,,,,,,,,,,,,,,,,,NATURE.impact_std,,,,,,NCP.impact_std,,,,,,GQL.impact_std,,,,,,OTHER.impact_std,,,,,,Positive,Negative,
,,,,,,,,,# PAs included in sample,,,,,,,,count,49,,,,,count,9,,,,,count,25,,,,,count,7,,,Total,68,57,13,
,,,,,,,,over 100,10,,,,,,,,positives,0.87755102,,,,,positives,1,,,,,positives,0.6,,,,,positives,0.285714286,,,Single-outcome studies,49,38,8,
,,,,,,,,over 100,2,,,,,,,,mixed,0.06122449,,,,,mixed,0,,,,,mixed,0.2,,,,,mixed,0.285714286,,,Multi-outcome studies,19,14,1,
,,,,,,,,median,56,,,,,,,,negative,0.06122449,,,,,negative,0,,,,,negative,0.2,,,,,negative,0.285714286,,,,,,,0.210526316
,,,,,,,,,,,,,,,,,% of total,0.544444444,,,,,% of total,0.1,,,,,% of total,0.277777778,,,,,% of total,0.077777778,,,,,,,
Order,Study,First author,Title,DOI / ISBN,Year,Name of Protected Area,Country/Region,"Biome? (select: Temperate, tropical, arid boreal, arid; AND Terrestrial, freshwater, marine, coastal)",# PAs included in sample,Study_obj.env,Study_obj.socioeconomic,Comparisons,Comp_type,Design type (from keywords tab),Design (from methods of paper; snippet of text),NATURE (stated in authors' words),NATURE.impact_std,,NATURE.impact,NATURE.measured,NATURE.mechanisms,NCP. (stated in authors' words),NCP.impact_std,,NCP.impact,NCP.measured,NCP.mechanisms,GQL(stated in authors' words),GQL.impact_std,,GQL.impact,GQL.measured,GQL.mechanisms,OTHER(stated in authors' words),OTHER.impact_std,,OTHER.impact,OTHER.measured,OTHER.mechanisms,outcomes_count,outcomes_pos,outcomes_neg,outcomes_balance
44,1,Alva,Marine protected areas and children’s dietary diversity in the Philippines,https://doi.org/10.1007/s11111-015-0240-9,2015,# Country level study,Philippines,"Tropical, marine",526,0,1,1,"Intervention, distance to MPA, user groups",regression analysis,see outcome 1,,,,,,,,,,,,,GQL / children’s dietary diversity - proximity to children’s community,positive,1,« positive association between MPAs and children’s dietary diversity when the MPAs were located closer than 2 km to a child’s community »,"data from the 2008 Philippines Demographic and Health Survey  (13,594 women aged 15–49 interviewed -> 4,382 children aged 6–59 months in 2008; use of WHO definition of minimum dietary diversity) and MPA data from the Coastal Conservation and Education Fund (distance to a child’s community, MPA creation year, size, management); weighted logistic regression model based on the survey regression techniques to account for the sampling design of the survey (logit)","assumption : « proximity to effective MPAs will have a positive effect on children’s dietary diversity as a result of fish abundance »- no measurement of fish abundance
« MPA characteristics such as age or type of management were not consistently associated with dietary diversity. »",,,,,,,1,1,0,
45,2,Andam,Measuring the effectiveness of protected area networks in reducing deforestation,https://doi.org/10.1073/pnas.0800437105,2008,# country level study,Costa Rica,"Tropical, terrestrial",N.A.,1,0,1,inside and outside,"empirical evaluation, matching methods",see outcome 1,NATURE / 1) Deforestation - Avoided deforestation estimates 2) Deforestation spillover,positive,1,"1) « approximately 10% of the protected forests would have been deforested had they not been protected » 2) negligible « Given the weak evidence for spillover effects from Costa Rica’s protected areas, we conclude that the matching estimates in Table 2 and Fig. 1 reflect the full effect of protected areas on deforestation within and outside protected areas between 1960 and 1997. »","PA between 1960 and 1997 (broken in 3 cohorts of time); forest-cover data (1960, 1986 and 1997, satellite and aerial photographs); biais control (eg. land productivity from soil, climate, topography; distance to road; to cities and forest edge) ; deforested,  80% canopy cover ; comparison with traditional methods; avoided deforestation - Sensitivity Test to Hidden Bias  - robust conclusions, only strong hidden bias could change them",,,,,,,,,,,,,,,,,,,,1,1,0,
33,3,Andam,Protected areas reduce poverty in Costa Rica and Thailand,https://doi.org/10.1073/pnas.0914177107,2010,# country level study,Costa Rica,"tropical, terrestrial",N.A.,0,1,1,Inside and outside,quasi-experimental matching,see outcome,,,,,,,,,,,,,"GQL / Poverty - ""impacts of protected area systems on poverty""",positive,1,"« although communities near protected areas are indeed substantially poorer than national averages, an analysis based on comparison with appropriate controls does not support the hypothesis that these differences can be attributed to protected areas. In contrast, the results indicate that the net impact of ecosystem protection was to alleviate poverty.»
« ∼10% of the poverty reduction observed in treated segments over time is attributable to protected areas. »","cofounders : preprotection poverty, forest cover, land productivity, and access to transportation and market infrastructure (control variables) (more details available if needed) ; measures based on national census data of household characteristics and assets (2000) -> poverty index at census segment level (tract) ; PA created by 1980 at the latest, area chosen from segment with 10% or more of PA (control less than 1%) (category I to IV)","""our analysis does not elucidate the specific mechanisms through which protected areas may have reduced poverty. We speculate that benefits to local residents have included tourism business opportunities, investments in human and physical capital by national and international agents, and the maintenance of ecosystem services (39, 45, 46). Research to understand these mechanisms is a clear future priority.""
robustness check: seems to be no population displacement, no negative spillover (more likely positive, no hidden biaises",,,,,,,1,1,0,
34,3,Andam,Protected areas reduce poverty in Costa Rica and Thailand,https://doi.org/10.1073/pnas.0914177107,2010,# Subnational region level study,North and Northeastern Thailand,"tropical, terrestrial",N.A.,0,1,1,Inside and outside,quasi-experimental matching,see outcome,,,,,,,,,,,,,"GQL / Poverty - ""impacts of protected area systems on poverty""",positive,1," although communities near protected areas are indeed substantially poorer than national averages, an analysis based on comparison with appropriate controls does not support the hypothesis that these differences can be attributed to protected areas. In contrast, the results indicate that the net impact of ecosystem protection was to alleviate poverty.» 
« ∼30% of the counterfactual poverty level, which is represented by the mean poverty headcount ratio for the matched control subdistricts. »","cofounders : preprotection poverty, forest cover, land productivity, and access to transportation and market infrastructure (control variables) (more details available if needed) ; measures based on national census data of household characteristics and assets (2000) -> poverty headcount ratio ; subdistrict ; PA created by 1985 at the latest ; subdistrict with 10% or more of PA (control less than 1%)","""our analysis does not elucidate the specific mechanisms through which protected areas may have reduced poverty. We speculate that benefits to local residents have included tourism business opportunities, investments in human and physical capital by national and international agents, and the maintenance of ecosystem services (39, 45, 46). Research to understand these mechanisms is a clear future priority.""
robustness check: seems to be no population displacement, no negative spillover (more likely positive, no hidden biaises",,,,,,,1,1,0,
114,4,"Arriagada, R. A., Echeverria, C. M., & Moya, D. E. (2016). Creating Protected Areas on Public Lands: Is There Room for Additional Conservation? PLOS ONE, 11(2), e0148094. https://doi.org/10.1371/journal.pone.0148094",Creating Protected Areas on Public Lands: Is There Room for Additional Conservation?,https://doi.org/10.1371/journal.pone.0148094,2016,(8 selected PAs in Chile),Chile,"Tropical, terrestrial",8,1,,1,"compares PAs to unprotected private lands, unprotected public lands",,"Matching (""i) comparison of means, (ii) conducting statistical matching, (iii) adjusting for remaining bias post-matching, and (iv) testing for unobservables that may bias causal estimates... We used a simple difference-in-differences (DID) estimator (called the Before-After-Control-Impact estimator in the ecology literature), which can control for time-invariant unobservable characteristics... we chose one-to-one, nearest-neighbor covariate matching with replacement using a generalized version of the Mahalanobis distance metric and genetic matching algorithm that maximizes covariate balance... Matching estimators of PAs’ impacts may still be biased due to discrepancies between the covariates of the matched protected pixels and their unprotected matches. We reduced this bias by using regression methods for the matched data... We used the sensitivity test
proposed by [Rosenbaum] and based on the Wilcoxon statistical test that assumes that each observation unit has a fixed value of one unobserved covariate (or a compound of unobserved covariates)."")",NATURE / Deforestation,positive,1,4.7% - 4.8% of protected plots would have been deforested by 2011 in the absence of protection,Remote sensing (Landsat),,,,,,,,,,,,,,,,,,,,1,1,0,
10,5,"Bauch, S.C.; Birkenbach, A.M.; Pattanayak, S.K.; Sills, E.O.",Public health impacts of ecosystem change in the Brazilian Amazon,https://www.pnas.org/content/pnas/112/24/7414.full.pdf,2005,# Subnational region level study,Brazil Amazon,"Tropical, terrestrial",N.A.,0,1,1,"Across different PA (strict, sustainable use) and sites (indigenous lands)",random-effects and quantile regression models,"(a) Datasets on (i) public health data (malaria, acute respiratory infection (ARI), and diarrhea + control), (ii) climate (2 seasons) and biophysical characteristics (temperature, precipitation, altitude, and water bodies), (iii) demographic and socioeconomic characteristics (population, migration, health services, and income) and (iv) land use drivers (roads, mining, and three types of PAs) ; (b) Timeline and scale: 4 years 2003-2006 at municipality level (700 municipalities) ; (c) PA: strict, sustainable use + indigenous lands ; (d) Statistical mathods: random-effects and quantile regression models",,,,,,,,,,,,,GQL / health,mixed,-1,"Strict PAs are negatively correlated with the three major diseases: malaria, diarrhea, and acute respiratory infection (ARI). 
Sustainable-use PAs, which allow for human use and/or occupation, are not associated with diarrhea and ARI and are positively correlated with malaria. 
Indigenous reserves are negatively correlated only with diarrhea, possibly reflecting differences in health outcomes among the indigenous populations who live in those reserves.
