, with the help of this tool.
6 References
[1] Catskill Watershed Corporation. “History of the NYC Water Supply”, 2020, cwconline.org/history-of-the-nyc-water-supply/
[4] Agar, Jesse. “Valuation of Ecosystem Services: Classes of Values.” Valuation of Ecosystem Services: Classes of Values | Conservation Strategy Fund, 2015, www.conservation-strategy.org/csf-economic-video/classes-values
[5] Ulster County Government. “Industries: Ulster County.” Industries | Ulster County, 2020, ulstercountyny.gov/node/35417.
[6] Cedar Lake Ventures, Inc. “Industries in Ulster County, New York (County).” The Demographic Statistical Atlas of the United States - Statistical Atlas, 2020, statisticalatlas.com/county/New-York/Ulster-County/Industries.
[7] Ryan, Patrick K. “Arts & Culture: Ulster County Alive!” Ulster County Tourism, 2020, ulstercountyalive.com/landing/arts-and-culture.
Additional References
Iyer, Sangamithra. “Environmental Reviews.” Extended New York City Watershed Land Acquisition Program - DEP, 2010, www1.nyc.gov/site/dep/about/extended-nyc-watershed-land-acquisition-program.page.
Downeast Development Consulting Group. “The New York City Watershed Economic Impact Assessment Report.” Delaware County Chamber of Commerce, Delaware County Board of Supervisors, May 2009, dev.co.delaware.ny.us/wsa/wp-content/uploads/sites/2/2017/06/NYCWatershedImpactStudy-FinalReport2.pdf.
City of New York. “NYC Watershed Memorandum of Agreement.” EPA, Environmental Protection Agency, 2016, archive.epa.gov/region02/water/nycshed/web/html/nycmoa.html.
U.S, Department of Labor. “North American Industry Classification System (NAICS) at BLS.” U.S. Bureau of Labor Statistics, U.S. Bureau of Labor Statistics, 27 Feb. 2020, www.bls.gov/bls/naics.htm.
Appendix
A: Data
Exploration of Other Economic Areas
Agricultural Productivity
The agriculture industry has been highlighted by DEP as one of the industries they were interested in further analyzing. One research question this project proposed was on whether a reduction in agricultural land available within the watershed counties has instead motivated farmers to prioritize increasing agricultural productivity and efficiency. In other words, it may be a positive policy impact if yield per acre increased over the years and when compared to non-watershed farms. Available annual yield per acre data for a few commodities within the WOH watershed counties was acquired from the US Department of Agriculture (USDA). However one main challenge prevented further research in this direction: depending on the product, yield was recorded in at least two different units - tons and bushels. Plotting annual yield as a stacked area plot below, it is observed that the two groups share a somewhat inverse relationship consistently across the five counties. This is perhaps due to crop rotation and geographical factors. As the team lacked domain expertise in agricultural data, we were unable to correctly aggregate the two different units of measurement into one composite index of yield per acre. If this issue is resolved, computing percentage change in yields per acre and comparing those statistics to neighboring non-watershed counties may help validate or refute the hypothesis that land acquisition has motivated more productive agricultural activity.
Figure 11: Annual yields of select agricultural commodities in the WOH watershed counties
Impact of Land Use Change on Flood Damage
Another research question explored was whether land use change resulting from LAP has impacted flood severity and damages to property. It is hypothesized that watershed protection may reduce the flood plain and provide better flood defence to local communities. To consider the impacts of flood in the WOH region, county level data of each flood occurrence, property damages and deaths data was accessed from the National Oceanic and Atmospheric Administration’s (NOAA) National Centers for Environmental Information (NCEI). However, exploratory analysis revealed that the data was sparse and poorly populated. Further, as seen below, pearson correlations computed against annual acres of land acquired in the respective counties were mostly insignificant particularly in terms of property damage. Given low data density and little number of data points, flood impacts were removed from the final analysis.
