Biodiversity Conservation

Université Paris 1 Panthéon Sorbonne - M1 Economie Internationale - Topics in Environment

Simon Jean
simon.jean@agroparistech.fr

AgroParisTech - CIRED - PSAE

Lecture plan

  1. Biodiversity: concept, measures and decline
  2. Rationales for biodiversity conservation
  3. How to conserve biodiversity?

Introduction

COP 16 and biodiversity conservation

Multilateral policy for biodiversity

Keunming-Montreal Global Biodiversity Framework:

  • Follows the failure of the Aichi targets to halt biodiversity loss
  • Structured around 4 goals by 2050 :
  1. “Protect and Restore” : enhance ecosystem health, halt human induced extinction, genetic diversity is maintained
  2. “Propsper with Nature” : biodiversity is sustainably used
  1. “Share Benefits Fairly”: monetary and non monetary benefits from various uses of biodiversity are shared, including indigenous communities
  2. “Invest and Collaborate”: means of implementation are accessible, especially to developping countries
  • And 23 targets by 2030 , such as the fammous “30 by 30” (targets 2 and 3)
  • With associated metrics and reporting

With scientific support from IPBEs

  • Interdisciplinary Panel on Biodiversity and Ecosystem Services
  • Founded in 2012, under the UNEP
  • Gathers scientific information about the state of biodiversity, and ecosystem services
  • Basically the equivalent of IPCC for living world

1. Biodiversity : concepts, measures and decline

A. Definition

  • From conservation biology, a subfield of biology created to halt the decline of biological diversity (M. Soulé)

“Biological diversity”means the variability among living organisms from all sources including, inter alia, terrestrial, marine and other aquatic ecosystems and the ecological complexes of which they are part; this includes diversity within species, between species and of ecosystems

Article 2, Convention on Biological Diversity, 1992

  • Different levels : within an ecosystem, between ecosystems etc
  • Different diversities : structural, functional, genetic

B. Different types of measures

For an overview of measures used, see Marcon, 2018 1

Species neutral diversity

  • Richnesss : number of classes in a given area
    ex : number of species of trees in a forest
    • \(\alpha\) diversity for community level
    • \(\gamma\) diversity for meta community level
  • Evenness : similarity of abudances of each species in an environment
    ex : a forest with 99% spruce & 1% oak v. a forest with 50% spruce and 50% oak
  • Shannon index :
    • Mean measure of surprise that a sample population gives : individual with low probability of occurence gives more information
    • with \(p_i=\frac{N_i}{N}\) and \(S\) species richness
    • \(H = -\sum_{i=1}^S p_i \log_2(p_i)\)
  • Simpson index :
    • probability that two individuals randomly drawn belong to the same species
    • \(S' = \sum_{i=1}^S p_i^2\)

Non neutral diversity measures

  • Dissimilarity : take into account the distance between members of a class or a species (DNA, branches of a phylogenetic tree)

Take home

  • Biodiversity measures are different but highlight the fact that the sum is more than its parts
  • As we’ll see, depending on what you choose, definitions yield different conservation strategies

C. A worldwide decline

2. Rationales for biodiversity conservation

A. The concept of Total Economic Value

The concept of total economic value emerges from John Krutilla’s work (1967, Conservation Reconsidered):

  • Use values can be :
    • Direct : associated with the direct use of ressources from the environment through consumption (fish, timber etc)
    • Indirect : values associated with indirect contributions from ecosystems, such as flood protection, or as inputs in productive activities (pollination)
    • Option : the possibility of using a resource, either for preservation or exploitation, in the future, when technical progress makes it useful
  • Non use values :
    • Bequest Value: The value placed on preserving ecosystems for future generations.
    • Existence Value: The benefit derived from simply knowing that an ecosystem or species exists, even if one never directly interacts with it.

