Modelado de Diseños de Redes de Gobernanza Colaborativa Alternativa

Journal of Public Administration Research and Theory, 2020, 1–20
doi:10.1093/jopart/muaa013
Article
Article
Modeling Alternative Collaborative Governance
Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Network Designs: An Agent-Based Model of
Water Governance in the Lake Champlain Basin,
Vermont
Patrick Bitterman,*,† Christopher J. Koliba†
*University of Nebraska–Lincoln; †University of Vermont
Address correspondence to the author at [email protected].
Abstract
With the widespread use of collaborative governance mechanisms for mitigating water pollution,
an opportunity exists to test alternative institutional designs based on collaborative governance
theory using computer simulation models, particularly when there is a clear relationship between
governance networks, observable resource allocation decisions, and measurable outcomes. This is
especially the case for wicked problems like nonpoint source water pollution where there are com-
pelling questions regarding how best to design policies, allocate funds, and build administrative
capacity to meet water quality standards. We present an agent-based model (ABM) of water gov-
ernance for the Lake Champlain Basin to simulate the impacts of alternative collaborative govern-
ance arrangements on the development of suites of water quality projects. The ABM is connected
or coupled with land use and phosphorus load accumulation models that are informed by existing
hydrologic models, project datasets, and state-set load reduction targets. We find that regionally
arranged collaborative governance in water quality project planning and implementation can lead
to better water quality outcomes, thereby affirming one of the central premises of collaborative
governance regime theory. We also find that externally mandated collaboration, as opposed to
voluntary, self-initiated collaboration, can lead to better water quality outcomes, adding to our
understanding of which type of collaborative governance arrangement is best suited to the spe-
cific contexts of this case. Further, without adequate administrative capacity in the form of human
resources located in central network actors to manage project funds, “administrative bottlenecks”
may form and money can go unspent. This research demonstrates the efficacy of using simula-
tions of alternative institutional design for theory testing and tuning, and policy prototyping.

The authors would like to thank our co-collaborators on the VT EPSCoR Introduction
Basin Resilience to Extreme Events (BREE) project including those
from state government: Julie Moore, Neil Kammen, Kari Dolan, David Over the last 20 years, theoretical frameworks have
Mears, Debra Markowitz, Karen Bates, Emily Bird, and Tim Pricer; and been advanced to inform the design and implementa-
core collaborators in BREE research: Asim Zia, Elizabeth Doran, Jory tion of effective collaborative governance arrangements
Hecht, Scott Turnbull, Donna Rizzo, Scott Hamshaw, Andrew Schroth,
Scott Merrill, Arne Bomblies, Judy Van Houton, Carol Adair, Richard
Clark, Stephanie Hurley, Richard Kujawa, Clare Ginger, Patrick Clemins Committee. We also thank summer interns Samantha Christopher and
and all of the members of the BREE Policy and Technical Advisory Alexandra Cole for their assistance.
© The Author(s) 2020. Published by Oxford University Press on behalf of the Public Management Research Association. 1
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.
org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly
cited. For commercial re-use, please contact [email protected]
2 Journal of Public Administration Research and Theory, 2020, Vol. XX, No. XX
(Ansell and Gash 2008; Bryson, Crosby, and Stone of a transdisciplinary research project that studies this
2015; Emerson and Nabatchi 2015). Those advancing social-ecological system as a complex adaptive system
collaborative governance theories often posit that col- (Koliba et al. 2016). Using this model, we explore and
laboration among policy actors tends to produce better test specific collaborative governance questions relating
results (Booher 2004; Emerson and Nabatchi 2015; to the efficacy of collaborative governance over “non-
Koontz and Johnson 2004; Pahl-Wostl et al. 2007), and collaborative” governance arrangements (H1); the ef-
there is building evidence to suggest this is so (Booher ficacy of self-initiated, voluntary collaboration versus
and Innes 2010; Rogers and Weber 2010; Scott 2015). externally mandated collaboration (H2); the extent to

As collaborative governance networks are increasingly which the functional aims of collaboration (planning
pursued as key strategies to drive “collective impact” and/or implementation) impacts performance (H3); and
(Kania and Kramer 2011) and improve outcomes, pol- the role that the persistence of administrative bottle-
icymakers, rulemakers, and public managers are also necks in the lead organization hinders mitigation (H4).
increasingly interested in questions of collaborative in-
stitutional design and operation (Kamensky 2019). Testing and Tuning Collaborative and Network
Calls for the development of experimental collab- Governance Theory
orative platforms to aid in this development through The use of collaborative governance to address envir-
the testing of “design rules” have been made (Ansell onmental problems has long been documented empiric-
and Gash 2018). While computer simulation models ally through comparative case study analysis (Imperial
have long been used to help stakeholders envision pos- 2005; Koontz et al. 2010), a large-N comparative
sible policy solutions, the use of simulations to model study (Scott 2015), social network analysis (Berardo
alternative governance design arrangements using and Lubell 2016; Knieper and Pahl-Wostl 2016), and
meso-level theories of governance originating from observed in historical assessments of eras of environ-
the field of public administration has begun to take mental policy, specifically found in the increased uses
root (Eckerd, Campbell, and Kim 2012; Kim 2007; of partnerships and incentives beginning in the 1980s
Koliba, Zia, and Merrill 2019; Maroulis 2016; Scott (Gerlak 2005). Collaborative governance theory has
and Thomas 2017; Zia and Koliba 2013). Drawing on been widely applied to environmental policy and gov-
a case involving collaborative governance design for ernance situations and is increasingly being considered
nonpoint source water pollution mitigation, we em- more intentionally by policy and rulemakers (Booher
ploy an agent-based modeling approach to test specific 2004; Emerson and Nabatchi 2015; Koontz and
design elements found in collaborative governance the- Johnson 2004; Pahl-Wostl et al. 2007; Scott 2015).
ories, and inform actual policy design in the process. Collaborative governance can be realized as a form of
This manuscript highlights how an agent-based model institutional design, leading to the question: “How best
(ABM) constructed to simulate alternative institutional to design an effective and usable collaborative govern-
design configurations is used to test aspects of collabora- ance platform to mitigate environmental pollution?”
tive governance theory and inform public policy. Taking Emerson and Nabatchi (2015, 18) define collabora-
an ABM approach to modeling different institutional tive governance as “the process and structures of public
designs allows us to estimate performance outcomes (in policy decision-making and management that engage
the form of reduced phosphorus loads) relative to dif- people across boundaries of public agencies, levels of
ferent design parameters. These performance outcome government, and/or the public, private and civic spheres
estimates are determined on a regional as opposed to to carry out a public purpose that could not otherwise
local scale. With the ability to manipulate the collabora- be accomplished.” Theories of collaborative governance
tive dynamics of simulated actors (in this case municipal focus on the critical inputs, processes, and outcomes as-
and state government actors) through generating dif- sociated with effective collaborative governance (Ansell
ferent collaborative (and non-collaborative) pathways, and Gash 2008), the internal dynamics of collaborative
the ABM platform enables the aggregation of individual actors and nature of collaborative life cycles (Emerson
actor decision criteria to larger scales of operation, al- and Nabatchi 2015), and the critical challenges and ex-
lowing for the emergence of variations in outcomes that ogenous factors that support or hinder effective collab-
are contingent upon different governance designs. orative governance arrangements (Bryson, Crosby, and
The ABM highlighted here integrates spatial context Stone 2015). Collaborative governance is said to be ini-
and dynamics of the collaborative water governance tiated and evolved within a “multi-layered system con-
system being employed and considered for the Lake text,” which includes resource conditions, policy and
Champlain Basin (LCB) in Vermont, United States. legal frameworks, power relations, and network char-
This water governance model and the framing of initial acteristics (Bryson, Crosby, and Stone 2015, 27)
scenarios were informed by stakeholder input, including The key premise of collaborative governance re-
consultation with key policymakers in the LCB as part gime theory is the notion that principled engagement
Journal of Public Administration Research and Theory, 2020, Vol. XX, No. XX 3
“will produce determinations that are fairer and network actors. Network governance hinges on the ef-
more durable, robust, and efficacious” (Emerson and ficacy of resource exchanges (Rhodes 1997) that lead to
Nabatchi 2015), and we may infer, lead to better per- network outputs and performance (Turrini et al. 2009).
formance outcomes. Most collaborative arrangements The multi-scale, multiplex water governance net-
call for the intentional use of institutional design work in the LCB contains a large number of actors who
(Emerson and Nabatchi 2015, 69; see also Ansell and possess multiple objectives and engage with one an-
Gash 2018; Koliba et al. 2018; Ostrom 1990) and the other through different types of ties (Koliba et al. 2014;
utilization and exchange of resources (Emerson and Scheinert et al. 2015). The State of Vermont has flexi-