Impacts depending on land-use drivers ""Conservation scenarios based on estimated regression results suggest that malaria, ARI, and diarrhea incidence would be reduced by expanding strict PAs, and malaria could be further reduced by restricting roads and mining.""","Health: annual municipality-level observations on cases of malaria, dengue, AIDS, arthritis, and leukemia in the population and cases of diarrhea and ARI in children under 2 years of age; negative control (dengue, AIDS, arthritis, and leukemia) ; land use drivers: see design","""This may be because of the combined effects of reduced de-forestation and reduced exposure, which mean that strict PAs can serve as a barrier to disease transmission"" + ""This supports the contention that roads can plausibly both damage ecosystems and improve access to formal health services."" + complete set of assumptions available
Other covariates: a) ""The number of migrants is positively correlated with each disease"" 
b) ""Population is negatively correlated with diseases"" 
c) ""Natural water bodies are positively correlated with all three diseases, altitude is negatively correlated with malaria, and in the summer higher temperature and rainfall (which could, for example, flush pathogens down river) are negatively correlated with all three diseases""",,,,,,,1,0,1,
46,6,"Börner, J., Schulz, D., Wunder, S., & Pfaff, A. (2020). The Effectiveness of Forest Conservation Policies and Programs. Annual Review of Resource Economics, 12(1), annurev-resource-110119-025703. https://doi.org/10.1146/annurev-resource-110119-025703",The Effectiveness of Forest Conservation Policies and Programs,https://doi.org/10.1146/annurev-resource-110119-025703,2020,#Meta-analysis,Global,(multiple),Not stated,1,,1,(varies by study),,"counterfactual-based: ""After applying the selection criteria, we identified 99 studies reporting counterfactual-based treatment effects of conservation mechanisms. Within these studies, we were able to calculate 198 effect sizes from 51 papers. After taking means as described above, the final number of comparable estimates was 136. One study (Abman 2018) yielded a total of 68 country-specific effect sizes for PAs, but our results are robust to excluding this subsample of PA effect estimates.""",NATURE / avoided deforestation,positive,1,"""PAs and PES are moderately effective"" - Cohen's d of 0.1-0.2 which ""are considered small effects. However, similar ranges were reported for interventions in other sectors, such as for education interventions in developing countries""","(varied by study, most used remote sensing)","Fig 4 shows IPLs are several times more effective than PAs - ""Designation of indigenous areas is on average the most effective intervention in our sample. Using analysis of variance, we found that the difference to both PES and PAs was significant at the 95% level of confidence (Figure 4). The estimates for demarcation of indigenous territories require cautious interpretation because of the unique legal status of these lands in Latin America. All observations are from Brazil, where indigenous land differs from nature reserves with sustainable use. On other continents, indigenous communities often have privileged access to areas legally defined as nature reserves, merging the effectiveness of indigenous demarcation into the categories “Sustainable Use” or “PAs,” the latter being predominantly strict protection.""",,,,,,,,,,,,,,,,,,,1,1,0,
11,7,"Canavire-Bacarreza, G.; Hanauer, M.M.",Estimating the impacts of bolivia's protected areas on poverty,https://doi.org/10.1016/j.worlddev.2012.06.011,2013,# Country level study (23 PA at national level but authors calculated them differently in the study; see design),Bolivia,"Depending on the location, terrestrial",23,0,1,1,Inside and outside,matching,"(a) data: (i) temporally distinct boundary mappings of terrestrial protected areas; (ii) boundary mappings of municipalities for the 1992 and 2001 censuses and the underlying demography, and; (iii) key biophysical characteristics believed to jointly affect the establishment of protected areas and poverty (e.g. road networks, forest cover, elevation and slope, distance to major cities); (b) PA = municipalities with 10% or protected land (56 out of 307); (c) Unit of analysis: municipalities",,,,,,,,,,,,,GQL / poverty reduction,positive,1,« municipalities with at least 10% of their area occupied by a protected area established between 1992 and 2000 exhibited differentially greater levels of poverty reduction between 1992 and 2001 compared to similar municipalities unaffected by protected areas »,"1992 and 2001 census data (municipalities boundaries, demography, etc.) -> estimation of a poverty index and Necesidades Basicas Insatisfechas (percentage of the population within a municipality with unsatis
fied basic human needs)",further studies needed,,,,,,,1,1,0,
47,8,"Carranza, T., Balmford, A., Kapos, V., & Manica, A. (2014). Protected Area Effectiveness in Reducing Conversion in a Rapidly Vanishing Ecosystem: The Brazilian Cerrado. Conservation Letters, 7(3), 216–223. https://doi.org/10.1111/conl.12049",Protected Area Effectiveness in Reducing Conversion in a Rapidly Vanishing Ecosystem: The Brazilian Cerrado,https://doi.org/10.1111/conl.12049,2014,# Subnational region level study; 107 out of 116,"Cerrado, Brazil","tropical, terrestrial",107,1,0,1,"presence and absence, across PA (strictly protected, multiple use) and sites (indigenous lands)",matching,"SIG : a) type of PA (strictly, SP - IUCN I to IV, multiple-use, MU - IUCN V to VI) + indigenous land ; b) forest cover = conversion data from the satellite imagery (from 2002 to 2009) ; c) counterfactual : points within 10 km ; Matching variables=predictors: measures of accessibility (distance to paved roads and nearest state or federal capital city), rainfall and agricultural suitability, vegetation type, elevation",NATURE / 1) lower habitat conversion 2) Leakage / Conversion in buffers vs. control areas,positive,1,"1) Positive absolute effect for PA overall and subsets (SP, MU, IL), same for networks; larger for SP than MU
Positive relative effect for overall PA and subsets (SP, MU, IL),  same for networks ; larger for SP than MU 
PAs and ILs broadly similar to those seen in comparable studies for tropical forests ; larger effect for SP, smaller for MU (pooled relative effects)
2) no higher conversion, no evidence of local-scale leakage","1) absolute effect measured as difference between control sample and matched PA; PA with the same weight or according to the native vegetation area (network) ; relative effect measured as difference divided by conversion in matched unprotected sites (i.e., the percentage of baseline conversion represented by the absolute effect) ; pooled relative effect measured as mean conversion found in the pool of matched control locations minus the mean conversion found in all matched locations within the network, and divided the result by the mean conversion found in the pool of matched controls
2) comparing conversion in a 10 km buffer around PAs or ILs with that in the buffer’s matched control sample","« This may be expected as multiple-use PAs are under less restrictive land-use rules. » « This is particularly true in Cerrado, as most MUs are classed as environmental protected areas (locally termed APAs), which broadly allow more land-use change than other MU categories, and are commonly used for buffering stricter reserves such as National Parks »
SP = Strictly protected PA
MU = multiple use PA
IL = indigenous lands",,,,,,,,,,,,,,,,,,,1,1,0,
48,9,"Cazalis, V., Princé, K., Mihoub, J.-B., Kelly, J., Butchart, S. H. M., & Rodrigues, A. S. L. (2020). Effectiveness of protected areas in conserving tropical forest birds. BioRxiv, 2020.01.21.912345. https://doi.org/10.1101/2020.01.21.912345",Effectiveness of protected areas in conserving tropical forest birds,https://www.nature.com/articles/s41467-020-18230-0,2020,"# No specific area; 8 tropical forest biodiversity hotspots : four in the Americas (Atlantic Forest, Tropical Andes, Tumbes-Chocó-Magdalena, Mesoamerica), one in Africa (Eastern Afromontane), and three in Asia (Western Ghats and Sri Lanka, Indo-Burma, Sundaland)",Global,"tropical, terrestrial",N.A.,1,0,1,inside and outside,Counterfactual analysis,"a) biodiversity database (2005-2008): ebird « world’s largest citizen science programme », « confounding effects (observer experience, sampling effort, seasonality) that can affect the length and composition of recorded bird lists » ; b) site characteristics: control for location biases (altitude, remoteness, agricultural suitability)",NATURE / 1) biodiversity « whether protected areas differ from unprotected sites in terms of their bird species diversity » 2) forest cover/quality  « effects of protected areas on forest presence […] or quality » compared to unprotected sites,positive,1,"1) « positive effects of protection on the diversity of bird species that are forest-dependent, endemic to the hotspots, or threatened or Near Threatened, but not on overall bird species richness. »  overall richness = not a suitable indicator of local biodiversity
2) « significant positive effects of protection on forest presence across all hotspots analysed, with a protected site having on average 17.8% higher probability of being forested than a non-protected counterfactual »
« a general positive effect of protection on forest quality »","1) 4 indices (overall species richness, specialists, endemic species, species classified as threatened or near threatened in the IUCN red list) 2) canopy height, forest contiguity and wilderness","1) « the effects of either forest presence (IIIa) or forest quality (IIIb) on each of the four above-mentioned indices of bird diversity: Our results suggest that protected areas effectiveness at retaining species of concern is mainly driven by their effectiveness at mitigating forest loss » ; « Stronger evidence of effectiveness for South American protected areas » perhaps results reflecting reductions in other types of pressures such as hunting, selective logging, or invasive species ; « This may reflect variation in the effectiveness of protected area implementation across the world, or simply differences in statistical power » 2) « Stronger evidence of effectiveness for South American protected areas » perhaps results reflecting reductions in other types of pressures such as hunting, selective logging, or invasive species ; « This may reflect variation in the effectiveness of protected area implementation across the world, or simply differences in statistical power » ",,,,,,,,,,,,,,,,,,,1,1,0,
49,10,Cernea and Schmidt-Soltau,Poverty risks and national parks: policy issues in conservation and resettlement,https://doi.org/10.1016/j.worlddev.2006.02.008,2006,# No specific area,Central Africa,"tropical, terrestrial",12,0,1,1,Before and after,model of impoverishment risks and reconstruction,IRR model data coming from 10 to 100 days in each site 1996-2005 (consultancy/research visits : while villages visited or representative samples ; literature review ; land cost estimation),,,,,,,,,,,,,"GQL / Security / 1) Percetions of facing the risk of landlessness / loss of common property
2) Perceptions of facing the risk of joblessness (loss of productive work, income and subsistence)
3) Facing the risk of homelessness
4) Rights a) Facing the risk of marginalization / b) Facing the risk of social disarticulation
5) Health // a) Facing the risk of food insecurity / b) Facing the risk of increased morbidity and mortality",negative,-2,"1) Increased « They commonly express the view that conservation has taken their forest and forced them into poverty » « there is hardly a substantive difference between the risk of losing land (or forest-land) and thus becoming landless, and the risk of losing the access to the common property resources from the forest, since the forest in its total meaning is both the ‘‘individual’’ and common property. »
2) increased (estimated annual income loss from hunting/gathering in Euros : 7,108,612 in total)
3) increased (no housing, infrastructure or social services)
4) a) marginalization: increased (for hunter-gatherers) b) social disarticulation:  (« impoverishment fact » fir hunter-gatherer societies)
5) a) food insecurity: mostly absent in the short run but basically by default; uncertain in the long run b) morbidity/mortality: not clear","1) involved 41 000 km2 and 54 000 persons ; look at existence of resettlement policy
2) « This research has reconstructed the pre- conservation income based on a livelihood survey in one of the remotest but un-conserved regions in Central Africa » (subsistence and cash loss from hunting and gathering)
3) interviews, « empirical evidence showing it is not happening »
4) interviews; literature review
5) interviews","1) losses neither compensated nor replaced by any alternative income source; with 2 exceptions, no assignement to new settlement areas
2) hypothesis  « They are not only vulnerable [mainly depending on susbistence income - gathering and hunting] but also very poor […] among the poorest populations in Africa and the world»
3) decrease in health (same habitations but for a different lifestyle: hunter-gatherer -> farmer) ; seen as strangers
4) loss of traditional rights
5) a) law not fully implemented
b) new settlement closer to formal health services and facilities but decrease in income make it impossible to pay for these services and medecine",,,,,,,1,0,1,
50,11,"Eklund, J., Blanchet, F. G., Nyman, J., Rocha, R., Virtanen, T., & Cabeza, M. (2016). Contrasting spatial and temporal trends of protected area effectiveness in mitigating deforestation in Madagascar. Biological Conservation, 203, 290–297. https://doi.org/10.1016/j.biocon.2016.09.033",Contrasting spatial and temporal trends of protected area effectiveness in mitigating deforestation in Madagascar,https://doi.org/10.1016/j.biocon.2016.09.033,2016,# No specific area; 39 (46 for second period),Madagascar,"tropical, terrestrial and marine",39,1,0,1,inside and outside; 2 distinct decades; type of forest,counterfactual methodology (new methods for large dataset),"forest cover: 1990, 2000 and 2010; three major forest types (humid, dry and spiny ; includes mangroves); Data: distance to roads, rivers, major cities and altitude, slope and annual rainfall, proxy for agricultural sustainability; PA : 1990 (or 2000) or earlier ; categories II (national parks) and IV (special reserves) ; Timeline:  2 periodes (1990-2000, 39 PA and 2000-2010, 46 PA); Method: Mahalanobis distance for large dataset; Wilcoxon signed ranks tests",NATURE / Forest cover,positive,1,"PAs were effective to some extent ; some of this decrease can be attributed to the presence of PAs, not just to the confounding factors rendering the land assigned for protection less likely to be deforested (humid> dry > spiny), equally effective in second period)  BUT:« Overall deforestation rates have decreased from the first to the second time periods for all forest types. » « The largest reduction in deforestation rates was in the humid forest »","forest cover: 1990, 2000 and 2010; three major forest types (humid, dry and spiny ; includes mangroves); ; layers from institutions, 30m spatial resolution
mechanisms: distance to roads, rivers, major cities and altitude, slope and annual rainfall, proxy for agricultural sustainability ; 90m resolution",deforestation mainly where PA pixels tend to be absent ; PA at high altitudes and steep slopes ; spiny forest : deforestation spreading up to altitude and steep slopes in later period;  in general lower deforestation in the later time period has meant that the PAs have less pressures to resist.,,,,,,,,,,,,,,,,,,,1,1,0,