B. Linking Nature to its Values

Ecosystem services

  • Ecosystem services emerged as a pedagogical metaphor to highlight the crucial dependence of humans on ecosystems
  • Became a way to measure the values of ecosystems in the 1990s
  • Different types of ecosystem services :
    • Provision
    • Regulation
    • Cultural and Recreation
  • Who provides these services?
    • Individual species for direct use values
    • Ecosystems overall for indirect use values such as water filtration etc
  • Does species diversity lead to more ecosystem services?
    • Sampling effect : larger diversity ensures more productive species are present
    • Niche differentiation effect : more species will use resources more efficiently
    • Insurance effect : greater diversity is linked to lower variance of services
  • Results are supported across a wide range of ecosystems

Towards Nature Contributions to People

  • With criticism surrounding the commodification of Nature and to include non instrumental values, IPBES recently moved towards the Nature Contributions to People framework
    • Emphasizes the multifaceted relationships between people and nature, capturing both tangible/intangible, positive/negative contributions
    • Recognizes the importance of indigenous and local knowledge, cultural heritage, and ethical considerations in understanding human–nature connections.

C. Evaluating Ecosystem Services

  • Costanza et al, The value of the world’s ecosystem services and natural capital, 1997, Nature
  • Find that the natural capital and associated ecosystem services is worth $33 trillions
    • Still larger than today’s GDP
    • But lower than the potential infinity of its value if disappeared
  • How would you value ecosystem services? What do you need to know?

A recent example : biological pest control

Eyal Frank, The economic impacts of ecosystem disruptions: Costs from substituting biological pest control, Science, 2024

Causal study of the impact of the onset of white nose syndrom in the US using staggered difference in difference on pesticide use and infant health

  • After the onset of bat die-offs, farmers in the county increase their insecticide use by 31.1%, on average
  • Substitution between a declining natural input and a human-made input
  • Infant mortality rates due to internal causes of death (i.e., not due to accidents or homicides) increased by 7.9%, on average, in the affected counties

3. How to conserve biodiversity?

A. Promoting reasonable use of resources and ecosystem based management?

1. Bioeconomic analysis of fishery

  • Defining a model :
    • Mechanistic : based on equations with identified variables
    • To understand the link between a scenario and social ecological consequences
    • Thus excluding strictly correlative methods
  • Featuring population dynamics
    • Of a (set of) weakly managed or wild species
  • A decision process emerging from economic theory:
    • Rules for decision making
    • Can be optimisation as well as agent based etc
  • A link between subsystems:
    • Economics impact biodiversity
    • Biodiversity impacts the economy

Population dynamics

  • Only one species is considered
  • Track the evolution of biomass (in tons) through time

\[ X_{t+1} = F(X_t) + X_t = G(X_t) \]

  • \(F(X_t)\) is the growth function of the population (abuse of language : mean growth function)
  • It includes a density effect : growth level depends on the population (or stock level) :
    • It can be large for low levels of population
    • And low for large levels of population (saturation effect)
    • A same growth level can happen at two population levels
  • Main example : logistic growth (Verhulst, 1838)

\[F(X_t) = rX_t \left(1 - \frac{X_t}{K}\right)\]

with \(r\) the intrinsic growth rate (for low population levels) and \(K\) the carrying capacity of the ecosystem (in tons)

Biological equilibria

We have an equilibrium when the population is not changing through time :

\[ \begin{align*} X_{t+1} &= X_t = X^*\\ \Rightarrow & X_{t+1} -X_t = F(X_t) = 0 \end{align*} \]

We’re looking at a convergence, a long term case, where things are no longer changing as determined by fixed external factors

Extraction technology and sustainable yield

  • Extraction technology : links harvesting effort to harvested quantity such that \[Y_t = h(X_t, E_t)\]
  • Population dynamics change with extraction :

\[\begin{align*} X_{t+1} - X_t &= F(X_t) - Y_t= F(X_t)-h(X_t, E_t) \end{align*}\]

  • Equilibrium population is such that fishing exactly equals growth: \(X_{t+1}-X_t=0 \iff F(X_t) = h(X_t, E_t)\)

  • An equilibrium fishing level, with the associated stock level is called sustainable yield if the fish growth is equal to the fish harvest

  • If \(Y_t>F(X_t)\), \(X_{t+1}<X_t\)