Nabatchi 2015, 73; see also Rhodes 1997; Emerson bility in the development and application of policies to
and Gerlak 2014; Koliba et al. 2018). Across the col- meet Total Maximum Daily Load (TMDL) targets set
laborative governance literature, the role of networks by the US Environmental Protection Agency (USEPA).
as the dominant structure shaping institutional rules In this framework of “pragmatic federalism” (Gerlak
and resource exchange is fairly consistent (Bryson, 2005), the state has considerable control over the shape
Crosby, and Stone 2015; Emerson and Nabatchi of the collaborative governance response by choosing
2015; Keast and Mandell 2014; Koliba et al. 2018). how it codifies requirements, structures (dis)incentives,
Both collaborative governance (Ansell and Gash and facilitates coordination. Figure 1 illustrates pos-
2008; Emerson and Nabatchi 2015) and network gov- sible alternative configurations of the water governance
ernance theories (Provan and Kenis 2007) project typ- network. In figure 1A, we show a traditional top-down,
ologies relating to idealized states. Both collaborative “non-collaborative” governance structure, where
and network governance arrangements may be exter- the State generates policy tools to directly manage or
nally initiated and led, independently convened and incent municipal behavior through mandates, grants,
network administratively organized, or self-initiated and matching funds. The edges in the network are
and governed through shared leadership arrangements. formed by the legal requirements for clean water ac-
Collaborative governance regime theory tends to em- tions flowing from the State (S) to all municipalities
phasize how collaborative networks emerge, whereas (M), forming the basis of a typical hub and spoke or
network governance theory stresses the relatively stable star network. This configuration represents the cur-
structures that exist to govern goal-directed networks rent business-as-usual arrangement that involves an,
(Provan and Kenis 2007; Milward and Provan 2006). essentially, formal, non-collaborative governance ar-
The extent to which a particular arrangement fits a given rangement. Vermont does not have strong county-level
context is a question that many theorists and researchers governance, making figure 1A the norm. In figure 1B,
in this space have asked (Emerson and Nabatchi 2015; an alternative network structure describes regional
Keast and Mandell 2014; Milward and Provan 2006). water districts (D) that coordinates the involvement of
One of the common critiques of collaborative gov- municipalities and serves as an intermediary for regu-
ernance theory is the “black box” treatment of net- lation and funding streams. In this instance, we can im-
work structures in these models (Koliba et al. 2018), agine a legal framework that mandates municipalities
despite the fact that network characteristics are viewed to participate in districts that collectively manage water
as being an essential feature of collaborative govern- quality into the regional jurisdictions, a scenario that
ance regimes (Emerson and Nabatchi 2015; Keast and was eventually enacted by the state. Within a district,
Mandell 2014). To model the structures of collabora- total phosphorus (TP) mitigation efforts could be pri-
tive governance regimes, we need a more fully expli- oritized at a regional scale, whereas municipal planning
cated model of network structures to empirically track and implementation capacities could be pooled at the
and describe resource flows over time and between or- regional scale. Funding mechanisms (e.g., block grants)
ganizational actors. could further take advantage of the network to reduce
In network governance theory, the unit of analysis the administrative burden on the state government
is the interorganizational governance network. The and on specific municipalities. Such a configuration
network is multi-layered (or multi-level) and is com- may create new inefficiencies in distributed govern-
prised of actors that can include public, private, and ance and in principal-agent relationships. However, it
nonprofit actors (Koliba et al. 2018; Rhodes 1997). may also surface local knowledge and develop social
Network architecture allows for the explicit identifica- capital to better facilitate project feasibility. This scen-
tion of actors as being linked or tied to other network ario best aligns with the externally directed type of col-
actors. The functions carried out by the whole network, laborative governance regime outlined by Emerson and
sub-network, or network cliques can include formal or Nabatchi (2015). Finally, in figure 1C we see a hybrid
informal processes of principled engagement, deliber- of A & B, with some districts in place, but direct state-
ation, and decision-making. In essence, processes of col- municipal interactions also present. Such a structure
laboration unfold as interactions or exchanges between may indicate a voluntary regionalization policy based

Figure 1. Alternative Network Configurations of Water Governance. In panel A, the state agency directly communicates requirements and
funding to municipalities. In panel B, regional water districts act as intermediaries between state and municipal actors. In panel C, the
adoption of regional districts is incomplete, and the state must manage relationships with both districts and municipalities.
on a “self-initiated” collaborative governance regime To better understand how and to what extent col-
(Emerson and Nabatchi 2015), in which the propensity laborative governance arrangements work best and
of a municipality to voluntarily collaborate with others under what conditions, computer simulation modeling
at the regional scale can vary. is increasingly being used to test and tune theory
Across these three scenarios, the State (S) serves as (Johnson 1999; Schlüter et al. 2017). Recently, the no-
the central actor responsible for the distribution of tion of “collaborative platforms” has been advanced
resources across the network. As the network actor to answer the question of “how can collaborative
with the highest degree centrality, the resources of the governance be purposefully extended and scale-up?”
state, both in terms of financial and human capacity, (Ansell and Gash 2018, 16). Ansell and Gash propose
are critical to the overall performance of the network, the collaborative platform as a “generic organizational
regardless of its configuration. Municipalities (M) rely logic” (Ansell and Gash 2018, 17) or as a strategy for
on state funds to pay in part for the planning and im- societal problem-solving (Nambisan and Baron 2009).
plementation of stormwater projects. The two cap- These platforms can serve one of three different func-
acities of human and financial capital are very often tions: to explore and frame the nature of problems,
coupled—as it takes human resources to process re- to experiment with potential solutions, and then exe-
quests for funding, and limitations on funding will cute specific solutions using collaborative governance
predicate the number of projects that may be planned frameworks (Ansell and Gash 2018; Nambisan and
and enacted. In essence, the resources available to the Baron 2009). In promoting the concept of collaborative
central actor in the collaborative governance network platforms, Ansell and Gash observe that “more needs
configurations studied here likely matters a great deal to be done to translate the logic of platforms from the
to the performance of the whole water governance net- domain of technology, software development and even
work. The degree to which the state’s human resource organization theory to that of governance and public
capacity limits or enables the flow of funding through administration” (Ansell and Gash 2018, 17). The use
each type of network is likely to matter a great deal. of computer simulation models to experiment with
alternative collaborative designs is increasing, particu- include that same group as well as private contractors.
larly in the experimental stages of collaborative devel- We expect that water districts that regionalize both
opment in which stakeholders may prototype and test project planning and implementation will be more ef-
the efficacy of these governance arrangements against fective than districts that only regionalize planning. In
performance targets. Although Ansell and Gash’s in- other words, extending the value proposition of col-
terpretation of collaborative platforms extends well laboration from an “end-to-end” perspective is likely
beyond the prototyping phase and into the realms of to better amplify the positive impacts of collaborative
actual implementation, we focus here on one example governance on performance.

of constructing and using a computer simulation H3: The Mandated Collaboration Hypothesis.
model to test the efficacy of specific collaborative gov- Mandated collaborations for planning, and for plan-
ernance arrangements. As we will note, stakeholder en- ning and implementation will generate higher nutrient
gagement around the design of the model and the use load reductions than voluntary collaboration for plan-
of model output to initialize new collaborative govern- ning, and for planning and implementation. The results
ance arrangements (e.g., moving from experimentation of studies of mandated, externally directed collabor-
to execution) is discussed. Borrowing from Ansell and ation have provided mixed results relative to the value
Gash’s language, this experimental collaborative plat- and efficacy of mandated collaboration. A comparative
form can test a series of alternative design rules. case study of mandated health care delivery networks
To better understand which design rules are more by Rodriguez et al. (2007) found the mandated collab-
or less likely to be effective, we formulate a series of orative approach failed to enhance interorganizational
hypotheses to test in our model. relationships, particularly in cases when there is a lack
H1: The Collaborative Governance Hypothesis: of multiple governance mechanisms to enable well-
Those alternative water governance arrangements that resourced coordination. While Brummel et al.’s (2012)
are premised on collaborative governance designs will study of wildland fire planning found that mandated
generate higher nutrient load reductions than those collaboration among service providers provided more
water governance arrangements that are not prem- consistent and higher quality services because of the
ised on collaborative designs. Findings from prior ability to allow for flexibility and resource sharing. It
studies of collective action problems for water gov- is clear that the specific design of the mandated col-
ernance have found that greater coordination among laboration is critically important. We anticipate that
governance actors leads to more effective outcomes, mandated collaboration will allow for lower capacity
underscoring the assumptions of collaborative gov- municipalities responsible for higher nutrient loads
ernance regime theory that collaborative governance to receive aid in their phosphorus mitigation efforts.
leads to better outcomes (Koontz and Johnson 2004; thereby increasing the impact of mandated collabor-
Scott 2015). However, other studies have found that ation arrangements. The challenges associated with the
collaboration can sometimes reinforce existing power limited capacity of particularly smaller municipalities
relations (Scott and Thomas 2017) and lead to reduced are a clear concern to emerge during discussions with
effectiveness. Our first hypothesis centers on whether local municipal leaders.
collaboration leads to greater phosphorus mitigation. H4: The Administrative Bottleneck Hypothesis.
We postulate that greater rates of municipal collabor- The administrative capacity of the State (S) will be a
ation in water districts will lead to greater phosphorus significant factor in load reductions results of all con-
mitigation. figurations of water governance design considered
H2: The Collaboration for Planning and here. Drawing on qualitative assertions made by pol-
Implementation Hypothesis (H2). Collaboration in icymakers operating in the region, as well as the logic
both the planning and implementation phases of pro- of coupling of financial and human resources, the ad-
ject development will result in higher nutrient load ministrative capacity of the State (S) to evaluate water
reductions than those networks that only collaborate quality projects operates as a constraint on the alloca-
around the planning of projects. Our second hypoth- tion of resource. This assertion is supported by prior
esis focuses on the functional goals of collaboration. studies of regional governance that have found that the
In the realm of municipal stormwater management, flow of financial resources to infrastructure projects is
clean water projects, like all engineering or capital hindered by a lack of administrative capacity to scope,
improvement projects, must follow a clear planning- design, and implement projects (Milio 2007).
to-implementation, “end-to-end” process. Planning
phases include the scoping and design of potential pro- Case Study: Stormwater Project Planning and
jects and involve the coordination of engineering firms, Implementation in the Lake Champlain Basin
public works departments, and planning offices. While The wicked problem of nonpoint pollution provides a
the implementation phase of stormwater projects can context to study the relationship between alternative
institutional designs and specific outcomes, in this water quality problem in Lake Champlain and many
case, reductions in nutrient loads. In the LCB in the other freshwater lakes all over the world. HABs close
northeastern United States, challenges of nutrient pol- beaches, threaten public health, negatively impact re-
lution and harmful cyanobacteria blooms in Lake gional economies, and reduce property values (USEPA
Champlain are met by a variety of cooperating and 2018). In Lake Champlain, HABs occur most fre-
competing governance actors that operate at federal, quently during late summer in shallow bays where cli-
state, regional, municipal, and even international levels mate has a greater influence on nutrient mixing and
(figure 2B). Collectively, these actors comprise multi- bacteria have easier access to nutrients in lake sedi-