88,12,"Eklund, J., Coad, L., Geldmann, J., & Cabeza, M. (2019). What constitutes a useful measure of protected area effectiveness? A case study of management inputs and protected area impacts in Madagascar. Conservation Science and Practice, 0(0), e107. https://doi.org/10.1111/csp2.107",What constitutes a useful measure of protected area effectiveness? A case study of management inputs and protected area impacts in Madagascar,https://doi.org/10.1111/csp2.107,2019,(35 PAs in Madagascar),Madagascar,"Tropical, terrestrial",35,1,0,1,Matching,,"Matching - compare each focal pixel (i.e., each sample pixel from inside a PA) to a group of pixels with similar covariate characteristics. [Rachel's note: the authors included both protected and non-protected pixels in the ""control"" groups, this seems like a strange approach as it's effectively comparing PA pixels to a combination of PA and non-PA pixels]",NATURE / Deforestation,positive,1,"29 out of the 36 PAs had an impact in mitigating deforestation within the PA borders;  PAs in Madagascar differ substantially in how much pressures they experience, some show high deforestation rates despite being effective compared to the counterfactual, whereas other show low deforestation while still effective  (earlier in the paper they say their sample included 35 PAs but this paragraph summarizing results says 36)",Remote sensing,We found no association between PAME (Protected Area Management Effectiveness scores) and avoided deforestation pressure.,,,,,,,,,,,,,,,,,,,1,1,0,
38,13,Ferraro,"More strictly protected areas are not necessarily more protective: evidence from Bolivia, Costa Rica, Indonesia, and Thailand",https://doi.org/10.1088/1748-9326/8/2/025011,2013,# Country level study,Bolivia,"tropical, terrestrial",N.A.,1,0,1,"Inside and outside, type of PA (strictly restricting human use, allowing some extractive use)",quasi-experimental design,see outcome 1,"NATURE / Deforestation - ""Which areas, on average, experienced more avoided deforestation?""",neutral,0,difference statistically insignificant,"pixel size (100m2), data on entire country, strict definition (IUCN II), less strict (VI - integrated management), forest cover period (1991-2000) and protection period (1992-2000), matching weighting = inverse covariance;""we define the strictness of protection based on de jure criteria; i.e., what the law says the regulations are. Thus the analyses in all four sites contrast protected areas that strongly restrict human use to protected areas that allow extractive uses.""",,,,,,,,,,,,,,,,,,,,1,0,0,
39,13,Ferraro,"More strictly protected areas are not necessarily more protective: evidence from Bolivia, Costa Rica, Indonesia, and Thailand",https://doi.org/10.1088/1748-9326/8/2/025011,2013,# country level study,Costa Rica,"tropical, terrestrial",N.A.,1,0,1,"Inside and outside, type of PA (strictly restricting human use, allowing some extractive use)",quasi-experimental design,see outcome 1,"NATURE / Deforestation - ""Which areas, on average, experienced more avoided deforestation?""",positive,1,"""Costa Rica, Sumatra and Thailand’s strictly protected areas experienced an estimated 10–13 percentage points less deforestation than less strictly protected areas."" ","pixel size (3 ha), data on entire country, strict definition (IUCN Ia, II, IV), less strict (VI), forest cover period (1960-1997) and protection period (1973-1980), matching weighting = Mahalanobis; ""we define the strictness of protection based on de jure criteria; i.e., what the law says the regulations are. Thus the analyses in all four sites contrast protected areas that strongly restrict human use to protected areas that allow extractive uses.""","« On average [the 4 countries], stricter protection reduced deforestation rates more than less strict protection, but the additional impact was not always large and sometimes arose because of where stricter protection was assigned rather than regulatory strictness per se.» ""strictness drives differences in Costa Rica.""",,,,,,,,,,,,,,,,,,,1,1,0,
40,13,Ferraro,"More strictly protected areas are not necessarily more protective: evidence from Bolivia, Costa Rica, Indonesia, and Thailand",https://doi.org/10.1088/1748-9326/8/2/025011,2013,# Subnational region level study,"Sumatra, Indonesia","tropical, terrestrial",N.A.,1,0,1,"Inside and outside, type of PA (strictly restricting human use, allowing some extractive use)",quasi-experimental design,see outcome 1,"NATURE / Deforestation - ""Which areas, on average, experienced more avoided deforestation?""",positive,1,"""Costa Rica, Sumatra and Thailand’s strictly protected areas experienced an estimated 10–13 percentage points less deforestation than less strictly protected areas."" - ""site selection constant, changing strictly protected areas to less strictly protected status would reduce the protected area impacts substantially (forgo 21 percentage points of avoided deforestation), whereas moving less strictly protected areas to stricter protection status would increase impacts substantially (19 percentage points increase in avoided deforestation)""","pixel size (1 km2), data on Sumatra, strict definition (IUCN Ia, II, IV), less strict (VI), forest cover period (2000-6) and protection period (prior to 2000), matching weighting = Mahalanobis; ""we define the strictness of protection based on de jure criteria; i.e., what the law says the regulations are. Thus the analyses in all four sites contrast protected areas that strongly restrict human use to protected areas that allow extractive uses.""","« On average [the 4 countries], stricter protection reduced deforestation rates more than less strict protection, but the additional impact was not always large and sometimes arose because of where stricter protection was assigned rather than regulatory strictness per se.» site selection constant, changing strictly protected areas to less strictly protected status would reduce the protected area impacts substantially (forgo 21 percentage points of avoided deforestation), whereas moving less strictly protected areas to stricter protection status would increase impacts substantially (19 percentage points increase in avoided deforestation)""",,,,,,,,,,,,,,,,,,,1,1,0,
41,13,Ferraro,"More strictly protected areas are not necessarily more protective: evidence from Bolivia, Costa Rica, Indonesia, and Thailand",https://doi.org/10.1088/1748-9326/8/2/025011,2013,# Subnational region level study,North and Northeastern Thailand,"tropical, terrestrial",N.A.,1,0,1,"Inside and outside, type of PA (strictly restricting human use, allowing some extractive use)",quasi-experimental design,see outcome 1,"NATURE / Deforestation - ""Which areas, on average, experienced more avoided deforestation?""",positive,1,"""Costa Rica, Sumatra and Thailand’s strictly protected areas experienced an estimated 10–13 percentage points less deforestation than less strictly protected areas."" - ""site selection constant, changing strictly protected areas to less strictly protected status would reduce the protected area impacts substantially (forgo 21 percentage points of avoided deforestation), whereas moving less strictly protected areas to stricter protection status would increase impacts substantially (19 percentage points increase in avoided deforestation)""","pixel size (30 m2), North and Northeastern part, , strict definition (IUCN I, II), less strict (VI - forest reserves), forest cover period (1973-2000) and protection period (prior to 1985), matching weighting = inverse covariance; ""we define the strictness of protection based on de jure criteria; i.e., what the law says the regulations are. Thus the analyses in all four sites contrast protected areas that strongly restrict human use to protected areas that allow extractive uses.""","« On average [the 4 countries], stricter protection reduced deforestation rates more than less strict protection, but the additional impact was not always large and sometimes arose because of where stricter protection was assigned rather than regulatory strictness per se.» site selection constant, changing strictly protected areas to less strictly protected status would reduce the protected area impacts substantially (forgo 21 percentage points of avoided deforestation), whereas moving less strictly protected areas to stricter protection status would increase impacts substantially (19 percentage points increase in avoided deforestation)""",,,,,,,,,,,,,,,,,,,1,1,0,
36,14,Ferraro,Quantifying causal mechanisms to determine how protected areas affect poverty through changes in ecosystem services and infrastructure,https://doi.org/10.1073/pnas.1307712111,2014,# country level study,Costa Rica,"tropical, terrestrial",N.A.,0,1,1,Inside and outside,quasi-experimental matching,"(a) unit of analysis: 2000 census tract ; (b) PA: established before 1980, « protected areas were in International Union for Conservation of Nature (IUCN) categories Ia, I, II, IV, and VI. As in ref. 12, a census tract is considered protected if at least 10% of its area is occupied by protected land of any IUCN category. »; (c) poverty: index derived from census data (1973 and 2000); (d) tourism mechanism: locations of entrances derived from GPS data, « A protected census tract is con- sidered exposed to a park entrance if it is occupied by a protected area in which at least one entrance was established. »; (e) land cover mechanism: Adam et al., 2008; (f) infrastructure mechanism: roadless volume between 1969 and 1991, « sum of the product of area and distance to the nearest road for every 1-ha parcel within the census tract »; (g) method: two-stage estimation procedure with confounders (baseline poverty, forest area, agricultural productivity classes, roadless volume, and distance to markets), one-to-one Mahalanobis near- est-neighbor covariate matching with a bias-adjustment procedure and counterfactual poverty estimation for protected census tracts had protection not impacted the mechanism",,,,,,,,,,,,,"GQL / Poverty  « Estimated poverty reduction that can be attributed to changes in infrastructure, tourism services, and other ecosystem services» 
a) Infrastructure
b) tourim
c) ecosystem services",positive,1,« the establishment of protected areas before 1980 in Costa Rica caused a 16% reduction in poverty in neighboring census tracts by the year 2000 » (Adam et al. (2012)),"Poverty index from Cavatassi et al., 2004 + indirect mechanisms measurement - 
a) « Infrastructure » : « We use changes in roadless volume (35) between 1969 and 1991 to capture the impact of protected areas on road infrastructure. Higher values of roadless volume imply a smaller road network within a given area (i.e., less infrastructure). »
c) « tourism and recreational services » :  « Our correlate is the establishment of a formal entrance for the protected area »
c) « other ecosystem services » : « we estimate the net effect on poverty from the changes in forest cover caused by protection between 1960 and 1986 »","« Although protected areas reduced deforestation and increased regrowth, these land cover changes neither reduced nor exacerbated poverty, on average. Protected areas did not, on average, affect our measures of infrastructure and thus did not contribute to poverty reduction through this mechanism »;
a) ""Changes in infrastructure or land cover contributed little, on average, to poverty reduction""
b) «Nearly two-thirds of the impact is causally attributable to opportunities afforded by tourism. The rest is attributable to unobserved mechanisms.»
c) ""Changes in infrastructure or land cover contributed little, on average, to poverty reduction.""",,,,,,,1,1,0,
12,15,"Ferraro, P. J.; Hanauer, M. M.",Protecting Ecosystems and Alleviating Poverty with Parks and Reserves: 'Win-Win' or Tradeoffs?,https://link.springer.com/article/10.1007/s10640-010-9408-z,2011,# Country level study,Costa Rica,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,quasi-experiment,"(a) data from Andam et al. (2008, 2010) (deforestation from 1960 to 1997) ; (b) PA established prior to 1980 (International Union for Conservation of Nature (IUCN) management categories Ia, I, II, IV and VI), PA if 10% of pixel occupied by protected land ; (c) census tract (1973 and 2000, poverty index) ; (d) variables (land productivity, distances to road, forest edge and cities) ; (d) scale : census tract",NATURE / avoided deforestation,positive,1,"a) ""we find that protected land parcels with high land use capacities display significantly higher levels of avoided deforestation (32.4%) than those with low capacities (9%)""
b) ""Avoided deforestation from protection on high-slope forest parcels is estimated to be 1.4%, which is significantly lower than the estimated avoided deforestation of 15.9% on low-slope parcels""
c) ""protected land parcels that are located further from one of Costa Rica’s three major cities experience significantly higher levels of avoided deforestation (15.3%) than parcels that are closer (5%)""
d) ""Avoided deforestation estimates are significantly higher on parcels that fall in census tracts with high percentages of agricultural workers (13.3%) compared to those in census tracts with lower percentages of agricultural workers (4.5%).""
e) ""Although we find the point estimates of avoided deforestation due to protection to be higher on land parcels that fall within census tracts with high levels of baseline poverty, the difference between high (11.6%) and low (8%) subgroups is statistically insignificant.""",see design and table 1 for more information on covariates description,"Comparison between high and low
a) Land use capacity - ""we find that protected land parcels with high land use capacities display significantly higher levels of avoided deforestation (32.4%) than those with low capacities (9%)""
b) Slope- ""Avoided deforestation from protection on high-slope forest parcels is estimated to be 1.4%, which is significantly lower than the estimated avoided deforestation of 15.9% on low-slope parcels""
c) Distance to major cities -  ""protected land parcels that are located further from one of Costa Rica’s three major cities experience significantly higher levels of avoided deforestation (15.3%) than parcels that are closer (5%)""
d) Agricultural workers - ""Avoided deforestation estimates are significantly higher on parcels that fall in census tracts with high percentages of agricultural workers (13.3%) compared to those in census tracts with lower percentages of agricultural workers (4.5%).""
e) Poverty - ""Although we find the point estimates of avoided deforestation due to protection to be higher on land parcels that fall within census tracts with high levels of baseline poverty, the difference between high (11.6%) and low (8%) subgroups is statistically insignificant.""\",,,,,,,GQL / Poverty alleviation,mixed,-1,"a) ""the results suggest that protection on high-capacity land may have exacerbated poverty (positive rather than negative ATT). In contrast, the poverty reduction impacts on low-capacity lands are quite large.""
b) ""poverty alleviation associated with protection is greater on census tracts with high average slopes than those with low average slopes.""
c) Protection ""yields higher socioeconomic impacts when located near cities.""
d) ""census tracts with high percentages of agricultural workers exhibited significantly lower socioeconomic outcomes due to protection (0.008) than did census tracts with low percentages of agricultural workers (−1.802).""
e) Same as for avoided deforestation",see design and table 1 for more information on covariates description,"Comparison between high and low
a) Land use capacity- ""the results suggest that protection on high-capacity land may have exacerbated poverty (positive rather than negative ATT). In contrast, the poverty reduction impacts on low-capacity lands are quite large.""
b) Slope - ""poverty alleviation associated with protection is greater on census tracts with high average slopes than those with low average slopes.""
c) Distance to major cities - Protection ""yields higher socioeconomic impacts when located near cities.""
d) Agricultural workers - ""census tracts with high percentages of agricultural workers exhibited signif- icantly lower socioeconomic outcomes due to protection (0.008) than did census tracts with low percentages of agricultural workers (−1.802).""