  • If \(Y_t<F(X_t)\), \(X_{t+1}>X_t\)

  • In equilibrium, \(h_t(X_t,E_t) = F(X_t)\)

  • The maximum sustainable yield is the maximal fishing level sustainable in the long term

Measurements

  • Effort is difficult to measure, as it depends on the technology : km of lines trawled, hooks/km/day etc
  • Catchability is difficult to measure and to rely on : here we assume a fixed rate, but it can change! \[Y=q(X)XE\]
  • North Atlantic Cod collapsed in the 1990s after decades of overfishing and overcapitalization to 0.1% of its historical level
  • A moratorium was declared in 1992, causing 37,000 fishermen to lose their job (Canada’s single largest layoff in history), that has started to be lifted in 2024
  • Among the reasons for its collapse was a wrong evaluation of thelink between catch per unit of effort (\(Y/E\)), catchability and stock (\(qX\)) (see Rose et Kulka, 1999:
  • CPUE was used as an indicator for abundance, while the stock actually displayed hyperstability!

Plugging the economics

  • How do we go from effort to population to dollars?
  • First, recognize that the equilibrium relationship implies that : \[ \begin{align*} &rX^* \left(1 - \frac{X^*}{K}\right) = qX^* E\\ \Rightarrow & X^* = \left(1- \frac{q}{r}E \right) K \\ \Rightarrow & Y = qE\left(1- \frac{q}{r}E \right) K \end{align*} \]
  • All that’s left to do is bring in some economics around effort
  • Define total revenue by \(RT= pY\) with \(p\) the price of the resource
    • Assume a competitive output market (most of the time, but not always the case)
  • Define total costs as \(CT = cE\) :
    • Assume total costs do not depend on the population of fish
    • And long term perspective, where entry is at its largest and costs depend on effort proportionally (no need to drag additional people)
    • Hence, average cost is marginal cost
  • Profit is thus : \[ \Pi = RT - CE = pY - cE = pqE \left(1 - \frac{q}{r}E \right)K - cE\]

Explaining overfishing

  • Gordon introduces the fundamental difference between agriculture and fishing, through the analysis of differentiated rents (i.e. economic returns above the “normal” profit level due to varying productivities)
    • We now come to the point that is of greatest theoretical importance in understanding the primary production phase of the fishing industry and in distinguishing it from agriculture.
    • In the sea fisheries the natural resource is not private property : hence the rent it may yield is not capable of being appropriated by anyone. The individual fisherman has no legal title to a section of ocean bottom. Each fisherman is more or less free to fish wherever he pleases
    • The result is a pattern of competition among fishermen which culminates in the dissipation of the rent of the intramarginal grounds
  • Key ideas and concepts are intramarginal grounds, heterogeneous rents, and lack of property (i.e. open access)

Let :

  • Average revenue (per unit of effort) be: \[RT = pY(E)/E = pq\left(1 - \frac{q}{r}E \right)K\]
  • Marginal revenue : \[Rm = \frac{\partial pY(E)}{E} = pq\left(1 - 2E\frac{q}{r} \right)K\]
  • Average and marginal cost : \[CM = Cm = c\]

Two fishing grounds

  • When both grounds are optimally managed, effort levels are A and B. Why?
  • That’s what matters for a social planner
  • But what would you do, as a fisherman? Where would you go?
    • What matters is not the marginal profit, on aggregate
    • But as a person, it is the mean profit: where will you earn the most?