scale, multiplex water governance networks (Koliba ment (Isles et al. 2017; Zia et al. 2016). Blooms in Lake
et al. 2018; Lubell, Robins, and Wang 2014; Scheinert Champlain are fed by nutrient nonpoint runoff (pri-
et al. 2015; 2017). It is well-recognized that the multi- marily phosphorus) from multiple sources, including
and-cross-scale nature of environmental problems— agriculture, urban stormwater, streambank erosion,
including water-related issues of quantity, quality, and and forested land use (Lake Champlain Basin Program
the degradation of concomitant ecosystem services— 2018). Each nutrient source represents a potential
requires governance approaches that themselves span target for policy tools aimed (directly or indirectly)
similar scales (Bodin, Sandström, and Crona 2017; at reducing nutrient runoff, and ultimately bloom se-
Hamilton and Lubell 2018; Koontz et al. 2010; Pahl- verity and frequency.
Wostl et al. 2010; Scott and Thomas 2017). Yet, under- A TMDL regulation, put in place by the USEPA, cur-
standing how the actors in water governance networks rently regulates the maximum amount of phosphorus
might react and adapt to novel policy or environmental that is allowed to reach Lake Champlain and still meet
pressures (e.g., regulations, global climate change) re- water quality standards. The amount of phosphorus
quires dynamic modeling techniques that integrate reaching a waterbody is termed “load” and is meas-
geographic context, spatial and temporal lags, and the ured in units of mass per time (e.g., kg/day, Mt/year)
complex structure of the social-ecological network. (figure 2A). Thus, the TMDL does not directly regu-
Harmful cyanobacteria blooms (sometimes grouped late the presence of HABs in Lake Champlain, but ra-
with harmful algal blooms, or HABs) are a significant ther one of the main causal inputs (Koliba et al. 2014).
Figure 2. One View of the LCB Water Governance System. Panel A shows the spatial distribution of TP loads at the NHDPlus catchment
scale. Panel B shows the mismatch between hydrology (broadly represented by HUC-8 watersheds), municipal jurisdictions, and the
modeled regions to which municipalities are assigned.
USEPA used a physics-based SWAT (Soil and Water sustainability in complex social-ecological systems
Assessment Tool) model (Arnold and Fohrer 2005) (Bitterman and Bennett 2018). For example, ABMs
combining land use, hydrological, and erosional pro- have been used to study land use change in the Yucatán
cesses to estimate the contribution of various land peninsula (Manson 2005), the movement of elk as af-
uses in a spatially explicit manner across the LCB fected by land management practices in Yellowstone
(figure 2A). National Park (Bennett and McGinnis 2008), the
On June 15, 2015, Act 64, colloquially the “Vermont transmission of disease across livestock production
Clean Water Act,” was signed into law to address chains (Wiltshire et al. 2019), and in common pool

degrading water quality in Lake Champlain and its resource management situations (Deadman 1999;
tributaries. The law set forth new rules for managing Schlüter and Pahl-Wostl 2007). ABMs have been expli-
water quality in the state, including restrictions on citly used to model alternative institutional designs in
agricultural practices and permitting processes for ex- other settings including models of hazardous waste re-
isting and new development. It created the Clean Water mediation (Eckerd, Campbell, and Kim 2012), school
Fund, a pool of financial resources to be distributed by district governance (Maroulis 2016), the diffusion of
the state to fund projects related to clean water goals. fraudulent claims across service delivery networks
Subsequently, on June 6, 2018, Act 76 was signed into (Kim 2007), and the prioritization of transportation
law outlining the expectations to design and implement projects (Zia and Koliba 2013).
a system through which every watershed in the LCB In their work on Balinese water temples, Lansing and
be assigned a Clean Water Service Provider (CWSP) Kremer (1993) demonstrated how simulation models
to provide regional coordination of non-regulatory could be used to understand the relationship between
water quality improvement projects (Vermont General water resources and collaborative governance. Their
Assembly 2019). These new CWSPs would coordinate work showed how self-organization emerged in spe-
actions to collectively develop clean water priorities cific collaborative governance arrangements. Building
and to apply for state funding. While implementa- on this legacy, a suite of models and research programs
tion details are not yet finalized, this collaborative have emerged at this nexus of water, land use, col-
governance arrangement is in part due to the findings lective decision-making, and governance. Schluter and
stemming from the sharing of these modeling results. Pahl-Wostl (2007) developed an ABM to simulate the
resilience of different water-management institutions
Agent-Based Modeling of Water Governance to changes in environmental conditions in a semi-arid
ABMs have been proven to be useful tools as experi- Amu Darya river basin in Central Asia. Smajgl et al.
mental collaborative platforms that can help policy- (2009) used an ABM to simulate water trading mar-
makers make informed decisions (Ligmann-Zielinska kets and alternative market configurations in Australia,
and Jankowski 2007). ABMs are a “bottom-up” whereas Bellaubi and Pahl-Wostl (2017) modeled cor-
method of simulating the interactions among heter- ruption in a water management system in Kenya and
ogenous actors and their environment, commonly in Ghana. While governance actors can be modeled as
a spatially explicit manner. Agent-based modeling has agents, their internal dynamics can have substantial
been called “the only technique available today to for- influence on collective and individual decision-making
malize models based on micro-foundations, such as as well, as illustrated by the ABM in Watkins et al.
agents’ beliefs and behavior and social interactions, all (2013). Recent communications behavior-focused re-
aspects that we know are of a certain importance to search has also shown how ABMs can be used to test
understand macro outcomes” (Squazzoni and Boero behavioral theories (Janssen and Baggio 2017; Bucini
2010). ABMs often reflect the dynamics in real-world et al. 2019) and how different policy-targeting can im-
systems, including how system structures change due pact flood mitigation and coping (Erdlenbruch and
to internal processes or outside disturbances (Batty Bonté 2018). In many cases, ABMs are calibrated to
et al. 2011). One strength of ABMs lies in their ability historical patterns (e.g., land use, migration) (Grimm
to simulate emergent system-scale dynamics from the 2005; Magliocca and Ellis 2013) such that the model
repeat interactions of individual agents on a landscape can be used to create scenarios of future system trajec-
(Parker et al. 2003). Through the application of rela- tories. When calibration targets are sparse or unavail-
tively simple rules that govern individual agent be- able, or when the subject of the study is novel, ABMs
havior, patterns of collective group behavior, including are commonly more “stylized” and used not for pre-
cooperation in environmental management, can diction, but to improve understanding of system dy-
emerge (Goldstone and Janssen 2005; Scott, Thomas, namics. In this study, the alternative water governance
and Magallanes 2018). This ability to link multi-scale regime embodied by districts in Vermont would be
actors, processes, and structures means ABMs are novel to this system. Thus our ABM is conceived as a
particularly well-suited to the study of resilience and stylized, exploratory model grounded in collaborative
governance theory and not subject to model valid- Additional data were collected through a survey that
ation expectations. We purposefully simplify agent went out to all municipalities across the State (Clark,
decision-making process and have not, for example, in- Hurley, and Koliba 2018).
cluded political or legal motivations of behavior. Given As detailed in figure 3, the coupled (connected)
these assumptions, the simulations explore model sen- model platform has three primary components: (1) a
sitivity to policy-relevant inputs. land use model that places municipal stormwater pro-
While the above models capture many of the key jects on the LCB landscape, (2) a simplified load accu-
dynamics that can be found in watershed governance, mulation model, and (3) the ABM of networked water

a gap remains in modeling the translation of policy governance. Communication among model compo-
objectives to policy tools in a manner that is both nents is facilitated by an infrastructure that tightly
spatially explicit and incorporates environmental feed- couples the models at an annual timescale, meaning
backs. Effective reduction of nonpoint pollution has that the land use, phosphorus load, and governance
proven to be extremely difficult. The extent to which models can generate outputs that can be aggregated
these difficulties stem from challenges associated with across specific and consistent scales of temporality.
the design and operation of explicit water governance Within the ABM, agent–agent interactions occur at
networks is the subject of this study. sub-annual intervals. Agent–environment interactions
occur when municipalities or districts react to the state
of the environment and make decisions that lead to the
Materials and Methods creation of stormwater projects, which are then instan-
An adaptive management perspective guided the devel- tiated on the simulated landscape. When implemented,
opment of this governance ABM (Norton 2005), with these projects differentially affect annualized phos-
a key feature of this approach being the pursuit of col- phorus loads that accumulate to the lake. The changes
laborative learning between social and natural scien- in loadings are read by the governance actor agents,
tists and policymakers and other stakeholders (Daniels affecting their behavior in future timesteps and closing
and Walker 2001). At the start of this project, the re- the environment-agent portion of the feedback loop. In
search team undertook a series of focus groups and addition to the modeling framework shown in figure 3,
interviews with key informants in the LCB, and source the Overview Design Concepts, and Details (ODD)
document analysis of major pieces of legislation, rules, model specification (Grimm et al. 2010) for the ABM
white papers, and TMDL memorandums of under- is included in the Supplementary Material. The ABM
standing. This phase of research was used to develop framework is built upon the MASON Multiagent
a qualitative appreciation of the multi-level govern- Simulation Toolkit (Luke et al. 2005) and its extension
ance networks operating in this region. A Policy and GeoMason, which provides additional geospatial sup-
Technical Advisory Committee (PTAC) met with the port capabilities.
research team at regular intervals to inform the model Together, the ABM, land use, and load accumulation
development. Observations about model assumptions, models trace the lifecycle of clean water projects as they
data sources, and intended uses of model outputs were proceed from an initial state of unknown/unplanned to
made, summarized and drawn on by the research team. a final state where the project is implemented on the
Figure 3. Generalized Schematic of the Coupled Model. The Coupler is a software architecture that facilitates communication among
connected models—in this instance the governance ABM, the land use model, and the phosphorus load accumulation model
landscape. When municipalities utilize their project- capacity corresponds to the number of projects they
planning capacity, they may coordinate with other can take from concept to a “shovel ready” state that is
actors to prioritize clean water projects according to ready to be funded. The distribution of municipal plan-
their individual characteristics (e.g., available capacity) ning and implementation capacity parameters were es-
and rules put in place by policy. The state agency agent timated using empirical data from the VTANR CWIP
evaluates and prioritizes projects, allocating funds and database. We estimated each municipality’s implemen-
utilizing its state-level capacity. Finally, agents imple- tation capacity by calculating the annual mean number
ment projects “on the ground,” affecting land use and of projects it completed over the last 4 years (the length