e) Poverty- Same as for avoided deforestation","Trade-offs: avoided deforestation, poverty alleviation",,,"Trade-offs « We demonstrate that these environmental and social impacts were spatially heterogeneous. Importantly, the characteristics associated with the most avoided deforestation [highly suitable lands for agriculture, far from major cities and infrastructure or where a high percentage of adults are employed in agriculture] are the characteristics associated with the least poverty alleviation. »","see design : deforestation from 1960 to 1997, poverty index from census data (1973, 2000)",see impact,2,1,1,0
13,16,"Ferraro, P.J.; Hanauer, M.M.; Miteva, D.A.; Nelson, J.L.; Pattanayak, S.K.;Nolte, C.; Sims, K.R. E.",Estimating the impacts of conservation on ecosystem services and poverty by integrating modeling and evaluation,https://doi.org/10.1073/pnas.1406487112,2015,# Country level study,Brazil,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,quasi-experiment,"(a) data, CO2, PA: see outcomes measures ; (b) ""one-to-one nearest-neighbor matching with replace- ment to select unprotected units that are similar to protected units on av- erage, based on observable covariates.""; (c) analysis of heterogeneity: semiparametric partial linear model with distance to cities and slope",,,,,,,NCP / CO2 Additionality,positive,1,"1,247.8 Mt CO2 y−1 ; « protected areas in these conservation hotspots [the 4 countries] have stored at least an additional 1,000 Mt of CO2 in forests and have delivered ecosystem services worth at least $5 billion »","(a) Carbon and CO2: global distribution of aboveground woody biomass data ->  conversion factors -> social cost of carbon: $5/t (or 100) ; 
(b) PA: protected area networks established between 2000 and 2008 in the Brazilian Amazon ; 
(c) Unit analysis:  1-km2 parcels selected at random from land that was forested in 2000.","Moderators of Impacts (spatial heterogeneity : slope and distance to cities""
a) CO2 and deforestation - For  full  details  on  data,  see  Nolte  et  al. (2013) 
b) CO2 and poverty - No data on poverty",,,,,,,,,,,,,1,1,0,
14,16,"Ferraro, P.J.; Hanauer, M.M.; Miteva, D.A.; Nelson, J.L.; Pattanayak, S.K.;Nolte, C.; Sims, K.R. E.",Estimating the impacts of conservation on ecosystem services and poverty by integrating modeling and evaluation,https://doi.org/10.1073/pnas.1406487112,2015,# Country level study,Costa Rica,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,quasi-experiment,see outcomes measures,,,,,,,NCP / CO2 Additionality,positive,1,"3.67 Mt CO2 y−1 ; « protected areas in these conservation hotspots [the 4 countries] have stored at least an additional 1,000 Mt of CO2 in forests and have delivered ecosystem services worth at least $5 billion »","(a) Carbon and CO2: global distribution of aboveground woody biomass data ->  conversion factors -> social cost of carbon: $5/t (or 100) ; 
(b) PA: protected area networks established before 1997 in Costa Rica ; 
(c) Unit analysis: 3-ha parcels selected at random from either forested or unforested land at the relevant baseline year (1960 or 1986).","Moderators of Impacts (spatial heterogeneity : slope and distance to cities""
a) CO2 and deforestation - Possible trade-offs : ""the locations where protected areas generate the most avoided deforestation [according to slope and distance to major cities] are not necessarily the areas with the highest carbon densities.""  (For  full  details  on  data,  see  Andam  et  al. (2008, 2010, 2013)
b) CO2 and poverty - Same results for the slopes and distances associated with the greatest poverty alleviation. (1973 census tracts (average area of 9 km2); poverty data from the 1973 and 2000 censuses.)",,,,,,,,,,,,,1,1,0,
15,16,"Ferraro, P.J.; Hanauer, M.M.; Miteva, D.A.; Nelson, J.L.; Pattanayak, S.K.;Nolte, C.; Sims, K.R. E.",Estimating the impacts of conservation on ecosystem services and poverty by integrating modeling and evaluation,https://doi.org/10.1073/pnas.1406487112,2015,# Country level study,Indonesia,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,quasi-experiment,see outcomes measures,,,,,,,NCP / CO2 Additionality,positive,1,"385.4 Mt CO2 y−1 ; « protected areas in these conservation hotspots [the 4 countries] have stored at least an additional 1,000 Mt of CO2 in forests and have delivered ecosystem services worth at least $5 billion »","(a) Carbon and CO2: global distribution of aboveground woody biomass data ->  conversion factors -> social cost of carbon: $5/t (or 100) ; 
(b) PA: protected area networks established between 1988 and 2008 in Indonesia ;
(c) Unit analysis: Indonesia, carbon: 463-m forest or peatland parcels selected at random from a 1988 land cover map. ","Moderators of Impacts (spatial heterogeneity : slope and distance to cities""
a) CO2 and deforestation - Possible trade-offs : ""the locations where protected areas generate the most avoided deforestation [according to slope and distance to major cities] are not necessarily the areas with the highest carbon densities."" 
b) CO2 and poverty - Same results for the slopes and distances associated with the greatest poverty alleviation (data from villages (average area of 22 km2); poverty data from the 2006 village census (PODES).)",,,,,,,,,,,,,1,1,0,
16,16,"Ferraro, P.J.; Hanauer, M.M.; Miteva, D.A.; Nelson, J.L.; Pattanayak, S.K.;Nolte, C.; Sims, K.R. E.",Estimating the impacts of conservation on ecosystem services and poverty by integrating modeling and evaluation,https://doi.org/10.1073/pnas.1406487112,2015,# Subnational region level study,North and Northeastern Thailand,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,quasi-experiment,see outcomes measures,,,,,,,NCP / CO2 Additionality,positive,1,"58.7 Mt CO2 y−1 ; « protected areas in these conservation hotspots [the 4 countries] have stored at least an additional 1,000 Mt of CO2 in forests and have delivered ecosystem services worth at least $5 billion »","(a) Carbon and CO2: global distribution of aboveground woody biomass data ->  conversion factors -> social cost of carbon: $5/t (or 100) ; (b) PA: protected area networks established before 1985 in Thailand ; (c) Unit analysis: 3-ha parcels selected at random from land that was forested in 1973 in northern and northeastern Thailand, where the majority of the protected areas are located.","Moderators of Impacts (spatial heterogeneity : slope and distance to cities""
a) CO2 and deforestation - Possible trade-offs : ""the locations where protected areas generate the most avoided deforestation [according to slope and distance to major cities] are not necessarily the areas with the highest carbon densities."" (For full details on data see Andam et al. (2010) and Sims (2010).
b) CO2 and poverty - Same results for the slopes and distances associated with the greatest poverty alleviation ( subdistricts (tambons; average area, 74 km2).)",,,,,,,,,,,,,1,1,0,
17,17,"Ferraro, P.J.; Hanauer, M.M.; Sims, K.R.E.",Conditions associated with protected area success in conservation and poverty reduction.,https://doi.org/10.1073/pnas.1011529108,2011,# Subnational region level study,North and Northeastern Thailand,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,Matching + LOESS + PLM,"(a) poverty: based on national census data of household characteristics and assets (Costa rica : 1973 and 2000 census tract poverty indices from a principal components analysis; Thailand: subdistrict poverty head- count ratio, which is the share of the population in 2000 with monthly household consumption below the poverty line; this information comes from a poverty mapping analysis); (b) PA: census tract or subdistrict as protected if at least 10% of its area is protected before 1980 (Costa Rica) or 1985 (Thailand; 249 census tracts and 192 sub- districts) ; Unprotected units, from which matched controls are selected, comprise units with less than 1% protected before 1980 or 1985 ; Protected areas comprise IUCN Categories I, II, IV, and VI in Costa Rica and International Union for Conservation of Nature (IUCN) Categories I and II in Thailand. ; (c) avoided deforestation and timeline: The unit of analysis for the deforestation data is a 3-ha land parcel (20,000 randomly selected) drawn from forested areas at base- line (Costa Rica in 1960 and Thailand in 1973). Each parcel is classified as deforested or forested by the end year (Costa Rica in 1997 and Thailand in 2000). ; (d) covariates: land productivity, distance to road, to major city, to forest edge, roadless volume, poverty baseline, slope, forest cover, etc.",NATURE / Avoided deforestation,positive,1,"""15% of protected forest in Thailand would have been deforested in the absence of protection""","see design and Adam & al., 2008",,,,,,,,GQL / Poverty traps,positive,1,"""no evidence that protected areas trap historically poorer areas in poverty. In fact, we find that poorer areas at baseline seem to have the greatest levels of poverty reduction as a result of protection""  (fyi poverty impacts in Adam & al., 2010: ""In Thailand, protected areas reduced poverty by about 30% "")","subdistrict poverty head- count ratio, which is the share of the population in 2000 with monthly household consumption below the poverty line; this information comes from a poverty mapping analysissee Adam & al., 2010",,Trade-offs: Avoided deforestation & poverty,,,"b) slope: ""negatively related to avoided deforestation and positively related to poverty reduction"" (more pronounced in Thailand)
c) distance to major cities: ""the largest reductions in poverty are observed at intermediate distances from major cities."" but doesn't apply for deforestation avoided ","ATT ; data : multiple spatial layers are used to create covariates for each census tract, subdistrict, or parcel (ex : land productivity, Distance to Forest Edge, to road or major city)",,2,2,0,2
18,17,"Ferraro, P.J.; Hanauer, M.M.; Sims, K.R.E.",Conditions associated with protected area success in conservation and poverty reduction.,https://doi.org/10.1073/pnas.1011529108,2011,# Country level study,Costa Rica,"Tropical, terrestrial",N.A.,1,1,1,Inside and outside,Matching + LOESS + PLM,"(a) poverty: based on national census data of household characteristics and assets (Costa rica : 1973 and 2000 census tract poverty indices from a principal components analysis; Thailand: subdistrict poverty head- count ratio, which is the share of the population in 2000 with monthly household consumption below the poverty line; this information comes from a poverty mapping analysis); (b) PA: census tract or subdistrict as protected if at least 10% of its area is protected before 1980 (Costa Rica) or 1985 (Thailand; 249 census tracts and 192 sub- districts) ; Unprotected units, from which matched controls are selected, comprise units with less than 1% protected before 1980 or 1985 ; Protected areas comprise IUCN Categories I, II, IV, and VI in Costa Rica and International Union for Conservation of Nature (IUCN) Categories I and II in Thailand. ; (c) avoided deforestation and timeline: The unit of analysis for the deforestation data is a 3-ha land parcel (20,000 randomly selected) drawn from forested areas at base- line (Costa Rica in 1960 and Thailand in 1973). Each parcel is classified as deforested or forested by the end year (Costa Rica in 1997 and Thailand in 2000). ; (d) covariates: land productivity, distance to road, to major city, to forest edge, roadless volume, poverty baseline, slope, forest cover, etc.",NATURE / Avoided deforestation,positive,1,"impacts in Adam et al., 2008 (""About 11% of the area protected in Costa Rica would have been deforested had it not been protected"")","see design and same method as Adam & al., 2008",,,,,,,,GQL / Poverty traps,positive,1,"""no evidence that protected areas trap historically poorer areas in poverty. In fact, we find that poorer areas at baseline seem to have the greatest levels of poverty reduction as a result of protection"" (fyi poverty impacts in Adam&al., 2010 ""Protected areas in Costa Rica accounted for about 10% of the poverty decline around the areas"")","dataset: national census tract poverty  indices (1973, 2000) from principal component analysis + covariates (more details in Adam & al., 2010 ; Adam & al., 2008)",,Trade-offs: Avoided deforestation & poverty,,,"a) ""The results confirm that avoided deforestation in Costa Rica is positive across observed baseline poverty values. In other words, protected areas did impose binding land-use restrictions."" ; ""The estimates suggest that protected areas achieved significant poverty reduction for most of the range above the median baseline poverty level""
b) slope: 'Protection on low-sloped land is associated with significant tradeoffs in joint outcomes.""