  • Fishers will redispatch until profits are equalized accross zones
  • To this end, profit levels are equal : spatial arbitrage opportunities are exhausted
  • A marginal fisherman would earn the same thing anywhere

  • This phenomenon is repeated for every fishing ground that is at least more productive than the marginal cost of harvesting, until the last ground, the marginal ground
  • All zones with higher productivity (or profit) are called intra marginal grounds
  • Profits get dissipated because of lack of clear property rights

Open access equilibrium

Comparing policy recommendations

  • We now know that open access is likely to yield to overharvesting and overcapitalization of the fisheries
  • What should be implemented?
  • Maximum sustainable yield was recommended as the basis for policy
    • It sill is recommended in the European Union, under the Common Fisheries Policy (2013) according to a precautionary principle
  • Gordon suggested to aim for the maximum economic yield, i.e. the effort level that maximizes profits, thus taking economics into account, implemented in Australia for example

p = 5
c = 2

# Define growth from effort
Croissance = function(Effort){
  y = q*Effort * (1 - q/r*Effort)*K
  return(y)
}

# Define monetary functions with negative return for optimization
profit = function(Effort, p_ = p , c_ = c){
  y = p_*Croissance(Effort) - c_*Effort
  return(-y)
}

marginal_revenue= function(Effort, p_ = p, c_ = c){
  y = p*q*K*(1-2*Effort*q/r)
  return(y)
}

# Run optimization with integrated solver 

solution = optimize(profit, interval = c(0,K))
e_look = solution$minimum

# Create data.frame 
data_fin <- data.frame(Effort = seq(0, 5, 0.01)) %>%
        mutate(Croissance = Croissance(Effort)) %>%
        mutate(Revenue = p * Croissance, 
               Cost = c * Effort) %>%
        mutate("Marginal Revenue" = marginal_revenue(e_look) * (Effort - e_look) +
                 p * Croissance(e_look))%>% # Tangent equation
        select(-Croissance)

# Treat data into tidy format
data_fin %>%
  pivot_longer(-Effort, #Pivot to long format
               values_to = "values", 
               names_to = "names") 
# A tibble: 1,503 × 3
   Effort names            values
    <dbl> <chr>             <dbl>
 1   0    Revenue           0    
 2   0    Cost              0    
 3   0    Marginal Revenue 20.2  
 4   0.01 Revenue           0.200
 5   0.01 Cost              0.02 
 6   0.01 Marginal Revenue 20.3  
 7   0.02 Revenue           0.398
 8   0.02 Cost              0.04 
 9   0.02 Marginal Revenue 20.3  
10   0.03 Revenue           0.596
# ℹ 1,493 more rows
# Then use ggplot to plot

And promote reasoned use

  • How can one do to steer OA towards the policy reference point?
  • Work on drivers of fishing effort :
    • Taxing inputs
    • Limit days at sea on aggregate
    • Include technical regulations
  • Work on catch:
    • Landing tax
    • Set limits to catch on aggregate : can cause race to fish situations
  • To explore the different policy levers within the Gordon Schaeffer framework, you can use the following app : https://simon-jean.shinyapps.io/gordon_schaeffer/

In practice in the EU

  • In practice, need to combine approaches due to monitoring and implementation difficulties
  • For a year
  • Based on stock assessments
  • Negociated with non EU countries for shared stocks
  • Shared between EU members with nationally determined quotas within the Common Fishery Policy
  • Assorted with mandatory landing :
    • Have to land everything that is caught, and count it wrt quota (even bycatch, or too small fish)
    • To value them in other industries (fish flour etc)
    • Incentive : today’s catch is tomorrow’s quotas, hence there will be a top-up of quotas
    • But can help respect the system and have better scientific information
    • Otherwise, risk of fine and licence suspension
  • Allow the limitation of administrative costs, economic profitability and conservation

RÈGLEMENT (UE) 2025/219 DU CONSEIL du 30 janvier 2025

(Individual) rights based fisheries management

  • The absence of clearly defined (bundels of) property rights in open access setting may dissolve incentives for conservation
  • Idea is to allocate individual quotas to fishermen, or geographic access rights
  • So that the integral flow of benefits and costs accrue to fishermen through time
  • Rights ought to be transferable for efficiency :
    • Low productivity fishing may overfish
    • High productivity fishing may underfish
    • Transferability can create a market where rights are traded such that marginal costs are equated \(c'(Y) = p - \pi_y\)
  • Could argue for conservation banks where people buy fishing rights to for conservation (See Leonard et al., Science, 2021)

Limites and vigilance points

Lire A Cautionary Note on Individual Transferable Quotas, R. Sumaila, Ecology and Society, 2010