fulfilling the lifecycle of a project. Once agents imple- of the dataset). As data for planned, but not funded,
ment clean water projects, these changes to the land- projects are currently unavailable, we assume that
scape are then translated to the environmental model, planning capacity is twice that of implementation cap-
affecting the amount of phosphorus reaching Lake acity. Additional parameter estimation details can be
Champlain. found in the supplemental ODD protocol. Municipal
The ABM simulates the interactions among three agents randomly select among unplanned projects in
agent types: municipalities (i.e., towns and cities), re- their boundaries, representing the (at times) disordered
gional water district agents, and state agencies in the process of managing multiple priorities from various
LCB. Each municipal agent represents one of the 126 constituencies and interest groups. Agents then use
Vermont towns or cities whose centroid falls within their planning capacity to “discover” properties of the
the Vermont portion of the LCB. Eight regional water project (e.g., estimated load reduction potential, esti-
district agents each correspond to actual regional plan- mated implementation cost, location) in a simulated
ning district (RPC) boundaries, with each municipal planning process. Municipal agents then rank planned
agent assigned to its corresponding geographic region. projects according to prioritization criteria. We follow
Due to our focus on the CWIP process, we simplify a simplifying assumption that municipal agents seek to
interactions among multiple government agencies by maximize utility by prioritizing projects that provide
using a singular state agency agent that is tasked with the greatest load reductions per dollar. All municipal
improving water quality and managing the allocation agents follow this institutional rule logic, maximizing
of public funds. Within each annual timestep, muni- project efficiency as defined in equation (1).
cipal agents may: (1) plan clean water projects, (2)
municipal
prioritize projects, and (3) implement projects. When
undertaking each of these steps, real-world muni- capacity
(1) est. phos. reductioni
cipalities are subject to dynamic budget and staffing max
est. implementation costsi
constraints, regulatory requirements, and public will. i= 1
We have simplified municipal decision-making and re- Once projects are planned, they are passed to the state
source constraints to five key municipal variables and agency agent to be ranked and funded (see below),
functions (table 1). then passed back to their corresponding municipality.
The objective of a municipal agent is to reduce Unfunded projects are queued, and funded projects are
the amount of phosphorus load generated by land available for implementation. Municipal agents then
use within its borders to the target level specified by rank funded projects according to equation (1), again
the “reasonable assurance” scenario generated by the optimizing for load reductions per dollar. The top n
USEPA during the TMDL process (US Environmental projects are then implemented, where n corresponds to
Protection Agency 2016). A municipal agent’s planning an agent’s implementation capacity. Projects that are
Table 1. Variables and Functions Governing Behavior of Municipal Agents

Variable Description Range
Planning capacity The number of clean water projects a municipality can plan in a year. 2–14
Implementation The number of clean water projects a municipality can implement in a year. 1–6
capacity
TMDL load Target value of phosphorus load reduction for urban land use in the municipality to meet 15.4–1,179.7
reduction target TMDL. Estimates generated by EPA. (kg/year)
Probability of The probability that a municipality will collaborate with other municipalities in regional water 0–1
collaboration districts. Empirical distributions generated using survey data of municipalities in the Lake
Champlain Basin.
Prioritization criteria The algorithm used by municipal agents to rank projects prior to submitting them for funding NA
and before implementation.
funded but not implemented are returned to the cor- criteria from equation (1). This agent’s capacity to
responding municipal agent’s implementation queue. evaluate projects is a scenario-driven parameter and
Once municipal agents have met their phosphorus re- corresponds to its throughput, or the number of pro-
duction targets, they stop all activity. jects it can evaluate in an annual timestep. Finally, the
In some scenarios (described below), state policy state agent sets the rules of the CWIP action arena,
is altered to mandate collaboration among muni- thereby affecting the allowed behavior of municipal
cipal agents at a regional scale. Stakeholder anecdotes and regional agents (table 3). The final class of agents
drawn from focus groups suggest that regional co- is regional facilitators that, depending on policy scen-

operation varies greatly among regions and their con- ario, coordinate the planning and implementation of
stituent municipalities. This is confirmed by a 2017 municipal agents at broader scales. We model these
survey of Vermont municipalities (Clark, Hurley, and agents on RPCs and utilities, but the framework can
Koliba 2018) that drew on responses to determine the be scaled to other geographic extents (e.g., watershed
probability that a municipality will collaborate with boundaries) or network configurations.
their regional planning commission (RPC) along four The land use model provides a bridge between gov-
dimensions: (1) information sharing, (2) technical as- ernance actions and water quality impacts. This con-
sistance, (3) sharing human/physical assets, and (4) nection takes the form of the clean water projects that
receipt of monetary support. This survey was sent to are planned, prioritized, funded, and implemented by
all 249 municipalities in Vermont, with a response rate municipalities in the LCB. While each municipality is
of 55% across Vermont, and 64% of those munici- itself a governance actor in the ABM, each is also con-
palities in the LCB. Each municipality’s response was nected to a spatial feature corresponding to its muni-
coded 1/0 for yes/no, then summed to estimate a “co- cipal boundaries. Empirical data from VTANR were
ordination capacity score” for each municipality. An used to generate a realistic set of possible projects for
empirical distribution was generated for each region. the simulation. Additional detail can be found in S7.2
At the state scale, these scores follow a normal distri- of the Supplementary Appendix.
bution, though distributions vary at the regional scale. The environmental model is a simplified load accu-
Accordingly, we estimate the distribution for each re- mulation model based on an EPA-created SWAT model
gion in the simulation and normalize to the interval (Arnold and Fohrer 2005; Gassman et al. 2007) that
[0,1]. In those scenarios where policies allow for re- calculated average annual phosphorus contributions
gional coordination, we sample from these distribu- to Lake Champlain (US Environmental Protection
tions and assign a probability of collaboration for each Agency 2016). As projects are implemented on the
municipality. Mean values are shown in table 2. landscape, the contribution of phosphorus to the lake
At the state level, the responsibility of water policy- is reduced in a spatially explicit manner. This simplified
rulemaking primarily falls on the state Vermont model excludes nutrient transport and the role of cli-
Division of Environmental Conservation (DEC). mate and assumes all phosphorus eventually reaches
Accordingly, we model a second agent type repre- the lake. The resultant model is a simplified represen-
senting a state agency tasked with managing water tation of nutrient export from the landscape (S7.1 of
quality. This singleton agency agent takes allocated the Supplementary Appendix). However, the relative
funds, evaluates municipal projects, and ranks funding lack of complexity reduces computational overhead
priorities as specified by policy scenarios. The state while isolating the effects of governance dynamics on
agent prioritizes projects using the same efficiency the landscape.
Through collaboration with policymakers at state,
Table 2. Mean of Probability Distribution for Each
regional, and municipal scales, we identified a set of
Municipality in a Region scenarios that included alternative policy configur-
ations, state-level capacity, and funding allocations to
Region Mean explore via simulation. The scenarios that alter policy
Chittenden 0.671 rules (table 3) modify the degree and geographic scale of
Addison 0.585 cooperation among municipal and regional water dis-
Central 0.568 trict agents. In the current state of this social-ecological
Rutland 0.545 system, municipalities are statutorily required to in-
Lamoille 0.521 dividually meet regulatory and non-regulatory load
Bennington 0.500 reductions and must participate in clean water im-
Northwest 0.429 provements, but have a choice as to whether work dir-
Northeastern 0.298 ectly with the state (figure 1A) or engaged in mandated
Note: Regions derived from Regional Planning Commission collaboration with other municipalities in their region
(RPC) jurisdictions in the LCB. to form a district (figure 1B). This second scenario also
Table 3. Scenario Parameters Include Coordination Policies (District Function), Resource Levels (Funding and
Capacity), and Landscape Configuration
Resource Parameters
Parameter Values
Allocated funds 1, 2, 3, 4, 5, 6, and 7 million USD
State agent capacity (throughput) 50, 75, 100, 125, 150, 200 and 225 projects/year
Initial landscape configuration 5 alternative distributions of projects on landscape