c) distance to major cities: ""there is a substantial overlap of poverty reduction and avoided deforestation (win–win) at intermediate distances (∼40–100 km)."" ","ATT ; data : multiple spatial layers are used to create covariates for each census tract, subdistrict, or parcel (ex : land productivity, Distance to Forest Edge, to road or major city)","""partly because they tend to be sited in areas with low agricultural potential and thus, low opportunity costs.""",2,2,0,2
51,18,"Gaveau, D. L. A., Epting, J., Lyne, O., Linkie, M., Kumara, I., Kanninen, M., & Leader‐Williams, N. (2009). Evaluating whether protected areas reduce tropical deforestation in Sumatra. Journal of Biogeography, 36(11), 2165–2175. https://doi.org/10.1111/j.1365-2699.2009.02147.x",Evaluating whether protected areas reduce tropical deforestation in Sumatra,https://doi.org/10.1111/j.1365-2699.2009.02147.x,2009,# Subnational region level study,"Sumatra and Siberut, Indonesia","tropical, terrestrial",N.A.,1,0,1,"inside and outside, distance to PA",propensity score matching and linear regression,"Sampling strategy : sites randomly chosen ""the 1264 sampled cells cumulatively captured 11% of the total 1990 forest cover making up the Sumatra-wide map"" ; Variables: deforestation see below as for leakage definition; PA: reserves created before 2000 (national parks, nature, wildlife and game reserves and recreational parks); Hydrological and conservation PAs were combined into one PA category ; inside PAs, in the 10 km adjacent area or in the wider landscape ; confouding variable: slope and elevation, distance to forest edge, distance to roads and distance to logging roads + factor for political province prior to matching (reducing the biais of unobserved socio-economic drivers of deforestation); statistical tests : Kolmogorov–Smirnov, t-tests ; Wilcoxon""",NATURE / Deforestation + leakage,positive,1,"25% of loss in forest cover ;41% in unprotected areas ; 5% in PA == PA effective and beneficial neighbourhood leakage effect, « During the period 1990–2000 deforestation rates were found to be lower inside PAs than in adjacent unprotected areas or in the wider landscape. Deforestation rates were also found to be lower in adjacent unprotected areas than in the wider landscape. »","Propensity score matching (effectiveness). Mapping tropical deforestation: 98 corresponding LANDSAT TM and ETM+ satellite images with a c. 800 m2 (28.5 · 28.5 m) resolution to map forest cover change from 1990 to 2000 across a total of 440,000 km2 ; forest : old-growth natural evergreen forest (canopy cover > 50%), either undisturbed or partially degraded by selective logging; ‘Deforestation’ is the long-term removal of old- growth natural evergreen forest cover;  The area of deforestation (1990–2000) was extracted for each cell, with values that ranged from 0% to 100% on a continuous scale. Neighbourhood leakage = deforestation displacement (short distance) ; adjacent unprotected areas as land lying within 10 km of PA boundaries to compare our results with the simple inside–outside comparisons",may be explained by an island-wide decreasing population growth effect near Sumatran PAs as human population moves closer to urban centres.,,,,,,,,,,,,,,,,,,,1,1,0,
53,19,"Geldmann, J., Barnes, M., Coad, L., Craigie, I. D., Hockings, M., & Burgess, N. D. (2013). Effectiveness of terrestrial protected areas in reducing habitat loss and population declines. Biological Conservation, 161, 230–238. https://doi.org/10.1016/j.biocon.2013.02.018",Effectiveness of terrestrial protected areas in reducing habitat loss and population declines,https://doi.org/10.1016/j.biocon.2013.02.018,2013,# No specific area,Global,"Depending on the location, terrestrial",N.A.,1,0,1,"inside and outside (immediate surroundings or unprotected areas), type of PA, before and after",systematic literature review of publications with counterfactual,"a) peer reviewed and grey literature  (14 databases, eight specialist sources and 13 websites in English; list of relevant search terms and used Boolean operators and multi term searches; summer 2010) ; b) effectiveness of PA on targeted biodiversity ; c) before/after and control/intervention; d) causal link, appropriate counterfactual; e) information on biodiversity outcome variables (measurement methods, rates of change, units of measurement, direction of change, fraction of species doing better, trophic impacts) ; f) information on PA management interventions and characteristics, as well as external drivers of habitat or species change (direction or effect size) ; g) information on other biological and geo- graphical variables, and study biases.",NATURE / 1) habitat cover/protecting habitats  2) species populations,positive,1,"1) positive effect of PA : « there is good evidence that PAs have conserved forest habitat » ; Ratio between 1.25 and 22.7 times lower loss ; larger for Latina America and Africa ; possible effect size given the methods use; increase effectiveness when stricter protection (probably explained by larger size effect) and less effectivness in areas with more threats (ex : high human density) 
2) Inconclusive « evidence remains inconclusive that PAs have been effective at maintaining species populations, although more positive than negative results are reported in the literature »","1) 76 studies from 51 papers ; four were global, 35 evaluated regional, national or subnational networks of PAs, and 34 evaluated five or fewer PAs; mostly in Latin America, forest cover in tropical area, satellite remote sensing technique; 36 inside/outside, 21 compared similar areas outside PA, 10 matching
2) 42 studies from 35 papers ; Africa bias (57%), mammal (74%), species population abundance (34/42), 5 or less PA studied (83%), before/after (15), control/intervention (27)","No evaluation « causal connections between management inputs and conservation outcomes in PAs are rarely evaluated in the literature » ; « a limited evidence base, and weak understanding of the conditions under which PAs succeed or fail to deliver conservation outcomes. »",,,,,,,,,,,,,,,,,,,1,1,0,
54,20,"Geldmann, J., Coad, L., Barnes, M. D., Craigie, I. D., Woodley, S., Balmford, A., Brooks, T. M., Hockings, M., Knights, K., Mascia, M. B., McRae, L., & Burgess, N. D. (2018). A global analysis of management capacity and ecological outcomes in terrestrial protected areas. Conservation Letters, 11(3), e12434. https://doi.org/10.1111/conl.12434",A global analysis of management capacity and ecological outcomes in terrestrial protected areas,https://doi.org/10.1111/conl.12434,2018,# No specific area,Global,Terrestrial,73,1,1,1,Across PA,protected area management effectiveness (PAME),"a) bring together database on the Management Effectiveness Tracking Tool (METT - questionnaire - 1,988 PA - global quantitative data sets on management inputs  site-level management), and Living Planet Database (LPD - multi-source - 1736 PA - time-series of animal populations) -> overlap :  217 populations from 73 terrestrial PAs in 29 countries (biases:  older larger PA, Africa and Asia, mammals); b) contextual factors: PA attributes (size, date of establishment), human pressures (mean human influence index), socio-economic context (GDP, HDI 2000, HDI 2005), landscape structure (mean elevation); c) 4 models («each with population trend as the dependent variable, and one of the management dimensions as well as our random effects ») -> Model selection based on Akaike information criterion (AIC)",NATURE / vertebrate abundance,positive,1,"Positive for capacity and resources « a positive relationship between our METT-based scores for Capacity and Resources and changes in vertebrate abundance, consistent with the hypothesis that PAs require adequate resourcing to halt biodiversity loss »
No relationship for other dimensions","a) METT -> 30 questions (2003—2014) > composite score for 4 management dimensions (design and planning, capacity and resources, monitoring and enforcement systems, decision-making arrangements (stakeholders involvement)) ; b) LDP (1990-2012) : annual rate of change (i.e., the slope), by fitting a generalized linear regression model (GLM) with a log-link function","Role of management : capacities and resources (adequacy of staff, budgets, and available equipment);  other dimensions are important but might be related to other performance measures
PA size and age: « PA age was negatively correlated with trends for the mammal subsets and PA size negatively correlated with population trends in the global subset »  older PA in locations with less threats and conversely for younger PAs; larger parks leading to « dilution of resources, higher risks of of encroachment, and decreased detection of threats »",,,,,,,,,,,,,,,,,,,1,1,0,
55,21,"Geldmann, J., Manica, A., Burgess, N. D., Coad, L., & Balmford, A. (2019). A global-level assessment of the effectiveness of protected areas at resisting anthropogenic pressures. Proceedings of the National Academy of Sciences, 201908221. https://doi.org/10.1073/pnas.1908221116",A global-level assessment of the effectiveness of protected areas at resisting anthropogenic pressures,https://doi.org/10.1073/pnas.1908221116,2019,# No specific area,Global,Terrestrial,12315,1,1,1,"inside and outside, across sites (biome and realms)",Counterfactual,"a) Impact and management effectiveness THPI; b) PA: 2017 World database on PA, established before 1995 and not too small (compared to THPI resolution) ; c) METT assessment: questionnaire conducted between 2003 and 2010; d) Methods : Propensity score matching (+ coarsened exact matching (CEM) and assessing Mahalanobis distance matching (MDM)) : control variables (elevation, slope, access, temperature, precipitation, initial human footprint, country, land cover, soil type, nutrients levels) ; e) 6 realms : ) the Afrotropics, 2) Australasia, 3) Indomalaya, 4) the Nearctic, 5) the Neotropics, and 6) the Palearctic",,,,,,,,,,,,,,,,,,,DRIVERS / Human pressure,mixed,-1,"Lower increase inside PA over the last 15 y thanthe counterfactuals (Paleartic, Australasia, Neartic) ; « our results also confirm previous studies restricted to forest PAs, where pressures are increasing, but less than in counterfactual areas » Higher (Indomalaya, Afrotropics, Neotroprics) « on average, human pressures have increased inside PAs, with the greatest changes observed in the tropics, characterized by low HDI and low initial pressure  »  By pressure type:
1) smaller impacts of human population density  inside PAs compared to matched areas outside (lower for Afrotropics ; higher in Neotropics)
2) smaller impacts of night lights inside PAs compared to matched areas outside (higher in Neartic, Paleartic, Afroptropics, Indomalaya
3) Increased pressure from agriculture (Afrotropics, Indomalaya, Paleartic) ; « in tropical regions average pressure change from cropland conversion has increased inside PAs even more than in matched unprotected areas. »
Indistinguisable effect from agriculture (Australasia, Neartic, Neotropics)","Temporal Human Pressure Index THPI : composite index with human population density, density of night-visible infrastructure, % of area  under cropland + matching",« countries with high national-level development scores have experienced lower rates of pressure increase over the past 15 y within their PAs compared with a matched outside area »,1,0,1,
56,22,"Gill, D. A., Cheng, S. H., Glew, L., Aigner, E., Bennett, N. J., & Mascia, M. B. (2019). Social Synergies, Tradeoffs, and Equity in Marine Conservation Impacts. Annual Review of Environment and Resources, 44(1), 347–372. https://doi.org/10.1146/annurev-environ-110718-032344","Social Synergies, Tradeoffs, and Equity in Marine Conservation Impacts",https://doi.org/10.1146/annurev-environ-110718-032344,2019,"# No specific area; 64 out of 75 studies (also PES, environmental certi cation, and community-based management (CBM))",Global,Marine,,1 (indirect),1,0?,None,Literature review,"a) literature review (75 studies out of 2,619 screened articles);  b) Biases : Southeast Asia (36%), local-scale (93%), in or near tropical nearshore environment (53%), MPA (85% includes 12% CBMPA); c) not only PA : also environmental certification, community-based management (CBM); d) Outcomes : 195 identified ",,,,,,,,,,,,,"GQL / economic, health, political, education, social capital, and cultural",mixed,-1,"""From the 75 studies included in our review, we identified 195 combinations of outcomes that acted synergistically, antagonistically (i.e., tradeoff), or had unequal impacts across various domains of wellbeing, social/demographic groups, or scale. Within the literature, synergies and tradeoffs between wellbeing domains were pervasive, particularly among political empowerment, economic wellbeing, and social capital domains.""
More synergies than trade-offs ""Authors commonly reported synergies between income (economic wellbeing) and food security (mostly catch rates)"" ""mixed outcomes commonly occurred between the political empowerment, economic wellbeing, and social capital domains"" ""Authors most frequently described differential impacts by occupational groups such as tourism operators, net fishers, etc."" ""When considering a wide array of impacts across economic and ethnic groups, gender, or age, attempts to identify discrete winners and losers are obscured by the fact that synergistic gains, losses, and tradeoffs can stem from the same intervention."" 