  • Property rights thus defined are not full
    • Finite duration
    • Risk discarding low value fish (low grading) and discard overall
    • Require monitoring and enforcement
  • Does not solve the discounting issue about fisheries : if the reproduction rate of the resource is low, and discount rate is high, may not be worth conserving
  • Gains come from reduced fishing intensity (so increased duration, or lower employment) which may conflict with the sector needs
  • Requires good scientific information
  • Initial allocation issue :
    • May face concentration due to gains from trade
    • Can trigger monopoly behavior, social inequalities, and lobbying

Design principles for common property resources

In real life, when resources can be held in common (implies low cost of exclusion - alpine meadows, some inland fisheries (Turkey)), other management schemes can be used to foster good resource use, as shown by Elinor Ostrom

  • 1A. User Boundaries : Clear and locally understood limits between legitimate users and non-users are in place
  • 1B. Resource Boundaries : Clear-cut boundaries that separate a specific common resource from a broader socio-ecological system are in place
  • 2A. Alignment with Local Conditions : The rules for appropriation and provision are in line with local social and environmental conditions
  • 2B. Appropriation and Provision : The distribution of costs is proportional to the distribution of benefits
  • 3. Collective Choice Mechanisms : Most individuals affected by a resource regime are allowed to participate in the formulation and modification of its rules
  • 4A. User Monitoring : Responsible individuals or users monitor the levels of appropriation and provision among users
  • 4B. Resource Monitoring : Responsible individuals or users monitor the state of the resource
  • 5. Graduated Sanctions : They start mild, but become more severe in the case of repeated infractions
  • 6. Conflict Resolution Mechanisms : Fast and low-cost local venues exist to resolve conflicts between users or with the authorities
  • 7. Minimum Recognition of Rights : The rights of local users to establish their own rules are recognized by the government
  • 8. Nested Enterprises : When a common resource is closely linked to a larger socio-ecological system, governance activities are organized at multiple nested levels

2. Towards ecosystem based managmeent

Limits of single species approaches

Ecosystem based management

FAO, 2003:

An ecosystem-based fisheries approach strives to balance different societal objectives, taking into account both the knowledge and uncertainties about the biotic, abiotic, and human components of ecosystems and their interactions, and applying an integrated fisheries approach within ecological limits.

Captures in multispecies fisheries

Tromeur et Doyen, 2018

Including different values

EBFM today in world’s fisheries

B. Conservation in agricultural settings

  • On land, land use change threatens species the most
  • Moreover, contracting with landowners may prove a more satisfactory option in a variable, dynamic context (as long as transaction costs are moderate)

1. The land sharing v. land sparing debate

Is it better to have a more intensive (with high yields) agricultural system and set more land aside?

That’s land sparing

Or to have an less intensive (i.e extensive, with lower yield) agricultural system with less land aside?

That’s land sharing

That all depends on the relationship between land, yield and biodiversity :

  • Knowing that increasing land yields to lower yield
  • But less land cultivated increases biodiversity
  • What should you do if the relationship between yield and biodiversity is :
    • Linear
    • Convex
    • Concave

Sparing

Sharing

Controversy

  • Green et al. (2005), Phalan et al. (2011) : use empirical data that the relationship tends to be convex, promoting land sparing

  • Sparing tends to be favored in several ecosystems :

  • However :

    • All biological diversities may not be impacted alike : soil v. floral and animal biodiversity
    • Not all impacts (especially dynamic) are factored in with intensive agriculture (pesticide use, local pollution, nutrient runoff)
    • The tradeoff can be potentially solved : how?