Coordination policy 1. No district
2. Voluntary planning district
3. Mandated planning district
4. Voluntary planning and implementation district
5. Mandated planning and implementation district
Coordination policies
Water district function Municipal aggregation rule Aggregated planning? Aggregated
implementation?
No district none no no
Voluntary planning district probability-based optional no
Mandated planning district all in region yes no
Voluntary implementation and planning district probability-based optional optional
Mandated implementation and planning district all in region yes yes
Note: Each coordination policy is composed of rules that govern municipal and regional behavior.
regionalizes planning capacity but imagines a policy capacity to implement projects, the funds sit unused.
tool that mandates municipal participation (prob- In our scenarios, we alter the throughput of the state
ability of coordination = 1). A third scenario is posited agent and the overall fund allocation to explore these
in which municipal agents have chosen to voluntarily dynamics. The parameters are based on past project
cooperate with others in the region or not (figure 1C). funding data and were developed in consultation
The voluntary participation of municipalities is prob- with agency stakeholders. In total, we simulate 1,225
abilistic, according to table 2. The fourth and fifth scenarios. Model stochasticity is introduced in (1) the
scenarios extend the previous scenarios to also include generation of projects, (2) the selection of projects by
collaboration for project implementation. Again, vol- municipalities to plan, (3) municipal participation in
untary and mandated policies allow for different levels regional cooperation, and (4) the initialization of mu-
of participation. These scenarios approximate a re- nicipal capacity. A simulation runs for 50 model years,
gional utility authorized to regulate the municipalities and each scenario was repeated for 30 Monte Carlo
in its jurisdiction. In some scenarios, collaboration is iterations.
mandated by the state agent, and regional facilitators
manage all municipal planning and implementation in
their jurisdiction. Results
In addition to alternative policy scenarios, we Hypothesis 1, the Collaborative Governance
modify the capacity (financial and human resources) Hypothesis, and hypothesis 2, the Collaboration for
of the state agency agent to evaluate and allocate funds Planning and Implementation Hypothesis, are closely
to stormwater projects (table 3). The State is tasked related. We expect collaborative policies to perform
with evaluating and prioritizing the projects identified better than non-collaborative policies, and we further
by municipalities. Administrative capacity bottlenecks expect that collaboration across several stages of pro-
may emerge at the project evaluation stage because the ject planning and implementation will perform better
State retains approval of all projects. A constrained ad- than partial efforts. We first compare the overall effi-
ministrative capacity can limit the throughput of the cacy of the different collaborative governance design
approval and funding to municipalities leading to re- scenarios by comparing the cumulative load reductions
duced load reductions on the landscape. Further, as for each scenario over the full simulation period. As
Acts 64 and 76 are implemented and as state, regional, shown in figure 4, the most effective (highest cumu-
and local capacities come online, there is uncertainty lative reduction) design was the creation of mandated
in the amount of funds that should be optimally allo- planning and implementation districts. The voluntary
cated to the problem. At a most basic level, if funding planning and implementation districts perform nearly
is low, loads will likely remain high. However, if too as well, followed by act alone, mandated planning
much money is allocated and there is insufficient (only), and voluntary planning (only) districts in that
order. The difference among outcomes was confirmed greater load reductions, specifically when resources are
to be significant by a Kruskal-Wallis test (p < .001). relatively high, and both planning and implementa-
The clustering of outcomes shown in figure 5A is a re- tion capacity are shared. Additional analysis confirms
sult of multiple scenarios with similar levels of funding three-way interactions among policy, funding, and cap-
and state-level capacity. However, statistically signifi- acity scenario parameters (table 4).
cant differences (p < .001) among collaborative gov- The mixed findings of H1 are relevant for evaluating
ernance scenarios hold when controlling for funding the Collaboration for Planning and Implementation
and capacity. In figures 4B and 4C, we see how differ- Hypothesis. Our findings very clearly prove H2—the

ences in policy efficacy change as funding and capacity coupling of collaboration for planning and implementa-
are increased. There is little difference among policy tion—leads to more efficacious outcomes. In our simu-
outcomes at low resource levels, whereas disparities lations, municipal water districts that only collaborate
become evident as resources increase. around planning functions are the least effective (as
While these findings partially confirm H1, our re- measured by cumulative load reductions) collaborative
sults are mixed with respect to the lower efficacy in governance scenarios. These results are explained by a
policies that incentivize regionalized planning schemes. spatial mismatch between planning and implementa-
Collaborative arrangements only sometimes lead to tion capacities caused by the newly enabled structure
Figure 4. Box-and-Whisker Plots of Cumulative Load Reductions. Panel A shows all scenarios by policy, obscuring the differences among
policies at various funding and capacity levels. Panel B shows little differences among policies at low capacity and funding levels, whereas
panel C shows that at higher levels of funding and capacity, reductions are greater in policies where both planning and implementation
capacity are aggregated.
Figure 5. Load Reductions as a Function of Policy, Funding, and Capacity Scenarios. Response is generally nonlinear, as the amount of
funding has little effect in lower capacity scenarios.
of the water districts. Because planning capacity is re- the implementation stage. We measure the depth of this
gionalized at the district level, but implementation cap- backlog using the average number of projects that have
acity is not, projects can be identified and planned in been planned and funded—but not implemented—for
areas where there is insufficient ability to implement each collaborative governance scenario.
them. Essentially, municipalities can become “over- The municipal-scale bottleneck becomes increas-
planned” when the number of to-be-implemented pro- ingly apparent in scenarios with greater state-level
jects exceeds a municipality’s implementation capacity. capacity (table 5). As the state processes additional
The regionalization policy creates a new bottleneck at projects, municipal capacity is exhausted, and the
implementation backlog increases. The backlog of de- we performed the non-parametric Kruskal-Wallis ana-
layed projects is much larger (confirmed by analysis of lysis of variance of each group individually, reported
variance [ANOVA], p < .001) when planning capacity in table 6. This Table should be read as what percent
is regionalized separate from implementation capacity, more effective is the mandated policy in mitigating TP
resulting in lower efficiency in the planning-only re- loads than the voluntary policy? We find that in all
gionalization scenarios. The issue is exacerbated as cases, the mandated policy performs better among the
capacity (throughput) of the state agency agent in- low-capacity, high-loads municipalities (all compari-
creases. Here, planning capacity, implementation cap- sons significant at p = .01). The relative effectiveness

acity, and prioritized policy targets (greater required of the mandated policy does not always improve as
load reductions) are mismatched spatially and tempor- state-level resources increase. This is sensible given that
ally. Therefore, the regionalization hypothesis is sup- the prioritization schemes in both policies optimize for
ported only when that mismatch is reduced. load mitigation independent of resource constraints.
Hypothesis 3, the Mandated Collaboration To confirm our findings, we also compared policy man-
Hypothesis, tests how municipalities with low op- dates among municipal agents with low-capacity and
erational capacity but high TP loads are affected by low TP loads (not shown here). In this alternative case,
an externally initiated mandate to collaborate. To test the result was the opposite, as the voluntary policy
this case, we created a subset of municipal agents pos- performed better (Kruskal-Wallis test, p = .01). This
sessing both low capacity (implementation capacity ≤ is also as expected, as low-load areas are disfavored in
1) and the highest 20% of TP loads. Given the imple- a collaborative scheme, independent of local capacity.
mentation bottlenecks discussed above and illustrated Thus, those municipalities with relatively low TP loads
in table 5, we only tested differences between the man- can improve their local conditions by not collaborating
dated and voluntary planning-and-implementation under a voluntary policy regime.
policies. In the mandated policy, we expect that in Our fourth hypothesis, the Administrative
aggregate the mitigated loads will be greater, as mu- Bottleneck Hypothesis, tests whether the human re-
nicipal agents with greater capacity can “lend” their source capacity at the state scale (as indicated by
resources to low-capacity collaborators. Accordingly, the number of projects that State administrators can
Table 4. Analysis of Variance of Interactions Among Model Input Parameters

df Sum Sq. Mean Sq. F p (>F)
Policy 4 1.5e11 3.6e10 40,214.9 <.001
State-level capacity 1 3.7e11 3.7e11 413,945.0 <.001
Funding 1 1.2e10 1.2e10 13,168.5 <.001
Policy:Capacity 4 4.3e10 1.1e10 11,972.6 <.001
Policy:Funding 4 2.3e9 5.7e8 628.6 <.001
Capacity:Funding 1 1.4e10 1.4e10 15,584.4 <.001
Policy:Capacity:Funding 4 2.9e9 7.1e8 792.9 <.001
Table 5. Project Implementation Backlogs, Measured as the Mean Number of Projects that are Funded but Unable
to be Completed due to Constraints on Implementation Capacity
State Agency Evaluation Capacity
50 75 100 125 150 200 225
Policy Compete 1.24 (0.26) 1.89 2.95 3.90 5.51 10.33 12.88
scenario (0.52) (0.99) (1.50) (2.59) (6.91) (9.33)
Voluntary planning 13.00 19.09 23.42 25.57 27.44 33.62 37.41
(2.70) (4.59) (6.78) (8.21) (9.54) (14.32) (17.47)
Mandated planning 5.42 (1.64) 8.56 11.88 14.72 17.93 24.18 26.59
(2.50) (3.62) (4.86) (6.66) (10.66) (12.43)
Voluntary planning 1.18 (0.21) 1.73 2.73 3.80 5.03 9.24 (5.65) 12.03
and implementation (0.44) (0.87) (1.42) (2.12) (8.22)
Mandated planning 1 (0.00) 1 (0.00) 1 (0.00) 1 (0.00) 1 (0.00) 1 (0.00) 1
and implementation (0.00)
Note: SDs in parentheses. Differences among regionalization policy, state evaluation capacity, and funding levels (not shown here) confirmed
to be significant (p < .001).
process) constrains the efficacy of the program and the phosphorus mitigated. However, we also see in Panel B
allocation of funds. When measured using the cumula- that as capacity increases, the effectiveness of spending
tive load reductions funded by alternative collabora- (kg / $1,000) decreases. This is a result of the priori-
tive governance scenarios, we find this to be the case. tization process—as the governance network imple-
Interactions among policy, capacity, and funding were ments additional clean water projects, it necessarily
confirmed by the ANOVA results (table 5). We plot funds more marginal projects, shifting the distribution
these interactions among policy, capacity, funding, and and lowering return on investment.
load reductions in figure 5. Load reductions in all col-