Figure 12: Pearson correlation between annual flood frequency and property damage, and annual acres acquired under LAP per WOH county
Assessment of Considered Datasets
B: Methods
Economic Metric Trends
Figure 13: West of Hudson Average employment numbers (left) and inflation-adjusted average annual salary (right), by industry
Mixed trends were observed in average employment across the five WOH counties. For instance, while employment in Delaware has gradually declined over the years (due to the shrinkage of Manufacturing and Retail Trade sectors), employment in Green and Sullivan have increased. Greene has seen a growth in the Retail Trade sector whereas Health Care and Social Assistance sectors have grown in Sullivan. On the other hand, annual average real salaries have gradually increased over the years for all five counties, with the introduction of the Utility industry being a primary factor in all five counties excluding Ulster.
Technology Stack
A web-based interactive data visualization platform is proposed given the following benefits:
It supports user interaction, such as toggling variables to visualize or explore in-depth aggregated/non-aggregated trends
It is easily shareable, such as in the form of a URL.
The software architecture and utilized technologies selected are as follows:
C: Results
Figure 14: Ulster County, Average Employment of the Arts, Entertainment, and Recreation Industry correlated against acquired land use types
Figure 15: Ulster County, Establishments in the Professional and Technical Services Industry correlated against acquired land use types
Figure 16: Ulster County, Establishments in the Manufacturing Industry correlated against acquired land use types
D: Ecosystem Services Valuation Literature Review per NYC DEP request
The DEP requested the team explore studied methods of ecosystem services valuation as a secondary output of the project. On the following page is what was delivered and was met with positive feedback, Project Sponsor Ira Stern of DEP stated, “thank you for this excellent literature review, it will serve us well.”
Ecosystem Services Valuation & Related Socio-Economic Factors
Literature Review
Center for Urban Science and Progress
Capstone Project - April 2020
The purpose of this document is to share with the New York City Department of Environmental Protection, our team’s Agency Partner, preliminary research and literature review findings on ecosystem service valuation and related information.
INDEX
The structure of this report is as follows:
Introduction……………………….……………………….……………………….……………………….……………………….……………………....… 2
Environmental Valuation Methods of Ecosystem Services……………………….……………………………................… 2
Classes of Values……………………….……………………….……………………….……………………….………………………………………... 3
Literature Review: Valuation of Ecosystem ……………………….……………………….………………………........................ 4
Literature Review: Socio-economic factors related to ecosystem valuation……………………...................... 7
Citations……………………….……………………….……………………….……………………….……………………….…………..………………….. 8
Introduction
Long-term investments made by city agencies often have a shortage in capacity to internally analyze the economic impacts in ways that are thorough, yet succinct, while remaining useful and informative for evaluation that can support future decision making. The project for our capstone with the New York City Department of Environmental Protection (NYC DEP) as Agency Partner and Professor Huy Vo of New York University’s Center for Urban Science and Progress as Faculty Advisor, takes on the challenge to synthesize and supplement previous analysis efforts and produce a methodology to illustrate the economic impacts of NYC DEP’s investment in the water supply. The origin of one billion gallons a day of unfiltered water, clean to drink from the tap, for nine million New York City and greater NYC residents dates back to a memorandum of agreement signed by the NYC Mayor in 1997 to ensure watershed operations, research, and protection of the Southeastern part of New York State where it is located [1].
An ecosystem service is a good or service provided by nature. In its valuation, it’s important to note that value can have subjective definitions since value can’t be measured from price alone or what people are willing to trade for it, a helpful distinction is consumer surplus which is the additional value-worth someone is willing to pay (or might have paid) for something [A]. There are 5 classes of valuation of ecosystem services which provide insight into the viability of quantifying different types of services [A]. In our literature review, we will begin outlining the several environmental valuation methods of ecosystem services that find alternative ways to measure the value-worth of consumer surplus.