  • Using agroecology :

    • To leverage biodiversity
    • Into producing larger yields
    • Such as in India etc

2. Leveraging models to guide policy : bioeconomic modeling

  • Use the framework originated by Clark see Jean and Mouysset, IRERE, 2022
  • But broaden the scope to include:
    • A definition of biodiversity with multiple species
    • Where land use choices are defined by economic principles
    • And affect species dynamics
  • To evaluate how a policy maker can build policies to improve biodiversity outcomes

V. Cocco et al. 2023, Relaxing the production-conservation trade-off: Biodiversity spillover in the bioeconomic performance of ecological networks, Ecological Economics

3. Using PES for incentive based conservation

  • Idea is to leverage incentives to promote biodiversity enhancing policies
  • In the form of Payment for Ecosystem Services : habitat provision etc
  • Farmers are being paid by a third party to change their agricultural practices, land use type (fallow etc)
  • Sometimes, this can be leveraged into a compensation, or offset, market:
    • Either through a government policy, i.e. need to pay for the offsetting of a residual impact : 81,000 km\(^2\) in 2018
  • Which rests upon 2 assumptions :
    • Comparability : aim for “like for like” compensation, within the same geographical region
    • Additionality : offsets are officially achieved, and in the absence of a payment, there would not have been offsets
  • In practice :
    • Biodiversity assessment is simplified and not thorough, as well not ambitious enough to achieve Not Net Loss
    • Additionality is overestimated

A cautionary tale : the Verra case

  • World leading carbon offset standard
  • Claimed to provide standards that could guarantee the additionality of carbon sequestration to provide offsets
    • In a counterfactual scenario, some share of land would have been deforestated
    • In the actual scenario, because of the PES, it’s not
  • The main question is : how robust is the counterfactual scenario?
    • That’s the main question of causal inference
  • Turns out : most of the credits where not additional

  • Question is : how do you estimate additionality of certified carbon sequestration projects?
  • Several issues :
    • No observation of counterfactual
    • Have several projects with very specific characteristics : not easy to design a control/treated group study
    • Not a lot of statistical units
  • Use the synthetic control method (See Abadie, Journal of Economic Literature, 2021
    • Solve the issue of small number of units in observational studies
    • By using a synthetic control i.e. :
      • Find weights that combine similar units
      • To replicate existing trends in the treated unit prior treatment
      • And use weight to simulate counterfactual post treatment

T. West et al., Action needed to make carbon offsets from forest conservation work for climate change mitigation, Science, 2023

  • Evaluate 26 REDD+ programs across 6 countries
    • Reducing Emissions from Deforestation and Forest Degradation : framework developped under UN Framework Convention on Climate Change launched in 2008
    • Aims to provide incentives in voluntary schemes to reduce deforestation
    • The “+” integrates conservation, sustainable management and enhanced carbon stocks
  • Using the synthetic control method

C. Protected areas

The seminal idea is basically : how would you fill Noah’s Ark?

  • That’s Martin Weitzman’s approach in The Noah’s Ark Problem, Econometrica, 1998
  • Using a philogenetic tree as the basis for diversity :
    • Value of a collection
    • Is the value of different features (set is union of all features)
  • Assuming no interactions among species, or that we value the collection as the sum of its parts

\[ R_i = (D_i + U_i)\frac{\Delta P_i}{C_i}\]

  • Conserving unit \(i\) stands for conserving a species \(i\)
  • Project \(i\) increases the probability of survival of a species \(i\) by \(\Delta P_i\) at a cost of \(C_i\)
  • With \(U_i\) the value we place on species \(i\)’s existence
  • \(D_i\) is the distance from the closest resembling species
  • \(R_i\) is the ranking : “expectde marginal distinctiveness plus utility per dollar”

1. But in practice?

Ando et al, Species Distributions, Land Values, and Efficient Conservation, Science, 1998

Use cost-effective analysis to guide conservation endeavors

  • Need to buy land that can support various species, with differentiated land prices?
  • How can I spend the minimal amount while protecting all species : the Set Coverage Problem

\[\begin{align*} \min \sum_{j \in J}c_jx_j\\ \text{such that}\\ \sum_{j\in N_i}x_j\geq 1 \forall i \in I \end{align*}\] With \(J\) the index set of reserve sites, \(I\) index set of species to be covered, \(N_i\) the subset of \(J\) containing \(i\) and \(x_{j}=1(j \text{ is selected})\)