significant knowledge gaps; ""Many conservation interventions reallocate property rights over marine resources, propagating a diversity of synergies and tradeoffs among other aspects of human wellbeing, particularly economic wellbeing and social conflict. »
main topics: political empowerment (resource rights = access, use, management rights), economic wellbeing (income), social capital and cohesion (changes in level of conflicts)","""we classify wellbeing outcomes into economic, health, political, education, social capital, and cultural domains. For every impacted wellbeing domain reported in the literature, we recorded observed synergies (positive or negative), tradeoffs, or cases where both synergies and tradeoffs occurred across multiple wellbeing domains.""",,,,,,,,1,0,1,
57,23,"Gill, D. A., Mascia, M. B., Ahmadia, G. N., Glew, L., Lester, S. E., Barnes, M., Craigie, I., Darling, E. S., Free, C. M., Geldmann, J., Holst, S., Jensen, O. P., White, A. T., Basurto, X., Coad, L., Gates, R. D., Guannel, G., Mumby, P. J., Thomas, H., … Fox, H. E. (2017). Capacity shortfalls hinder the performance of marine protected areas globally. Nature, 543(7647), 665–669. https://doi.org/10.1038/nature21708",Capacity shortfalls hinder the performance of marine protected areas globally,https://doi.org/10.1038/nature21708,2017,# No specific area; 433 (management) and 218 (ecological data),Global,"Temperate/tropical, marine",433,1,1,1,Across sites,Mixed methodologies,"(a) Management: to assess the efficacy and equity of MPA management processes: empirically supported governance and management theories to identify key management process indicators and data from 3 surveys (Likert-scale) -> binary threshold for effective management ; (b) Ecological impacts: MPA outcome data extracted from published studies(n = 40 MPAs, heavily centered on areas in the North Atlantic, US Pacific, and Australia) and transect- or site-level observations from unpublished regional and global datasets (n = 178 MPAs) -> logged response ratios (lnRR): comparison of mean fish biomass per unit of area in MPA and matched control site ; (c) Relationship between management and ecological data (n=62 MPA in 24 coutries): random forest and linear mixed-effects models to identify important management predictors",NATURE / MPA ecological outcomes,positive,1,"Postive and variable in magnitude; « Although 71% of MPAs positively influenced fish populations, these conservation impacts were highly variable. » (includes multi-use and restricted MPA)","MPA outcome data extracted from published studies(n = 40 MPAs, heavily centered on areas in the North Atlantic, US Pacific, and Australia) and transect- or site-level observations from unpublished regional and global datasets (n = 178 MPAs) -> logged response ratios (lnRR): comparison of mean fish biomass per unit of area in MPA and matched control site ; Relationship between management and ecological data (n=62 MPA in 24 coutries): random forest and linear mixed-effects models to identify important management predictors","« staff and budget capacity were the strongest predictors of conservation impact: MPAs with adequate staff capacity had ecological effects 2.9 times greater than MPAs with inadequate capacity. » ; other important variables : Clearly defined boundaries, and non management factors (MPA age and size, location (ecoregion, country), mean chlorophyll concentration, and mean shore distance)  «  these described processes have stronger, more direct impacts on resource users than on resource conditions, or that the indicators used in management assessments may imperfectly measure the governance and management processes from common pool resource theory »",,,,,,,,,,,,,INSTITUTIONS / **Procedural justice / MPA management processes,negative,-2,"negative « many MPAs failed to meet thresholds for effective and equitable management processes, with widespread shortfalls in staff and financial resources. » even if legal existence and appropriate regulations towards resource use (Few with scientific monitoring; Mainly state-managed (80%). Exception Oceania inclusive management and few legally gazetted MPA)",to assess the efficacy and equity of MPA management processes: empirically supported governance and management theories to identify key management process indicators and data from 3 surveys (Likert-scale) -> binary threshold for effective management,,2,1,1,0
19,24,"Gjertsen, H.",Can Habitat Protection Lead to Improvements in Human Well-Being? Evidence from Marine Protected Areas in the Philippines,dx.doi.org/10.1016/j.worlddev.2004.07.009,2005,Visayas region,Philippines,"Tropical, marine",40,1,1,1,Intervention,mixed methods,"(a) timeline: 2 years, 2000-2002, The fieldwork replicated a survey of the same sites in 2000 in order to obtain comparable information ; (b) community-based MPA :  mainly fishers population, community-based management, established between 1986-97;(c) measure of human welfare: village health officers data based on a regular weighing program of preschool children; (d) measure of ecosystem health: snorkel surveys (hard coral cover), interviews (perceived coral and fishes aboundance); (e) MPA management and contextual variables
: size and location of the MPA, penalties for violations, enforcement of the sanctuary rules, inputs to sanctuary management, and general village information, tourism, successful alternative income project, business owners, etc.",NATURE / ecosystem health (coral reefs),positive,1,« Community participation in determining the precise size and location of the MPA and financial or material municipal inputs both have a significant positive effect on reef health. » « Larger MPAs are significantly associated with decreased fish abundance in the MPA. » ,"snorkel surveys (hard coral cover), interviews (perceived coral and fishes aboundance)","« These results suggest that to improve reef conditions, the best options include providing alternative income projects and attractive wages to patrollers and charging high fines for violations [+ distance from municipality]. Once one controls for these key design variables and context, we find little evidence that developing tourism or marking MPA boundaries would further improve reef conditions appreciably. » « Community participation in determining the precise size and location of the MPA and financial or material municipal inputs both have a significant positive effect on reef health. »",,,,,,,GQL / child nutritional status,neutral,0,"""We have not found strong relationships between many of the MPA-related factors and child nutritional status (e.g., marking boundaries, paying patrollers, alternative income projects)"" -> no win-win situation with coral reef health",village health officers data based on a regular weighing program of preschool children,assumption MPA -> Habitat protection and fishery management tool -> income and fish consumption increase -> children health improvement,,,,,,,2,1,0,1
2,25,"Graham, N. A. J.; Ainsworth, T. D.; Baird, A. H.; Ban, N. C.; Bay, L. K.; Cinner, J. E.; De Freitas, D. M.; Diaz-Pulido, G.; Dornelas, M.; Dunn, S. R.; Fidelman, P. I. J.; Foret, S.; Good, T. C.; Kool, J.; Mallela, J.; Penin, L.; Pratchett, M. S.; Williamson, D. H.",From microbes to people: Tractable benefits of no-take areas for coral reefs,http://dx.doi.org/10.1201/b11009,2011,# No specific area,Global,"Tropical, marine",N.A.,1,1,1,"Inside/outside, before/after",empirical literature review,"(a) ""studies that adopt inside-outside or before-after designs""; no information on number of articles, 2005 reseach with key words ; (b)  looked at 5 broad categories: (1) tourism (, (2)  fisheries, (3) biodiversity, (4) ecological resilience, and (5) human well-being (human health,, values, empowerment)","NATURE / impacts on a range of ecological and social measures, from microbial communities to megafauna",positive,1,"Positive (very large to small and variable) : « increased abundance of target species of fishes and invertebrates, with smaller effects on the reef benthos, and variable effects on microbial communities, genetic diversity, marine megafauna »
Low : coral recovery (benthos)
Neutral : cetaceans (megafauna)
Negative : urchins and crown-of-thorns (motile invertebrates)",Literature review,"The effects of NTAs on fish, and to a lesser extent motile invertebrates, have received the most research attention. However, even within these groups, most studies assessed diversity, density and biomass, with less attention on trophic interactions, life history traits and adult and larval export. Surprisingly, the impacts of NTAs on the coral reef benthos are not well studied in most regions, and again those studies available focused on fairly simplistic measures, such as percentage cover. Effects on other aspects, such as megafauna abundance, microbes, genetic diversity and most social factors have also received limited attention.",,,,,,,"GQL / ""coastal societies"" ",neutral/positive,1,"Positive: overall increases in the economic revenue derived from reef-based tourism
Smaller: fisheries
Variable: various indicators of social well-being [values, human health, empowerment] 
Positive and negative effects expected under influence of climate change",,,,,,,,,2,2,0,2
58,26,Hanauer,Implications of heterogeneous impacts of protected areas on deforestation and poverty,https://royalsocietypublishing.org/doi/10.1098/rstb.2014.0272#d3e341,2015,"# Country level study (meaning ""all"" protected areas? ==123)",Bolivia,"Tropical, terrestrial",123,1,1,1,inside and outside,Quasi-experiment,"(a) PA : established between 1991 and 2000, a canton is (un)protected if at least 10% (less than 1%) of the canton is covered by one or more protected areas; (b) deforestation data: cover status between 1991 and 2006, from conservation international, 30m land parcel, control parcel never protected; (c) socioeconomic data:  1992-2012, from Bolivian National Statistical Office (INE), canton level, temporally comparable asset-based poverty index using data from the 1991, 2001 and 2012 censuses; (d) Moderating covariates:  (i) slope, (ii) distance to major city, (iii) baseline poverty and (iv) baseline proportion of indigenous peoples (+ distance to nearest road, elevation distance to the edge of the forest, roadless volume); (e) methods/Heterogeneity estimators: several non-parametric and semi-parametric econometric estimators; a multiple variable locally weighted scatter plot smoothing (i.e. three-dimensional LOESS) estimator, simple two-dimensional LOESS or partial linear model (PLM) approaches",NATURE / Deforestation,positive,1,"« Bolivia’s protected areas that were established prior to 2000 prevented approximately 1.9% of the protected forest from being deforested »; « protected areas avoided a modest amount of deforestation » (Ferraro et al., 2013)","deforestation data: cover status between 1991 and 2006, from conservation international, 30m land parcel, control parcel never protected + see design",« biophysical characteristics associated with the greatest avoided deforestation are the characteristics associated with the potential for poverty exacerbation from protection. » (less deforestation on flat and near cities land),,,,,,,GQL / Poverty + poverty traps,positive,1,"« Bolivia’s protected areas that were established prior to 2000 reduced poverty in surrounding communities between 1991 and 2012. » ; « reduced poverty in surrounding communities » (Canavire-Bacarreza and Hanauer, 2012) « Bolivia’s protected areas are not associated with poverty traps »","Creation of a poverty index from socioeconomic data: 1992-2012, from Bolivian National Statistical Office (INE), canton level, temporally comparable asset-based poverty index using data from the 1991, 2001 and 2012 censuses + see design","« However, protected areas tend to be placed relatively distant from major cities, where they have been less successful from a joint environmental and socioeconomic perspective. » ",**Equity,neutral,0,"« Overall, there does not appear to be evidence that protected areas had a differential impact on indigenous populations. »",,,3,2,0,2
59,27,"Herrera, D., Pfaff, A., & Robalino, J. (2019). Impacts of protected areas vary with the level of government: Comparing avoided deforestation across agencies in the Brazilian Amazon. Proceedings of the National Academy of Sciences, 116(30), 14916–14925. https://doi.org/10.1073/pnas.1802877116",Impacts of protected areas vary with the level of government: Comparing avoided deforestation across agencies in the Brazilian Amazon,https://doi.org/10.1073/pnas.1802877116,2019,# Subnational region level study,Brazil (Amazon),"Tropical, terrestrial",N.A.,1,0,1,"inside and outside, distance to PA (spillover), management scales",Matching,"(a) deforestation: using Programa de Cálculo do Desflorestamento da Amazônia (PRODES) data for 2000, 2004, and 2008 of the Brazilian space agency, Instituto Nacional de Pesquisas Espaciais (INPE) ; (b) PA: state or federal PA, indigenous lands; « we separated forests “in the arc of deforestation” (Rondônia, Mato Grosso, Pará, Maranhao, and Tocantins) from forests “not in the arc” (Acre, Amazonas, Roraima, and Amapá) » (no internal impacts or spillover expected, PA with « low pressure ») ; (c) Agencies: 3 types of agencies (federal indigenous lands, federal PAs, and state PAs);  (d) time period : « 2 time periods because other Amazonian policies (49) generated a fall in deforestation around 2004 » (2000-2004, 2004-2008); (e)Matching: land and site characteristics (slopes, soils, and characteristics such as distance to the nearest road, city or forest’s edge) ; propensity-score matching and covariate matching, plus postmatching regressions",NATURE / 1) Deforestation inside PA 2) Deforestation spillover,positive,1,"1) « Outside the region’s “arc of deforestation,” we confirm little internal impact »
2) no spillover outside the arc ; « For local spillover impacts, estimates for most arc states either are not significant or are not robust » « ; however, for Pará, federal PAs and indigenous lands feature both internal impacts and local spillovers. ","matching, inside PAs with the untreated or unprotected forests more than 30 km away",,,,,,,,,,,,,,,,,,,,1,1,0,
89,28,"Holland, M. B., de Koning, F., Morales, M., Naughton-Treves, L., Robinson, B. E., & Suárez, L. (2014). Complex Tenure and Deforestation: Implications for Conservation Incentives in the Ecuadorian Amazon. World Development, 55, 21–36. https://doi.org/10.1016/j.worlddev.2013.01.012",Complex Tenure and Deforestation: Implications for Conservation Incentives in the Ecuadorian Amazon,https://doi.org/10.1016/j.worlddev.2013.01.012,2014,Not stated,Ecuador,"Tropical, terrestrial",Not stated,1,0,1,"(1) Protected Areas (PAs), (2) forest patrimony areas (PF), (3) protected forests (BP), (4) indigenous lands, and (5) lands held privately (and combinations of these)",,"We use a random effects model to test forest change across different tenure categories, overlapping forms, and across two time periods (1990–2000 and 2000–08).",NATURE / Deforestation,positive,1,"(Complex results of multiple analyses, multiple tenure configurations, two time periods): over the 18-year period protected areas and all three categories of protected land that overlaps with indigenous communities saw significantly less deforestation than private-MAGAP lands after controlling for the location and proximity characteristics described above... this result was variable over time (two time periods)... The strictest protection (PA) and protection in areas with indigenous inhabitants avoided similar amounts of deforestation. Our models suggest these areas avoided roughly 100% more deforestation than private areas, on average.",Remote sensing,,,,,,,,,,,,,,,,,,,,1,1,0,
61,29,Kere,Addressing Contextual and Location Biases in the Assessment of Protected Areas Effectiveness on Deforestation in the Brazilian Amazonia.,https://doi.org/10.1016/j.ecolecon.2017.02.018,2017,# Subnational region level study,Brazil (Amazon),"Tropical, terrestrial",N.A.,1,0,1,"inside and outside, across different PA (sustainable use, integral use) and sites (indigenous lands), time of establishment",multilevel model and propensity score weighting,"(a) forest cover and PA: remotely sensed pixel data set from PRODES System of the Instituto Nacional de Pesquisa Espacial - INPE  (precision of 60 m approximately); PA established before 2004, distinguishing integral protection areas (IUCN I,II, III), sustainable use areas (IUCN IV, V, VI) and indigenous land ; (b) Location characteristics: slope, soil fertility, proximity to the nearest road, river and markets (towns) ; (c) Municipal context: legal amazon (nine states (Rondônia, Acre, Amazonas, Roraima, Pará, Amapá, Tocan- tins, and parts of Maranhão and Mato Grosso) and 788 municipalities); contextual effects : measuring the pressure on the resource (population density), the level of wealth (GDP per capita), forest cover and a proxy for biodiversity (the number of mammals) ; (d) timeline: 2005-2009",NATURE / Deforestation,positive,1,"a) Overall impact on deforestation : « Negative impact of protected areas on deforestation is consistent with results from previous studies. » (magnitude comparable - difference in geographical area considered) ; « these marginal effects are small when compared to the average deforestation rate observed in non-protected areas (4.44%). » ; « protection allow a decrease in the probability of deforestation of resp. 0.16%, 0.04% and 0.06% if the pixel becomes part of indigenous lands, sustainable use areas and integral PAs respectively. » ;  « The low level of marginal effects can be explained by the low rate of deforestation dur- ing the period under study: only 2.54% of the pixels have been cleared. » ;
b) Matched Unprotected versus Protected Areas: « protected areas have slowed down deforestation between 2005 and 2009, whatever the type of governance »; « indigenous protected areas are found to be marginally more efficient than sustainable use areas and integral use areas » Impact of Protected Areas by Establishment's Period (Before 2000 and Between 2000 and 2004): « Protected Areas that were more recently implemented are also found to avoid more deforestation than older ones » Multiple Treatments: confirm negative impact, indigenous lands to be more effective; no difference between the 2 PA","a) Multilevel nonlinear (logit) model + see design
b) Propensity score matching method + see design ","Main distance to the nearest road; « From a broader point of view, municipalities with high forest cover are far from main roads and main markets » « localization of recent PAs matters more than PAs seniority for protection purposes. This may suggest that recent PAs have a greater agricultural potential which means they are mostly located on fertile lands and near roads. » « This result can be explained by a better location of sustainable use areas which are characterized by higher fertility, lower slope, or shorter distance to cities. »
Type of PA/Site : « sustainable use areas not only help to avoid deforestation at the pixel level but also contribute to reduce the deforestation rate in the municipality. »;  « It is worth to remind that indigenous lands are more abundant and have the greatest marginal impact on the probability for a pixel to be cleared. »
Contextual effets: « A large forest area favors deforestation » not beyond 53% of forest in the municipality ; « A high level of income per head reduces deforestation », same as larger number of mammal species (no significant effect for population density)",,,,,,,,,,,,,,,,,,,1,1,0,
21,30,"Leverington, F.; Costa, K. L.; Pavese, H.; Lisle, A.; Hockings, M.",A global analysis of protected area management effectiveness,https://doi.org/10.1007/s00267-010-9564-5,2010,# No specific area,Global,N.A.,3038,1,1,1,Across PA,Cross-analysis,"We recorded over 8000 assessments from 100 countries, derived from more than 50 methodologies -> access results for 4092 evaluations from 3038 protected areas, 14 methodologies  -> 1800 indicators cross-analyzed through common scale and headlines indicators (statistical analysis) ; + reviewed and analyzed some 50 evaluation reports",NATURE / Conservation,positive,1,"For questions about whether protected areas are conserving their values, the overall mean score was 0.56, while the indicators relating to the contribution of protected areas to the well-being of their adjacent communities scored a mean of 0.58. For both these indicators, over half the assessments scored in the ‘sound’ range. These results indicate that in spite of deficiencies in inputs and management processes, many protected areas are achieving success in management. The positive scores might be questioned given that most assessments contributing to this study used primarily qualitative ratings, but there is no reason to believe that these indicators of outcome would be reported any more positively than other indicators","Defined by PAME as ""Conservation of nominated values—condition""","Factors most closely correlated with positive outcomes relating to values conservation included the skills of staff and other management partners, constraint or support by the external civil and political environment, achievements of outputs, adequacy of law enforcement. Though this indicates which aspects of management were most closely linked with overall performance, it should be stressed that these correlations do not necessarily establish causation.