First is the hedonic pricing method, as mentioned in the capstone write up as a possible added methodology overlaying a foundational map depicting place-based DEP investments over space and time with indicators relating to economic development. The hedonic pricing method tries to quantify the value of ecosystem services through examining housing prices [A]. Proximity to parks, lakes and nice views, the aspects people value when home buying, along with what they dislike, such as polluted areas, can be assumed to be reflected in property values that have the selected characteristics (or proximity to them) [A]. Then with data on property values and the selected characteristics one can use a regression analysis to build a function that would predict the change in value in the absence of the environmental services [A]. It would show the value of ecological services, given the perceived benefit of living near a more natural landscape[A]. Through the hedonic pricing method, it’s possible to be able to quantify the positive and negative impacts of specific characteristics [A]. A second valuation method is market based [A]. The aim here is to approximate the value of environmental goods bought and sold on the market such as timber, fish, water, animals, plants, berries, etc [A]. Since the goods are all things that can be traded, this valuation is done by measuring the consumer and producer surplus [A]. Replacement cost method is a third valuation method which calculates the amount of money it takes to replace an ecosystem service with a human made equivalent (in magnitude and service) [A]. For example, finding the infrastructure cost of installing a storm drain system in a place to control flooding provides insight on the valuation of the water holding capacity of the soil [A]. A fourth valuation method is called the stated preferences method, which uses surveys of the public and stakeholders to find ecosystem services value [A]. Other types of valuation methods of ecosystem services include the avoided cost method, travel cost method, contingent valuation, choice experiments, benefits transfer study to name a few [A].
Another practiced methodology in the valuation of an ecosystem, which can be independent of or in conjunction with the chosen method of those described earlier, involves adding the monetary total from each type of classification values. There are five ways to classify the benefits of a forest, for example. The first classification of values includes direct benefits [A]. This is the utility of using a substance directly either consumptive such as wood, water, and hunting or non-consumptive such as hiking on a trail or kayaking [A]. On the other hand, indirect use, encompases the usage (and willingness to pay) for a utilities output, such as the soil under the forest which allows for healthy infiltration which garners greater water availability [A]. In Use-Values, both direct and indirect, the ecosystem is producing something physical that people are willing to pay for. Another category is Non-Use value, which encompasses both bequest and existence value classifications [A]. Bequest places value on a community’s sentiment of the type of environment they would theoretically like to pass onto future generations, as opposed to which type of environment offers the greater present net-value [A]. Similarly difficult to quantify, existence value is the utility value of knowing something exists [A]. The last classification, option value, is found between Use and Non-Use value, is described as leaving the option open to utilize an environment in the future, and instead, paying a certain value to ensure any actions that would irreversibly alter the space would be postponed [A]. All these value classifications add up together to make the total economic value [A]. As one can imagine, it can be controversial to quantify and difficult to find data on the economic estimations of all five value classification of a land area before and after a specific ecological change to compare changes in value, but if related attributes are agreed on and transparent, it can provide a baseline for study [A].
To provide relevance, given the context the DEP provided of the change in forest cover near the watershed, the remainder of this literature review aims to bring to light academic research and discussion of the current academic landscape of value estimation methodology.
A starting point for social and ecological forest services, as discussed in “Development of the estimation methodology for the commercial balance of the market of social and environmental services of forests,” is finding a ratio of components to evaluate the commercial balance of the market [1]. This can be determined by both “quantitative and qualitative characteristics of the socio-ecological potential,” including, “the level of forest cover, the natural structure of the forest [1].” Additionally, it’s necessary to account for the “magnitude and direction of the demand for the socio-ecological benefits of forests [1].”
Suska proposes a methodology for “evaluation of the social and economic potential of forests” in three consecutive stages [1]. The first, a preemptive step, provides a baseline of both the environmental and social importance and danger of the project through identifying “environmental problems and social needs of the [project area] is planned to establish a market for social and environmental services of the forest [1].” The second stage calls for determining the range of services a wooded area may provide through conducting an evaluation that determines “the volume and quality and structure of forests (tree species, age) [1].” Lastly, the third step involves analysis of the marketing environment that allows to determine the possible volume and direction of demand [1].”