laborative governance scenarios exhibit some degree
Discussion
of non-linear response, as indicated by the surfaces on
the left-hand side of the figure. In general, when the Our model findings generally support the supposition
state’s human resource capacity to evaluate projects that designing for collaborative governance leads to
is relatively low (less than 100 projects/year) the ef- better performance outcomes in the form of reduced
fects of additional funds are minimal. For example, in phosphorus loads, affirming H1. We also find that H2,
scenarios that model a voluntary planning district with the Collaboration for Planning and Implementation
a capacity of 100 projects/year, our model indicates Hypothesis to be affirmed, as well as the Mandated
that that increasing funding from 1 to 7 million dollars Collaboration Hypothesis (H3). These findings are
will result in little-to-no change in load reductions. As likely intuitive in cases where all capacities and object-
capacity increases, however, additional funds do lead ives are aggregated, and processes are optimized using
to greater TP load reductions. The locations of thresh- collaborative governance mechanisms and norms. In
olds differ by active policy but are generally found in these scenarios, the policy essentially structures the
the 100–150 projects/year range in this model. The collaborative governance regime as a scale-based linear
response surface for planning-only districts is fairly optimization that maximizes utility at a broader scale.
flat, indicating a less-responsive system. Planning-and- While a useful benchmark, this type of governance re-
implementation districts, on the other hand, respond quires the participation of all finer-scale entities (here,
more strongly to changes in capacity and clean water municipalities) that is only achieved through man-
funding. The “act alone” scenarios fall in between. The dated collaboration. These results also support the
relationships among load reductions, funding, and Administrative Bottleneck Hypothesis (H4) in which
state capacity support the bottleneck hypothesis. we see the value of matching the administrative cap-
Finally, we can also explore the suite of options acity of the State to levels of funding.
available to policymakers from the perspective of As the real-world policy reality deviates from a
overall “bottom-line” performance indicators. As we mathematically optimal configuration, the resultant
have shown, load reductions respond strongly to state- social and environmental outcomes are less predict-
level capacity. A likely question within VTANR may be able. For example, voluntary participation scenarios
“what is the right level of staffing (capacity), and how led to lower reductions in load reductions and less-
might we measure the effectiveness of our actions?” In efficient spending of funds. Further, new bottlenecks
figure 6, we plot two alternative metrics for planning- and project backlogs were created by planning-only
and-implementation districts against various levels of districts at the municipal level, again leading to less-
state capacity. In Panel A, we see that as human re- efficient performance. We can conclude, therefore, that
source capacity to evaluate projects increases, so does the system is not strictly dependent on scaling effects—
a performance metric based on the total kilograms of rather, collaborative design considerations of how and
Table 6. The Relative Performance of The Mandated Planning-and-Implementation Policy Over the Voluntary Policy
State-Level Capacity (Projects/Year)
50 75 100 125 150 200 225
Annual funding (in millions of dollars) 1 22.9 13.4 13.9 23.1 23.2 23.3 22.5
2 22.7 20.2 19.3 26.6 27.8 20.5 24.1
3 18.5 22.0 20.7 21.8 23.8 31.5 22.5
4 20.5 19.3 24.9 20.5 18.5 15.2 33.6
5 23.5 15.6 23.1 23.4 22.2 31.7 29.0
6 17.2 23.8 19.8 18.0 26.1 24.2 35.5
7 11.2 13.3 11.2 20.6 20.4 21.2 37.4
Note: Values in the table are percentages, and should be read, for example, as “the mandated policy on average mitigates 22.9% more load
than the voluntary policy”. All comparisons confirmed significant by a Kruskal-Wallis analysis of variance.

Figure 6. Metrics of Policy Efficacy for Planning-and-Implementation Districts. As capacity increases, amounts of mitigated phosphorus
increase, but per-dollar efficiency decreases.
where to regionalize governance capacity is important could lead to more efficient policy outcomes. While
in achieving goals as well. the assumptions and simplifications in the model limit
These findings demonstrate the complex trade-offs normative predictions, the model allows for the explor-
that can surface when collaborative governance mech- ation of alternative scenarios with stakeholder input.
anisms are employed. Collaboration for collaboration’s The ability of the water governance ABM to both
sake is not a panacea. Our model results demonstrate test hypotheses and provide some practical consid-
that the links between planning and implementation, erations to policymakers offers us a unique ability to
and the need for administrative capacity to address test variations of collaborative governance theory. This
bottlenecks are critically important considerations. study contributes to the growing body of research that
Our model indicates that the efficacy of instituting has demonstrated the efficacy of collaborative govern-
a collaborative governance approach to municipal ance arrangements to address wicked environmental
stormwater management is highly dependent on the problems (Booher 2004; Emerson and Nabatchi 2015;
capacity of the state agency agent to evaluate clean Koontz and Johnson 2004; Pahl-Wostl et al. 2007;
water projects and allocate funds. The parameter Scott 2015). This assertion is rendered at a coarse-grain
ranges for our scenarios are based on empirical data level, as found in Hypothesis 1. The variation of differ-
from the Vermont DEC clean water projects data- ences between collaborative and non-collaborative de-
base, which shows that approximately 57 urban land signs is measurably higher, but not always decidedly so.
use projects are funded annually (VTDEC 2019). Our The non-collaboration scenario illustrated as the star
simulations show that significant increases in state cap- network found in figure 1A provides an opportunity
acity will have greater immediate impacts than short- for municipalities to directly access resources from the
term increases in funding. As shown in figure 6, system state. The lack of any intermediaries that may provide
response to an increase in capacity is non-linear. Across buffers against the state may place greater burdens
all modeled scenarios, at lower capacity levels (100 on municipalities to “fend for themselves.” For muni-
projects/year or less), funding can increase by 700% cipalities with higher capacity to plan and implement
and the response of the policy target (i.e., load reduc- stormwater projects, these arrangements provide them
tions) remains flat, as the system cannot utilize all of greater ability to comply with water quality standards.
the funding from the state. Though increased capacity But as we have noted, the problem locations or drivers
and funding improve the state’s ability to process pro- of nonpoint source pollution are not evenly distributed,
jects, this model holds municipal (and regional) cap- driving the potential for projects with lower returns
acity constant. Thus, additional resources eventually on investment to get implemented over other projects.
shift a bottleneck from the state agent to the munici- Pooling resources and aggregating project planning
palities (or to their regionalized districts). The values and implementation at the district scale should lead
in figure 6 heatmaps point to possible “sweet spots,” to higher net load reductions. However, this was not
or combinations of capacity and funding levels that the case for either voluntary collaboration scenarios. In
this case, voluntary collaborative governance regimes Although we cannot anticipate or model all possible
for planning alone and planning and implementa- futures or conditions, the examination of various con-
tion both scored lower mean load reductions than the figurations of collaborative governance platforms can
non-collaborative scenario. We should judge this par- increase learning and adaptive capacity (Ansell and
ticular outcome not as a generalizable outcome, but Gash 2018; Daniels and Walker 2001; Emerson and
as a property of the particular relationship between a Gerlak 2014). As we have seen, not all collaborative
municipality’s propensity to collaborate and the geo- governance schemes may lead to desirable outcomes.
spatial distribution of nonpoint pollution. This finding Thus, the value proposition of collaborative govern-

should rather be explained as a specific set of contin- ance is very likely contingent upon the specific design
gencies shaped by municipal willingness to collaborate of the collaboration network.
and the geophysical properties of the environmental
problem. By incentivizing or motivating municipal-
ities to increase their propensity to collaborate, we Conclusion
may find that load reduction measures would increase. With the nuances of modeling complex governance
Mandating collaboration for the planning and imple- arrangements recognized, there are a number of key
mentation of stormwater projects leads to generally observations to make relative to some of the core ques-
higher levels of load reductions because it allows for tions that concern public managers and policymakers.
a stronger prioritization of projects with higher ROIs. The first conclusion is one that is well-trodden, but
These findings suggest that in the context of here is empirically validated, is that you simply cannot
nonpoint pollution mitigation, at least, the efficacy “throw money at the problem” of nonpoint source
of collaboration governance regimes is contingent on pollution. A rush to fund at the expense of proper
the specific properties of both institutions and ecosys- planning, targeted focus, and appropriate levels of
tems (Raab, Mannak, and Cambré 2013). Taken more administrative capacity can leave a system rich in fi-
broadly, we can conclude that in circumstances when nancial capital, but poor in knowledge, human and
governance networks carry out specific regulatory and physical capital, and ultimately, results.
capital-intensive projects that the use of collaborative The second main conclusion to be drawn here pertains
governance regimes should be considered as contin- to the importance of administrative capacity. The kind of
gent upon actor’s propensities to collaborate and the process-based modeling demonstrated here can be very
administrative capacities of central actors. This latter useful in determining where and how administrative
point highlights the importance that network struc- bottlenecks occur, and their relation to program perform-
tures play in determining the optimal collaborative ance. Strategic investments in administrative capacity are
governance regime type. Here, the properties of the critically important, and an assertion that is likely not lost
network matter, signaling the need for deeper integra- on practicing public managers. This fact may, at times, be
tion of collaborative and network governance theories. lost on policymakers who are setting the policy agenda
The contributions of this study to theory develop- and making resource allocation decisions. ABMs and
ment include the demonstration of the relationship other process-based models of alternative governance de-
between network governance and collaborative gov- sign can be used to inform policy decisions and ensure
ernance. As noted earlier, there is compatibility be- that administrative capacity is dully considered.
tween collaborative governance typologies of origins The third main conclusion to be drawn pertains to
and network governance typologies of relatively stable designing for collaborative governance and the effi-
governance design. Future theory development should cacy of collaborative and network governance theories
focus on the relationship between specific structural to inform this process. The apparent inefficiencies of
properties of networks and collaborative capacity. voluntary collaboration at both the planning and im-
The ability of social-ecological systems to be re- plementation stages are worth noting again. When
silient is contingent on their ability to adapt and trans- agency is provided to actors such as municipal agents
form as conditions change (Folke et al. 2002; Gallopín at a finer, localized scale, their limited perceptions of
2006). There is congruence among dimensions of col- the whole system and the collective goals tied to that
laborative governance and adaptive capacity theories, system, may be lost in the specifics of local politics,
suggesting that as institutions enable collaboration, resource constraints and the like. The apparent value
their ability (and that of the system) to adapt may be of mandated collaboration, not just in planning, but in
increased (Emerson and Gerlak 2014). project implantation as well, suggests the need to “level
This study demonstrates the value that ABMs can the playing field” by pressing for stronger coordination
have for theory testing and institutional design and across scales of government. In this regard, our study
contributes to the growing literature regarding the use provides some very important insights regarding man-
of simulation modeling to test alternative design rules. dated versus voluntary collaboration.
This study highlights the potential uses of dynamic Bennett, David, and David McGinnis. 2008. Coupled and complex:
models of governance networks (Koliba, Zia, and Human–environment interaction in the Greater Yellowstone
Ecosystem, USA. Geoforum 39 (2): 833–45.
Merrill 2019) to inform actual public policy. The State Berardo, Ramiro, and Mark Lubell. 2016. Understanding what
of Vermont is set to enact mandated water districts, and shapes a polycentric governance system. Public Administration
this governance model has been credited with helping Review 76 (5): 738–51.
to shape this legislation (Kamman 2019). That this Bitterman, Patrick, and David A. Bennett. 2018. Leveraging
model was able to draw on actual project data, pro- coupled agent-based models to explore the resilience of
tightly-coupled land use systems. In Agent-Based Models
jected nutrient load reductions of those projects, and