  • How can I protect the maximal number of species while on a budget :the Maximal Coverage Problem

\[\begin{align*} \max \sum_{i \in I}y_i &\\ \text{such that } \\ \sum_{j \in N_i}x_j\geq y_i \forall i \in I \\ \text{ and } \sum_{j \in J}c_jx_j\leq b \end{align*}\]

2. Biodiversity hotspots

  • Introduced by Norman Myers in 1988, the concept aims to identify regions such that :
    • Spots present an exceptional endemic species richness
    • While being overly threatened by habitat degradation (70% of loss of primary vegetation)
  • Allows to prioritize conservation:
    • Focus resources on critical areas
    • With people benefiting from those
    • To optimize conservation investments : maximize bang for buck
    • For NGOs such as Conservation International (one of world’s largest NGOs, $ 9bn in assets)

3. Efficiency of protected areas

  • Within the protected areas, species richness and abundance tend to recover, both on land and at sea
  • Not always the case : places protected need to be of high ecological value
    • Recent results (Grupp et al, 2024) with the European Union, which protects 26% of land, close to the 30 by 30 target, especially through Natura 2000
    • Use staggered difference in difference (Callaway & Sant’Anna, 2021)
    • Show that the results on vegetation cover and nightlight pollution are close to 0, irrespective of the time of conservation

Callaway & Sant’Anna, 2021

Two way fixed effects (TWFE) have been showed to be biased in the case of staggered treatment, because the standard estimator compares treated units to already treated units

TWFE : \(y_{it} = \mu +\alpha_i + \lambda_t + \delta D_{it} + \beta x_{it} + \epsilon_{it}\)

With \(\alpha_i\) a unit-specific fixed effect (capturing time-invariant differences across units) and \(\lambda_t\) a time-specific fixed effect (capturing shocks common to all units at a given time).

Their approach

  • Finds an group time Average Treatement effect on the Treated
  • Comparing each cohort \(g\) (treatment starts in \(g\)) to not yet treated units at time \(t\)
  • And aggregates them over time to reflect the composition of the treated across time to find the ATE
  • For marine protected areas :
    • When monitoring and sanctioning are well enforced, see less harvest in protected areas and promote recovery (See Sala and Giakoumi, 2018)
    • Otherwise, results are more debated
  • Besides efficiency, the political economy of MPAs matter : the Blue (or Green) Paradox
    • Announcing a policy that will force harvesters to reduce their harvest in the futures pulls it back in time, resulting in increased pressure until implementation
    • McDermott, Meng et al., 2018, study the Phoenix Islands Protected Areas (Pacific Ocean), one of the world’s largest
    • Find that fishers doubled their effort once area was earmarked for protected status
    • Amounting to an empoverished starting point equivalent to 1.5y of banned fishing

4. Issues with protected areas

  • Approach relies on a interspecies comparison
  • Subject to more than 1 period constraint :
    • Dynamic problem of land allocation
    • Especially with climate change
    • And exogenously shifting location-specific probabilities for survival
    • Integer programming problem becomes more difficult to solve
  • May fail to integrate as such the drivers of biodiversity loss and forgets what people do
    • Political economy
    • Local populations not always associated
  • May be beneficial if strong monitoring and transfer of changed values (e.g. tourism etc) to local communities

Conclusion

  • Biodiversity underpins the functioning of essential ecosystem services
  • And is dramatically threatened by anthropogenic impacts :
    • Overexploitation
    • Habitat loss
  • Solving overexploitation of species
    • Implies regulating fisheries to promote sustainable use
    • But also feature larger ecosystem based goals to avoid dire ecological consequences
  • Adopting a more comprehensive approach is thus necessary
  • On land, policies aim at protecting habitat
    • Either through vertical policies : CAP incentives
    • Or more horizontal policies through voluntary PES
    • In more or less anthropized landscapes
  • The efficiency of policies is encouraging but can be improved with
    • More monitoring & communities empowerment
    • Better targeting, especially in Europe
    • Better forecasting for PES and favoring in kind compensation
    • Less red tape to avoid green paradox
    • and renewed commitment and effort for policy evaluation with causal methods