Corrected correlations were not strong between the mean score of overall effectiveness and headline indicators relating to the conservation of values (0.37) or effect on the community (0.30). This indicates that while measures of planning, inputs and processes give us important information, they do not appear to be adequate predictors of successful outcomes.",,,,,,,GQL / Well-being of adjacent communities,positive,1,"For questions about whether protected areas are conserving their values, the overall mean score was 0.56, while the indicators relating to the contribution of protected areas to the well-being of their adjacent communities scored a mean of 0.58. For both these indicators, over half the assessments scored in the ‘sound’ range. These results indicate that in spite of deficiencies in inputs and management processes, many protected areas are achieving success in management. The positive scores might be questioned given that most assessments contributing to this study used primarily qualitative ratings, but there is no reason to believe that these indicators of outcome would be reported any more positively than other indicators","Defined by PAME as ""Effect of park management on local community""","Community–related indicators such as the involvement of communities and stakeholders, the communication program, and appropriate programs of community benefit were most highly correlated to the outcome measure of impacts on communities. Though this indicates which aspects of management were most closely linked with overall performance, it should be stressed that these correlations do not necessarily establish causation.
Corrected correlations were not strong between the mean score of overall effectiveness and headline indicators relating to the conservation of values (0.37) or effect on the community (0.30). This indicates that while measures of planning, inputs and processes give us important information, they do not appear to be adequate predictors of successful outcomes.",**Sustainability,positive,1,"Improvement over time: Where time-series data is available (Fig. 3), most protected areas are showing improvements in management, with some scores increasing dramatically. The concept of protected areas becoming “consolidated” through defining and working towards minimum standards of management across a number of factors makes intuitive sense and has been applied, for example in the Parks in Peril Program (The Nature Conservancy 2007). This process takes time, so long-term commitments to protected area improvement are essential, as are efforts to build sustainability into all externally-funded programs.",,,3,3,0,3
62,31,"Mammides, C. (2020). Evidence from eleven countries in four continents suggests that protected areas are not associated with higher poverty rates. Biological Conservation, 241, 108353. https://doi.org/10.1016/j.biocon.2019.108353",Evidence from eleven countries in four continents suggests that protected areas are not associated with higher poverty rates,https://doi.org/10.1016/j.biocon.2019.108353,2020,# No specific area,Global,"Depending on the location, terrestrial",N.A.,0,1,1,inside and outside,Quasi-experimental matching,"(a) Data on poverty : a large dataset on poverty and inequality by the Socioeconomic Data and Applications Center (SEDAC) (between 1992-2005, districts or municipalities levels)—available for 5800 administrative regions in eleven developing countries (Albania Bulgaria Ecuador Guatemala Madagascar Malawi Mozambique Nicaragua Vietnam Bangladesh Bolivia Ecuador Madagascar Malawi Mozambique) in four continents ; (b) Data on PA: World Database on Protected Areas (WDPA; October 2018 version); established after the poverty index was compiled within each country ; (i) strict reserves (categories Ia, Ib, and II); (ii) non-strict reserves (categories III–VI); and (iii) reserves with no category reported ; units as “protected” if at least 10% of their territory was covered by protected areas (Andam et al., 2010; Canavire-Bacarreza and Hanauer, 2013), and as “non-protected” if there were no protected areas at all. I excluded all units in-between (i.e., with>0% of protected land but<10%) to avoid weakening the effect of protected areas, if present; (d) Method: control variables (elevation and slope, distance to nearest major city, human footprint index e.g. including build-up areas, agriculture, infrastructures, population densities), matching, random forest tech- nique (Cutler et al., 2007), a supervised machine learning method, to run a regression model for each administrative level (variables : (i) mean elevation; (ii) mean slope; (iii) distance to the nearest major city; (iv) mean human footprint; (v) percentage of protected land within each unit; and (vi) country) ; “mahalanobis” distance (same variables as matching), sensitivity test (« logit » distance, calipers)",,,,,,,,,,,,,GQL / Poverty,neutral/mixed,0,"a) « When protected units were compared to all unprotected units within each country (i.e., during the naïve comparisons), results were statistically significant for five out of the eleven countries. In two of those countries, Malawi and Mozambique, the poverty rates were higher within protected regions. »
b) « Protected areas do not appear to be associated with higher poverty rates. »","used the “Foster-Greer-Thorbecke (FGT) poverty measure” (Foster et al., 2010), and particularly the “headcount index”, which was available for all countries.
a) naive comparisons
b) Matched comparisons",,,,,,,,1,0,0,
64,32,"Miranda, J. J., Corral, L., Blackman, A., Asner, G., & Lima, E. (2016). Effects of Protected Areas on Forest Cover Change and Local Communities: Evidence from the Peruvian Amazon. World Development, 78, 288–307. https://doi.org/10.1016/j.worlddev.2015.10.026",Effects of Protected Areas on Forest Cover Change and Local Communities: Evidence from the Peruvian Amazon,https://doi.org/10.1016/j.worlddev.2015.10.026,2016,(29 PAs in Peruvian Amazon),Peru,"Tropical, terrestrial",29,1,1,1,"Matching PAs and unprotected areas, also compared PAs of different ages and IUCN categories",,Matching - We use Mahalanobis covariate matching because it generates the best covariate balance statistics.,NATURE / Deforestation,positive,1,"protected areas reduce deforestation plus disturbance by 0.15 percentage points over a six-year period, or by 0.03 percent per year",Remote sensing,"Deforestation alone falls by 0.08 percentage points, or 0.01 percentage points per year... If the average deforestation rate in the unprotected portions of the study area is a good representation of Peruvian deforestation rate (about 0.19 percent), then we can say that protected areas has reduced deforestation by 8 percent within the six-year period (approx 0.15/0.19). Our results suggest that protected areas established before 1990 and nonstrictly protected areas are more effective in reducing deforestation. Protected areas established before 1990 cut deforestation by 0.13 percentage points (twice the ATT for the pooled sample), did not have a statistically significant effect on disturbance, and reduced and deforestation plus disturbance by 0.20 percentage points. Results for protected areas established after 1990 are not statistically significant. With regard to protection type, mixed-use protected areas reduced deforestation by 0.10 percentage points (slightly more than the ATT for the pooled sample) and deforestation plus disturbance by 0.22 percentage points. In the case of strictly protected areas, deforestation is not statistically significant, but disturbance is. Strictly protected areas actually increase disturbance by 0.07 percentage points and deforestation plus disturbance by 0.10 percentage points.",,,,,,,"GQL / economic /  per capita income and expenditure, poverty rate, and extreme poverty rate",mixed,-1,"For households living in a 5-km buffer, protected areas exacerbate extreme poverty. However, protected areas do not have a statistically significant effect on the other three socioeconomic indicators. For households living within a 10-km buffer, the effect of protected areas on extreme poverty vanishes. The effect on expenditure is positive and weakly significant. Hence, there is some indication, but certainly not a strong one, that protected areas may have adverse socioeconomic effects on local communities. We find that older protected areas exacerbate extreme poverty for households living in a 5-km buffer zone.  We find no strong evidence of heterogeneity treatment effects for households in the 10-km buffer zone.",Peruvian National Household Survey,,,,,,,,2,1,1,0
96,33,"Miteva, D. A., Murray, B. C., & Pattanayak, S. K. (2015). Do protected areas reduce blue carbon emissions? A quasi-experimental evaluation of mangroves in Indonesia. Ecological Economics, 119, 127–135. https://doi.org/10.1016/j.ecolecon.2015.08.005",Do protected areas reduce blue carbon emissions? A quasi-experimental evaluation of mangroves in Indonesia,https://doi.org/10.1016/j.ecolecon.2015.08.005,2015,(84 PAs),Indonesia,"Tropical, marine",84,1,0,1,compares Marine Protected Areas (MPAs) to Species Management Areas (SMAs) (another PA category),,"Matching, Difference-in-differences (propensity score and covariate matching (Mahalanobis distance matching with replacement), differences-in-differences, and post-matching bias adjustments)",NATURE / Mangrove loss,positive,1,"marine protected areas reduced mangrove loss by about 14,000 ha; no evidence that species management PAs stalled the loss of mangroves... overall PAs have a moderately statistically significant effect of reducing mangrove loss rates — on average, 10% reduction in mangrove loss from 2000 to 2006; but impacts differ by PA type and time period: MPAs reduced mangrove loss by 13% from 2000 to 2006. In contrast, SMAs did not have a statistically significant effect on mangrove outcomes; impacts are smaller if we consider a longer time span from 2000 to 2010.",Remote sensing,,NCP / Carbon emissions,positive,1,avoided blue carbon emissions of approximately 13 million metric tons (CO2 equivalent),"first estimate the hectares of mangroves loss prevented by PAs, and then combine this estimate with published data on emissions per hectare lost",,,,,,,,,,,,,,2,2,0,2
65,34,"Miteva, D. A., Murray, B. C., & Pattanayak, S. K. (2015). Do protected areas reduce blue carbon emissions? A quasi-experimental evaluation of mangroves in Indonesia. Ecological Economics, 119, 127–135. https://doi.org/10.1016/j.ecolecon.2015.08.005",Evaluation of biodiversity policy instruments: What works and what doesn’t?,https://doi.org/10.1093/oxrep/grs009,2012,(Review),(152 countries),"(Global, terrestrial)",Not stated (review),1,0,1,(Review: matching),,(Review),NATURE / Deforestation,positive,1,"(Range of impacts from 1-2 percent deforestation reduction (Costa Rica, Pfaff et al. 2009) to 24 percent (Sumatra and Siberut, Gaveau et al. 2009)",(Review),,DRIVERS / Fire,positive,1,"some reduction in forest fires, impacts vary by intervention, time period, and distance to major city",,,,,,,,,,,,,,,2,2,0,2
67,35,"Naidoo, R., Gerkey, D., Hole, D., Pfaff, A., Ellis, A. M., Golden, C. D., Herrera, D., Johnson, K., Mulligan, M., Ricketts, T. H., & Fisher, B. (2019). Evaluating the impacts of protected areas on human well-being across the developing world. Science Advances, 5(4). https://doi.org/10.1126/sciadv.aav3006",Evaluating the impacts of protected areas on human well-being across the developing world,https://doi.org/10.1126/sciadv.aav3006,2019,(more than 600 PAs globally),(34 developing countries),(Global),600,0,1,,,,"We used quasi-experimental hierarchical regression to isolate the impact of living near a PA on several aspects of human well-being. Within countries and DHS survey years, we used a genetic matching algorithm that weighted the matching covariates to achieve optimal balance",,,,,,,,,,,,,"GQL / 1) Wealth 2) Poverty 3) Health / Children's height-for-age growth scores, Stunting",positive,1,"We find that all else equal, a hypothetical move of rural households to within 10 km of PAs with documented tourist visits would result in significantly higher wealth scores (by 16.7% on average) and a lower likelihood of poverty (by 16.1%) compared to similar rural households living further than 10 km from a PA. These impacts rise to 20.1 and 25.7% for wealth and poverty likelihood, respectively, for a scenario where households shift to living close to multiple-use PAs (IUCN categories V and VI), rather than those under stricter protection (IUCN categories I to IV), where tourism has been documented. Similarly, a hypothetical shift to living near multiple-use PAs where tourism has been documented would, all else equal, increase children’s height-for-age growth scores by 9.8% and reduce the likelihood of stunting by 13.4%, compared to similar children living further than 10 km from a PA. The likelihood of poverty would also be 8.8% lower for households that shift to live near multiple-use PAs, even with no documented tourism at these PAs. In contrast, no early childhood growth gains were observed for scenarios where children hypothetically move close to PAs where no tourism has been documented, nor would wealth scores be higher in households moving adjacent to PAs without such tourism. There was also no evidence for any negative impacts of PAs on human well-being in any of our scenarios.",DHS surveys,,,,,,,,1,1,0,