and Complexity Science in the Age of Geospatial Big Data,
hydrological modeling estimates used by the USEPA ed. Liliana Perez, Eun-Kyeong Kim, and Raja Sengupta, 17–
and the State to set TMDL targets, demonstrates the 30. Cham, Switzerland: Springer International Publishing.
practical use of coupled modeling approaches that in- doi:10.1007/978-3-319-65993-0_2
corporate alternative governance designs. That policy- Bodin, Örjan, Annica Sandström, and Beatrice Crona. 2017.
Collaborative networks for effective ecosystem-based manage-
makers in the State of Vermont were involved in the ment: A set of working hypotheses: Collaborative networks
design and analysis of this model also underscores the and ecosystem management. Policy Studies Journal 45 (2):
potential value of using dynamic modeling to shape le- 289–314.
gislation and governance arrangements. In this manner, Booher, David E. 2004. Collaborative governance practices and
policy implementation and governance design can be democracy. National Civic Review 93 (4): 32–46.
Booher, David E., and Judith E. Innes. 2010. Governance for resili-
understood as key components of a wider situational ence: CALFED as a complex adaptive network for resource man-
awareness of a social-ecological system. agement. Ecology and Society 15 (3): art35.
Brummel, Rachel F., Kristen C. Nelson, and Pamela J. Jakes. 2012.
Supplementary Material Burning through organizational boundaries? Examining inter-
organizational communication networks in policy-mandated
Supplementary material is available at the Journal of collaborative bushfire planning groups. Global Environmental
Public Administration Research and Theory online. Change 22 (2): 516–28.
Bryson, John M., Barbara C. Crosby, and Melissa Middleton Stone.
2015. Designing and implementing cross-sector collaborations:
Funding Needed and challenging. Public Administration Review 75 (5):
647–63.
This material is based upon work supported by the Bucini, Gabriela, Scott C. Merrill, Eric Clark, Susan M. Moegenburg,
National Science Foundation under VT EPSCoR Asim Zia, Christopher J. Koliba, Serge Wiltshire, Luke Trinity,
Grant No. NSF OIA 1556770, Lake Champlain and Julia M. Smith. 2019. Risk attitudes affect livestock
Basin Resilience to Extreme Events (BREE). This biosecurity decisions with ramifications for disease control in a
simulated production system. Frontiers in Veterinary Science 6
work also benefited from collaboration supported by
(June): 196.
the National Socio-Environmental Synthesis Center Clark, Richard, Stephanie Hurley, and Christopher Koliba. 2018.
(SESYNC) under funding received from the National Results of a Municipal Stormwater Capacity Study. Presented at
Science Foundation DBI-1052875. the BREE All Hands Meeting, Burlington, VT, June 12. https://
No potential conflict of interest is reported by the epscor.w3.uvm.edu/2/sites/default/files/general/public/009-chris-
bree-socialsystems-all-hands-june-2018.pdf.
authors.
Daniels, Steven E., and Gregg B. Walker. 2001. Working through
environmental conflict: The collaborative learning approach.
Westport, CT: Praeger.
References Deadman, P. J. 1999. Modelling individual behaviour and group per-
Ansell, Chris, and Alison Gash. 2008. Collaborative governance in formance in an intelligent agent-based simulation of the tragedy
theory and practice. Journal of Public Administration Research of the commons. Journal of Environmental Management 56 (3):
and Theory 18 (4): 543–71. 159–72.
———. 2018. Collaborative platforms as a governance strategy. Eckerd, Adam, Heather Campbell, and Yushim Kim. 2012. Helping
Journal of Public Administration Research and Theory 28 (1): those like us or harming those unlike us: Illuminating social
16–32. processes leading to environmental injustice. Environment and
Arnold, J. G., and N. Fohrer. 2005. SWAT2000: Current capabil- Planning B: Planning and Design 39 (5): 945–64.
ities and research opportunities in applied watershed modelling. Emerson, Kirk, and Andrea K. Gerlak. 2014. Adaptation in collab-
Hydrological Processes 19 (3): 563–72. orative governance regimes. Environmental Management 54 (4):
Batty, Michael, Andrew T. Crooks, Linda M. See, and 768–81.
Alison J. Heppenstall. 2011. Perspectives on Agent- Emerson, Kirk, and Tina Nabatchi. 2015. Collaborative governance
Based Models and Geographical Systems. In Agent-Based regimes. Washington, DC: Georgetown University Press.
Models of Geographical Systems, ed. Alison J. Heppenstall, Erdlenbruch, Katrin, and Bruno Bonté. 2018. Simulating the dy-
Andrew T. Crooks, Linda M. See, and Michael Batty, 1–15. namics of individual adaptation to floods. Environmental Science
Dordrecht, the Netherlands: Springer Science & Business Media. & Policy 84 (June): 134–48.
Bellaubi, Francesc, and Claudia Pahl-Wostl. 2017. Corruption risks, Folke, Carl, Steve Carpenter, Thomas Elmqvist, Lance Gunderson,
management practices, and performance in water service de- C. S. Holling, and Brian Walker. 2002. Resilience and sustain-
livery in Kenya and Ghana: An agent-based model. Ecology and able development: Building adaptive capacity in a world of trans-
Society 22(2):6. formations. AMBIO 31 (5): 437–40.
Gallopín, Gilberto C. 2006. Linkages between vulnerability, resili- Koliba, Christopher, Asim Zia, and Scott Merrill. 2019. Using
ence, and adaptive capacity. Global Environmental Change 16 agent-based models to study network and collaborative govern-
(3): 293–303. ance. In Researching Networks and Collaboration in the Public
Gassman, P. W., M. R. Reyes, C. H. Green, and J. G. Arnold. 2007. Sector: A Guide to Approaches, Methodologies, and Analytics,
The soil and water assessment tool: Historical development, ap- ed. J. Voets, R. Keast, and C Koliba, 210–31. New York, NY:
plications, and future research directions. Transactions of the Routledge.
ASABE 50 (4): 1211–50. Koliba, Christopher, Asim Zia, Andrew Schroth, Arne Bomblies,
Gerlak, A. K. 2005. Federalism and U.S. water policy: Lessons for Judy Van Houten, and Donna Rizzo. 2016. The Lake Champlain
the twenty-first century. Publius 36 (2): 231–57. Basin as a complex adaptive system: Insights from the Research