In Wilson’s “Forest Ecosystem-Service Transitions: the Ecological Dimensions of the Forest Transition” she proposes a conceptual framework to understand and compare the ecological services that arise from different types of forest transitions over time and calls this concept “forest ecosystem-service transition curves. [2]” By capturing the “dynamism that we observe in forest transitions”, Wilson illustrates that “different social drivers produce different types of ecosystem-service transitions. [2]” With the watershed land area having been under DEP ownership for the past almost three decades, this concept may likely come into play as the CUSP project takes into consideration how ecosystem services will “change over time at rates that differ among transition types. [2]”
There are a variety of ecosystem services provided by forests worldwide, including carbon sequestration, soil protection and water purification [3]. “Surface vs. Groundwater: The Effect of Forest Cover on the Costs of Drinking Water,” highlights that higher quality of drinking water is associated with forest coverage through both the “minimal use of pesticides and fertilizers” used in forests and the “the tree root system,” while having less influential impact on water treatment costs [3]. The results of Lopes case study of Portugal suggests an existence of “a positive and significant effect of local forest cover on water treatment cost savings of 0.056%,” and assess a “marginal effect of forest cover on drinking water treatment costs” [3].
A relevant article emerged in the literature review process entitled, “A Five-Step Protocol for Estimating Forest Cover and Rate of Change in the New York City Watershed” [4.] This work’s objective is to reduce the uncertainty surrounding the estimates of forest cover in NYC’s Catskill Mountain watersheds and it’s rate of change over time as an indicator found in “multiple published analyses of temporally approximate satellite imagery” [4]. Yavuz and Hall develop a “five-pronged protocol,” in which they “estimated the true percentage of forest cover as of 2002 is 72%,” a percentage that is “far less than that reported by other studies” [4].
This study, “Forest Cover Effects of Payments for Ecosystem Services: Evidence from an Impact Evaluation in Brazil” explores forest cover outcomes in a watershed made up of 81 properties serving the Rio de Janeiro region of Brazil [5]. Applying propensity score matching on the watershed properties over the 7 years of reforestation effort data, the study found that “impacts on forest cover were caused mostly by reduced deforestation rather than reforestation” [5]. The researchers conducted analytical ““placebo” impact tests” using baseline “pre-intervention” data, and observed “forest cover benefit came at the per-hectare cost of $32,963” which was mostly paid off with “off-site mitigation funds” [5].
According to research in the article titled, “Watershed Protection: Capturing the Benefits of Nature’s Water Supply Services,” through the “$1.5 billion spent on watershed protection over 10 years” there were at least $6 billion avoided in capital costs [6]. There is a key challenge in designing workable mechanisms to link beneficiaries with providers, either directly or indirectly [6]. The City opted to seek a waiver of the filtration requirement by investing in a comprehensive watershed protection program in Catskill-Delaware watershed, which supplies 90% of City’s drinking water [6].
Project team member Ross MacWhinney shared findings from, “Valuing Ecosystem Services” [7]. Several useful tools can be found in this paper, which include a brief description of the Ecosystem Valuation Methods that have emerged, a list of common ecosystem service types in a table format and lists the following for each: how amenable is it to economic valuation, which of the valuation methods described in are most applicable, and how transferable the analysis of these services is across sites [7]. Finally, there is a quantification of the number of peer-reviewed research papers published on the subject of ecosystem service valuation. The total number is 730 [7]. A summary of the sentiment of this article can be found in this quote, ". . .the fundamental challenge of valuing ecosystem services lies in providing an explicit description and adequate assessment of the links between the structure and functions of natural systems, the benefits (i.e., goods and services) derived by humanity, and their subsequent values" [7]. This assessment has led me to believe that the field of ecosystem service valuation is still very much developing and that straight forward valuation methods are difficult to come by.
Literature Review: Socio-economic factors related to ecosystem valuation
In “Selecting the Socio-Economic Metrics for Watershed Management”, this article argues that it’s critical in the decision making of “scientists, managers, and watershed citizen[s]” to document baseline conditions and analyze the dynamic relationships between ecosystems and human communities are important decisions for scientists, managers, and watershed citizen” [8]. The variety of social and economic data available must be used within the correct theoretical frameworks. The frameworks offered in this paper includes “social sanctions, sense of place, civic structure, and cultural differences [and underlying] all these frameworks are attitudes, beliefs, values, and norms that affect which questions are asked and which indicators are chosen” [8]. Additionally, “much work and significant challenges remain in developing a standard set of spatially based socio-economic metrics for watershed management” [8]. Morton and Padgitt share that “the knowledge needed to mediate between human communities and [river] ecosystems varies in terms of content, method, and purpose. The relationships are complex, dynamic, and affected by the scale of examination” [8].
Beisaw’s studies within the context of the New York City Watershed and in their research article “Water for the City, Ruins for the Country: Archaeology of the New York City Watershed,” writes in an austere tone that the “archaeological survey of city-owned watershed lands reveals a scarred landscape of ongoing colonial conflict” [9]. In speaking of the Board of Water Supply (BWS) and Department of Environmental Protection (DEP) actions as “shaping rural regions far outside city boundaries” focusing on the narrative of residents who were evicted and towns demolished and submerged to ensure the city’s water supply [9].
Citations
[A] Agar, Jesse. “Valuation of Ecosystem Services: Classes of Values.” Valuation of Ecosystem Services: Classes of Values | Conservation Strategy Fund, 2015, www.conservation-strategy.org/csf-economic-video/classes-values.
[1] Suska, Anastasiy. “Development of the Estimation Methodology for the Commercial Balance of the Market of Social and Environmental Services of Forests.” Technology Audit and Production Reserves, vol. 6, no. 5(38), 2017, pp. 20–24., doi:10.15587/2312-8372.2017.119595.
[2] Wilson, Sarah Jane, et al. “Forest Ecosystem-Service Transitions: the Ecological Dimensions of the Forest Transition.” Ecology and Society, vol. 22, no. 4, 2017, doi:10.5751/es-09615-220438.
[3] Lopes, Ana Faria, et al. “Surface vs. Groundwater: The Effect of Forest Cover on the Costs of Drinking Water.” Water Resources and Economics, vol. 28, Oct. 2019, pp. 100–123., doi:10.1016/j.wre.2018.06.002.
[4] Yavuz, Mehmet, and Myrna H. P. Hall. “A Five-Step Protocol for Estimating Forest Cover and Rate of Change in the New York City Watershed.” Environmental Monitoring and Assessment, vol. 190, no. 8, 2018, doi:10.1007/s10661-018-6855-4.
[5] Fiorini, Ana Carolina Oliveira, et al. “Forest Cover Effects of Payments for Ecosystem Services: Evidence from an Impact Evaluation in Brazil.” Ecological Economics, vol. 169, 2020, p. 106522., doi:10.1016/j.ecolecon.2019.106522.
[6] Postel, Sandra L., and Barton H. Thompson. “Watershed Protection: Capturing the Benefits of Nature's Water Supply Services.” Natural Resources Forum, vol. 29, no. 2, 2005, pp. 98–108., doi:10.1111/j.1477-8947.2005.00119.x.
[7] Liu, Shuang, et al. “Valuing Ecosystem Services.” Annals of the New York Academy of Sciences, vol. 1185, no. 1, 2010, pp. 54–78., doi:10.1111/j.1749-6632.2009.05167.x.
[8] Morton, Lois Wright, and Steve Padgitt. “Selecting Socio-Economic Metrics for Watershed Management.” Environmental Monitoring and Assessment, vol. 103, no. 1-3, 2005, pp. 83–98., doi:10.1007/s10661-005-6855-z.
[9] Beisaw, April M. “Water for the City, Ruins for the Country: Archaeology of the New York City Watershed.” International Journal of Historical Archaeology, vol. 20, no. 3, 2016, pp. 614–626., doi:10.1007/s10761-016-0363-4.
Last updated