Goldstone, Robert L., and Marco A. Janssen. 2005. Computational on Adaptation to Climate Change (RACC) Project. Vermont
models of collective behavior. Trends in Cognitive Sciences 9 (9): Journal of Environmental Law 17: 533–63.
424–30. Koontz, Tomas M., and Elizabeth Moore Johnson. 2004. One size
Grimm, Volker. 2005. Pattern-oriented modeling of agent-based does not fit all: Matching breadth of stakeholder participation
complex systems: Lessons from ecology. Science 310 (5750): to watershed group accomplishments. Policy Sciences 37 (2):
987–91. 185–204.
Grimm, Volker, Uta Berger, Donald L. DeAngelis, J. Gary Polhill, Koontz, Tomas M., Toddi A. Steelman, J. Carmin, K.S. Korfmacher,
Jarl Giske, and Steven F. Railsback. 2010. The ODD protocol: C. Moseley, and Craig W. Thomas. 2010. Collaborative environ-
A review and first update. Ecological Modelling 221 (23): mental management: What roles for Government-1. New York,
2760–8. NY: Routledge.
Hamilton, Matthew, and Mark Lubell. 2018. Collaborative govern- Lake Champlain Basin Program. 2018. 2018 State of the Lake and
ance of climate change adaptation across spatial and institutional Ecosystem Indicators Report. Grand Isle, VT: Lake Champlain
scales. Policy Studies Journal 46 (2): 222–47. Basin Program.
Imperial, Mark T. 2005. Using collaboration as a governance Lansing, J. Stephen, and James N. Kremer. 1993. Emergent prop-
strategy: Lessons from six watershed management programs. erties of Balinese water temple networks: Coadaptation on a
Administration & Society 37 (3): 281–320. rugged fitness landscape. American Anthropologist, New Series
Isles, Peter D. F., Yaoyang Xu, Jason D. Stockwell, and 95 (1): 97–114.
Andrew W. Schroth. 2017. Climate-driven changes in energy Ligmann-Zielinska, Arika, and Piotr Jankowski. 2007. Agent-based
and mass inputs systematically alter nutrient concentration and models as laboratories for spatially explicit planning policies.
stoichiometry in deep and shallow regions of Lake Champlain. Environment and Planning B: Planning and Design 34 (2):
Biogeochemistry 133(2):201–17. 316–35.
Janssen, Marco A., and Jacopo A. Baggio. 2017. Using agent-based Lubell, Mark, Garry Robins, and Peng Wang. 2014. Network struc-
models to compare behavioral theories on experimental data: ture and institutional complexity in an ecology of water manage-
Application for irrigation games. Journal of Environmental ment games. Ecology and Society 19 (4): art23.
Psychology 52 (October): 194–203. Luke, Sean, Claudio Cioffi-Revilla, Liviu Panait, Keith Sullivan, and
Johnson, Paul E. 1999. Simulation modeling in political science. Gabriel Balan. 2005. MASON: A multiagent simulation environ-
American Behavioral Scientist 42 (10): 1509–30. ment. Simulation 81 (7): 517–27.
Kamman, Neil. 2019. Policy and management implication of Magliocca, Nicholas R., and Erle C. Ellis. 2013. Using Pattern-
basin resilience to extreme events research. Presented at the Oriented Modeling (POM) to cope with uncertainty in multi-
Vermont EPSCoR Basin Resilience to Extreme Events All scale agent-based models of land change: POM in multi-scale
Hands Meeting, Colchester, VT, June 4. https://epscor.w3.uvm. ABMs of land change. Transactions in GIS 17 (6): 883–900.
edu/2/node/4675. Manson, Steven M. 2005. Agent-based modeling and genetic pro-
Kamensky, John M. 2019. What do they want to know? In gramming for modeling land change in the Southern Yucatán
Networks and Collaboration in the Public Sector: Essential Peninsular Region of Mexico. Agriculture, Ecosystems &
Research Approaches, Methodologies and Analytic Tools, ed. Environment 111 (1–4): 47–62.
Joris Voets, Robin Keast, and Christopher Koliba, 62. Abingdon, Maroulis, Spiro. 2016. Interpreting school choice treatment ef-
UK: Routledge. fects: Results and implications from computational experiments.
Kania, John, and Mark Kramer. 2011. Collective impact. Stanford Journal of Artificial Societies and Social Simulation 19 (1): 7.
Social Innovation Review 9 (1): 36–41. Milio, Simona. 2007. Can administrative capacity explain differences
Keast, Robyn, and Myrna Mandell. 2014. The collaborative push: in regional performances? Evidence from structural funds imple-
Moving beyond rhetoric and gaining evidence. Journal of mentation in southern Italy. Regional Studies 41 (4): 429–42.
Management & Governance 18 (1): 9–28. Milward, H. Brinton, and Keith G Provan. 2006. A manager’s guide
Kim, Yushim. 2007. Using spatial analysis for monitoring fraud in to choosing and using collaborative networks, Vol 8. Washington,
a public delivery program. Social Science Computer Review 25 DC: IBM Center for the Business of Government.
(3): 287–301. Nambisan, Satish, and Robert A. Baron. 2009. Virtual customer
Knieper, Christian, and Claudia Pahl-Wostl. 2016. A compara- environments: Testing a model of voluntary participation in
tive analysis of water governance, water management, and en- value co-creation activities. Journal of Product Innovation
vironmental performance in river basins. Water Resources Management 26 (4): 388–406.
Management 30 (7): 2161–77. Norton, Bryan G. 2005. Sustainability: A philosophy of adaptive
Koliba, Christopher, Jack W. Meek, Asim Zia, and Russell W. Mills. ecosystem management. Chicago, IL: University of Chicago
2018. Governance networks in public administration and public Press.
policy, 2nd ed. Oxfordshire, UK: Routledge. Ostrom, Elinor. 1990. Governing the commons: The evolution of in-
Koliba, Christopher, Adam Reynolds, Asim Zia, and Steven Scheinert. stitutions for collective action. Cambridge, UK: Cambridge uni-
2014. Isomorphic properties of network governance: Comparing versity press.
two watershed governance initiatives in the Lake Champlain Pahl-Wostl, Claudia, Marc Craps, Art Dewulf, Erik Mostert,
Basin using institutional network analysis. Complexity, David Tabara, and Tharsi Taillieu. 2007. Social learning and
Governance & Networks 1 (2): 99–118. water resources management. Ecology and Society 12 (2): art5.
Pahl-Wostl, Claudia, Georg Holtz, Britta Kastens, and Scott, Tyler A, Craig W Thomas, and José Manuel Magallanes.
Christian Knieper. 2010. Analyzing complex water govern- 2018. Convening for consensus: Simulating stakeholder agree-
ance regimes: The Management and Transition Framework. ment in collaborative governance processes under different net-
Environmental Science & Policy 13 (7): 571–81. work conditions. Journal of Public Administration Research and
Parker, Dawn C., Steven M. Manson, Marco A. Janssen, Theory 29 (1): 32–49.
Matthew J. Hoffmann, and Peter Deadman. 2003. Multi-agent Smajgl, Alexander, Scott Heckbert, John Ward, and Anna Straton.
systems for the simulation of land-use and land-cover change: 2009. Simulating impacts of water trading in an institu-
A review. Annals of the Association of American Geographers tional perspective. Environmental Modelling & Software
93 (2): 314–37. 24(2):191–201.

Provan, K. G., and P. Kenis. 2007. Modes of network governance: Squazzoni, Flaminio, and Riccardo Boero. 2010. Complexity-
Structure, management, and effectiveness. Journal of Public friendly policy modelling. London, UK: Routledge.
Administration Research and Theory 18 (2): 229–52. Turrini, Alex, Daniela Cristofoli, Francesca Frosini, and Greta Nasi.
Raab, J., R. S. Mannak, and B. Cambré. 2013. Combining structure, 2009. Networking literature about the determinants of network
governance, and context: A configurational approach to network effectiveness. Public Administration 88(2):528–50.
effectiveness. Journal of Public Administration Research and US Environmental Protection Agency. 2016. Phosphorus TMDLs
Theory 25 (2): 479–511. for Vermont Segments of Lake Champlain. Boston, MA: US
Rhodes, Rod AW. 1997. Understanding governance: Policy net- Environmetnal Protection Agency. https://www.epa.gov/tmdl/
works, governance, reflexivity and accountability. Buckingham, lake-champlain-phosphorus-tmdl-commitment-clean-water (ac-
UK:Open University Press. cessed May 1, 2019).
Rodríguez, Charo, Ann Langley, François Béland, and Jean- US Environmental Protection Agency (USEPA). 2018. Harmful Algal
Louis Denis. 2007. Governance, power, and mandated collab- Blooms. United States Environmental Protection Agency. https://
oration in an interorganizational network. Administration & www.epa.gov/nutrientpollution/harmful-algal-blooms (accessed
Society 39 (2): 150–93. May 1, 2019).
Rogers, Ellen, and Edward P. Weber. 2010. Thinking harder about Vermont Division of Environmental Conservation (VTDEC).
outcomes for collaborative governance arrangements. The 2019. Clean Water Portal. https://anrweb.vt.gov/DEC/
American Review of Public Administration 40 (5): 546–67. cleanWaterDashboard/WPDSearch.aspx (accessed May 5,
Scheinert, Steve, Christopher Koliba, Stephanie Hurley, 2019).
Sarah Coleman, and Asim Zia. 2015. The shape of watershed Vermont General Assembly. 2019. Bill Status S.96 (Act 76).
governance: Locating the boundaries of multiplex networks. Vermont General Assembly. https://legislature.vermont.gov/bill/
Complexity, Governance & Networks 2 (1): 65–82. status/2020/S.96 (accessed July 1, 2019).
Scheinert, Steve, Asim Zia, Christopher Koliba, and Scott Merrill. Watkins, Cristy, Dean Massey, Jeremy Brooks, Kristen Ross, and
2017. Growing collaborations forecasting changes in partner- Moira L. Zellner. 2013. Understanding the mechanisms of col-
ship networks using a bottom-up approach. Policy and Complex lective decision making in ecological restoration: An agent-based
Systems 3 (1): 3–27. model of actors and organizations. Ecology and Society 18 (2):
Schlüter, Maja, Andres Baeza, Gunnar Dressler, Karin Frank, 32.
Jürgen Groeneveld, Wander Jager, Marco A. Janssen, et al. 2017. Wiltshire, Serge, Asim Zia, Christopher Koliba, Gabriela Buccini,
A framework for mapping and comparing behavioural theories Eric Clark, Scott Merrill, Julie Smith, and Susan Moegenburg.
in models of social-ecological systems. Ecological Economics 2019. Network meta-metrics: Using evolutionary computation
131 (January): 21–35. to identify E Ective indicators of epidemiological vulnerability
Schlüter, Maja, and Claudia Pahl-Wostl. 2007. Mechanisms of resili- in a livestock production system model. Journal of Artificial
ence in common-pool resource management systems: An agent- Societies and Social Simulation 22 (2): 8.
based model of water use in a river basin. Ecology and Society Zia, Asim, Arne Bomblies, Andrew W. Schroth, Christopher Koliba,
12 (2): 4. Peter D. F. Isles, Yushiou Tsai, Ibrahim N. Mohammed, et al.
Scott, Tyler. 2015. Does collaboration make any difference? Linking 2016. Coupled impacts of climate and land use change across
collaborative governance to environmental outcomes: Does col- a river–lake continuum: Insights from an Integrated Assessment
laboration make any difference? Journal of Policy Analysis and Model of Lake Champlain’s Missisquoi Basin, 2000–2040.
Management 34 (3): 537–66. Environmental Research Letters 11 (11): 114026.
Scott, Tyler A, and Craig W Thomas. 2017. Winners and losers in Zia, Asim, and Christopher Koliba. 2013. The emergence of at-
the ecology of games: Network position, connectivity, and the tractors under multi-level institutional designs: Agent-based
benefits of collaborative governance regimes. Journal of Public modeling of intergovernmental decision making for funding
Administration Research and Theory 27 (4): 647–60. transportation projects. AI & Society 30 (3): 315–31.

Modelado de Diseños de Redes de Gobernanza Colaborativa Alternativa

Uploaded by

Copyright:

Available Formats

Modelado de Diseños de Redes de Gobernanza Colaborativa Alternativa

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Modelado de Diseños de Redes de Gobernanza Colaborativa Alternativa

Uploaded by

Copyright:

Available Formats

Journal of Public Administration Research and Theory, 2020, 1–20

Modeling Alternative Collaborative Governance

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Address correspondence to the author at [email protected].

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Table 1. Variables and Functions Governing Behavior of Municipal Agents

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Table 4. Analysis of Variance of Interactions Among Model Input Parameters

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

Downloaded from https://academic.oup.com/jpart/advance-article-abstract/doi/10.1093/jopart/muaa013/5820199 by guest on 25 April 2020

You might also like