68,36,"Negret, P. J., Di‐Marco, M., Sonter, L. J., Rhodes, J., Possingham, H. P., & Maron, M. (n.d.). Effects of spatial autocorrelation and sampling design on estimates of protected area effectiveness. Conservation Biology, n/a(n/a). https://doi.org/10.1111/cobi.13522",Effects of spatial autocorrelation and sampling design on estimates of protected area effectiveness,https://doi.org/10.1111/cobi.13522,2020,(116 PAs in Colombia),Colombia,"Tropical, terrestrial",116,1,0,1,"Matching PAs to unprotected areas, also compared PAs within different regions, and PA types (IUCN PA categories I-VI)",,Matching - we used propensity score matching to create control groups for each protected area cell in the country,NATURE / Deforestation,positive,1,"At the national level, areas under protection had 1.54% total forest loss compared with 2.61% for unprotected matched areas over the 15 year period. This represents 40% lower forest loss in protected areas than in matched unprotected areas. At the regional level, protection significantly reduced forest loss in the Amazon, Andes and Caribe regions. In contrast, protection significantly increased forest loss in the Orinoco and Pacific regions, with protection being most ineffective in the Pacific. ",Remote sensing,,,,,,,,,,,,,,,,,,,,1,1,0,
98,37,"Nepstad, D., Schwartzman, S., Bamberger, B., Santilli, M., Ray, D., Schlesinger, P., Lefebvre, P., Alencar, A., Prinz, E., Fiske, G., & Rolla, A. (2006). Inhibition of Amazon Deforestation and Fire by Parks and Indigenous Lands: Inhibition of Amazon Deforestation and Fire. Conservation Biology, 20(1), 65–73. https://doi.org/10.1111/j.1523-1739.2006.00351.x",Inhibition of Amazon Deforestation and Fire by Parks and Indigenous Lands,https://doi.org/10.1111/j.1523-1739.2006.00351.x,2006,(15 PAs),Brazil,"Tropical, terrestrial",15,1,0,1,"compares inside PAs to buffer area immediately outside, also compares PAs to inhabited reserves (indigenous lands, extractive reserves, national forests)",,Inside-outside (the ratio of deforestation and fire occurrence rates along the outside versus inside of the reserve perimeter),NATURE / Deforestation,positive,1,deforestation was 20 times higher outside reserves than inside,Remote sensing,,DRIVERS / Fire occurrence,positive,1,"average density of fires was 3.7 to 9.4 times higher along the outside of the reserves than along the inside (NOTE this result includes PAs, indigenous lands, national forests, not just PAs)",Remote sensing,,,,,,,,,,,,,,2,2,0,2
69,38,"Nolte, C., & Agrawal, A. (2012). Linking management effectiveness indicators to observed effects of protected areas on fire occurrence in the Amazon rainforest. Conservation Biology: The Journal of the Society for Conservation Biology, 27(1), 155–165. https://doi.org/10.1111/j.1523-1739.2012.01930.x",Linking management effectiveness indicators to observed effects of protected areas on fire occurrence in the Amazon rainforest,https://doi.org/10.1111/j.1523-1739.2012.01930.x,2012,(29 PAs in Amazon),"Bolivia, Peru, Brazil","Tropical, terrestrial",29,1,0,1,Matching PAs to unprotected areas,,matching to compare the estimated effect of protected areas with low versus high METT scores on fire occurrence. We divided our sample of protected areas with METT data into 2 groups—low and high composite METT scores—in order to compare their respective effects on fire occurrence. We used nearest‐neighbor matching (NNM) to estimate effects of protected‐area groups on fire occurrence. We used Mahalanobis distance NNM with replacement and bias adjustment. We examined the associations between METT scores and individual effects of protected areas on fire occurrence. ,,,,,,,DRIVERS / Fire,positive,1,"Protected areas with higher METT scores in 2005 did not seem to have performed better than protected areas with lower METT scores at reducing fire occurrence over the last 10 years. Protected areas with high and low METT scores reduced the occurrence of fires within their boundaries relative to similar unprotected areas, However, our results did not offer clear evidence that fire occurrence was lower in areas with high scores than in areas with low scores.",Remote sensing,"Many individual METT indicators did not exhibit observable or consistent differences between more effective and less effective groups. This was particularly true for management aspects traditionally assumed to be closely related to protected‐area effectiveness and thus to be classical targets of conservation investments, including adequacy (indicator 15) and security of budget (indicator 16), number of staff members (indicator 12), management plans (indicator 7), and boundary demarcation (indicator 6). The behavior of institutional variables was inconsistent: Although controlling access or use of the protected area (indicator 28) was positively associated with effectiveness, other variables such as mechanisms for controlling inappropriate land use and activities (indicator 2) and capacities and resources of staff to enforce regulations and legislation (indicator 3) had weak or negative associations with the relative effectiveness of protected areas in reducing fire occurrence.",,,,,,,,,,,,,1,1,0,
43,39,Pfaff,"Governance, Location and Avoided Deforestation from Protected Areas: Greater Restrictions Can Have Lower Impact, Due to Differences in Location",https://doi.org/10.1016/j.worlddev.2013.01.011,2014,# Subnational region level study,"Acre state, Brazil","tropical, terrestrial",N.A.,1,0,1,"inside and outside, type of PA (integral areas, sustainable us protection, indigenous lands)",Matching,see outcome 1,NATURE / Deforestation « Average protection’s forest impacts,positive,1,"« on average, protection in Acre tends toward lower clearing pressure, limiting deforestation impact » ; « Our matching (apples-to-ap- ples) impact estimate, based on unprotected land similar to protected land, suggests that a great deal of protected forest would have remained standing without policy »","sustainable use (IUCN V–VI), indigenous, integral governance (IUCN’s I–IV) ; deforestation during 2000–04 and 2004–08 (PRODES15 remotely sensed pixel data); location characteristics (distances to the forest’s edge, road and city + soil quality, rainfall, slope)","a) governance type: « Sustainable use protection targets areas with people, while integral protection seems to target an absence of local stakeholders. Thus, sustainable use protection occurs closer to clearing threats. ""deforestation in protected areas is lowest for integral areas and next lowest for indigenous lands and, finally, that it is highest for the sustainable use protection."" « We find that sustainable use protection, whose governance permits some local deforestation, is found on sites with high clearing threat. That allows more avoided deforestation than from integral protection, , which bans clearing but seems feasible only further from deforestation threats.» ; « Acre’s indigenous lands, which also permit livelihoods, did not have a significant impact, in contrast with other Brazilian Amazon findings »
b) distance to roads/cities: « For all governance types and for each type, protected areas closer to roads or cities avoided more deforestation than the distant protected areas »",,,,,,,,,,,,,,,,,,,1,1,0,
100,40,"Pfaff, A., Robalino, J., Sanchez-Azofeifa, G. A., Andam, K. S., & Ferraro, P. J. (2009). Park Location Affects Forest Protection: Land Characteristics Cause Differences in Park Impacts across Costa Rica. The B.E. Journal of Economic Analysis & Policy, 9(2). https://doi.org/10.2202/1935-1682.1990",Park Location Affects Forest Protection: Land Characteristics Cause Differences in Park Impacts across Costa Rica,https://doi.org/10.2202/1935-1682.1990,2009,(132 PAs),Costa Rica,"Tropical, terrestrial",132,1,0,1,compares PAs and matched unprotected areas,,Matching (propensity score matching),NATURE / Deforestation,positive,1,"PAs reduced deforestation by around 2% on average; PAs close to San Jose avoided over 4% deforestation, PAs close to national roads avoided 5% (versus 0 for PAs far from roads), PAs on low slopes avoided 14% of deforestation (vs 1% for PAs on high slopes over 7.12 degrees)",Remote sensing (Landsat 28 m resolution 1986 and 1997),,,,,,,,,,,,,,,,,,,,1,1,0,
79,41,"Rico, J., Panlasigui, S., Loucks, C. J., Swenson, J., & Pfaff, A. S. (2018). Logging concessions, certification & protected areas in the Peruvian Amazon: Forest impacts from combinations of development rights & land-use restrictions. Working Papers.","Logging concessions, certification & protected areas in the Peruvian Amazon: Forest impacts from combinations of development rights & land-use restrictions",,2018,(17 PAs in Peru),Peru,"Tropical, terrestrial",17,1,0,1,"uncertified logging concessions, FSC certification of logging concessions, PAs for strict conservation, and four types of multiple-use PAs",,Difference-in-differences,NATURE / Deforestation,positive,1,"on average logging concessions have no effect on tree-cover loss, while the PAs do reduce loss. Further, the PAs allowing limited private extraction save more forest than do more restrictive PAs.",Remote sensing (Hansen et al. 2013),,,,,,,,,,,,,,,,,,,,1,1,0,
26,42,"Robalino, J. and L. Villalobos-Flatt",Conservation Policies and Labor Markets: Unraveling the effects of national parks on local wages in Costa Rica,https://www.jstor.org/stable/resrep14924?seq=1#metadata_info_tab_contents,2010,# Country level study,Costa Rica,"Depending on altitude, terrestrial",28,0,1,1,"User groups, distance to the park/entrance",OLS and matching,"(a) PA : mostly established in the 70s ; (b) Timeline: 2000-2007 (census) ; (c) data set of workers:  census tracts (around 60 households per tracts), source = national institute of statistics and census; logarithm of hourly real wages, education level, gender, age, marital status, full-year employment, migration (2 years before census); (d) geographic characteristics :  distance to the park/entrance + average slope, average precipitation, and average elevation per census tract; (e) matching techniques: hourly real wages = dependent variable",,,,,,,,,,,,,GQL / Wages,positive,1,"positive for those « close to entrance » (6 % higher)
no impact for those « far from it »","- National Parks' Effects on Wages per Hour
- distance by road to park (5 km buffer) and their entrances (within or more than 20 km) for treated group; more than 15km for untreated group",results suggestion change in activities (higher paid : commercial and services) and higher wages for same activities near entrance,**Equity - National Parks’ Effects on Wages per Hour 1) by Subsamples (close to the entrance) 2)  by Economic Activity,neutral,0,"1) No better wages received by migrants (arrived in the previous 2 years); Positive for males and females (higher premium for the latest)
2) No difference for agricultural workers close to parks, wholesale and retail trade workers far from the entrance
Positive effect for restaurants and hotels for workers living close to entrances (12%), and wholesale and retail trade workers (9%)
Negative effect for restaurants and hotels for workers living far from entrances","wage premiums in the activities that will be more affected by land restrictions and tourism: agriculture, restaurants and hotels, and wholesale and retail trade",,2,1,0,1
80,43,"Robalino, J., & Villalobos, L. (2015). Protected areas and economic welfare: An impact evaluation of national parks on local workers’ wages in Costa Rica. Environment and Development Economics, 20(3), 283–310. https://doi.org/10.1017/S1355770X14000461",Protected areas and economic welfare: An impact evaluation of national parks on local workers’ wages in Costa Rica,https://doi.org/10.1017/S1355770X14000461,2014,(28 PAs in Costa Rica),Costa Rica,"Tropical, terrestrial",28,0,1,1,Matching,,"propensity score matching: We ran a probit regression in order to estimate the conditional probability of being assigned to each treatment group: being close to a park (CP), being close to the park and an entrance (CE), and being close to a park but far from the entrance (FE)...In order to account for any remaining differences in covariates, we run a standard OLS regression after matching, using only the matched observations to finally obtain the effects... the treatments we test are: (i) being close to the park; (ii) being close to the entrance of the park; and (iii) being close to the park but far from the entrance. We use the log of hourly real wages as a dependent variable.",,,,,,,,,,,,,GQL / Wages,positive,1,"on average, the effect on wages of being close to parks is positive and significant for local workers. The estimated effects range from 7.59 to 10.29 per cent. The estimated effects of living close to the entrance of a park range from 7.14 to 10.37 per cent. no robust evidence of positive or negative effects for workers living far from the entrance. although both females and males receive better wages close to park entrances, the premium for local females (from 15.52 to 20.18 per cent) is larger than the premium for local males (from 6.47 to 7.23 per cent)",Household surveys (Encuestas de Hogares de Propositos Multiples [EHPM] conducted by the National Institute of Statistics and Census of Costa Rica (Instituto Nacional de Estadıstica y Censos [INEC]) from 2000 to 2007),,**Equity,positive,1, the premium for local females (from 15.52 to 20.18 per cent) is larger than the premium for local males (from 6.47 to 7.23 per cent),,,2,2,0,2
102,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Brazil,"Tropical, terrestrial",Not stated,1,1,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",NATURE / Deforestation,positive,1,"(varies by study and PA type): Positive, Positive (older PAs); neutral (newer sustainable use PAs); positive (strict PAs), neutral (sustainable use PAs)",Not stated,,NCP / 1) Carbon storage 2) Fire,positive,1,1) Positive 2) Neutral,Not stated,Not stated,,,,,,,,,,,,,2,2,0,2
103,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Colombia,"Tropical, terrestrial",Not stated,1,0,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",NATURE / Deforestation,negative,-2,Negative,Not stated,,,,,,,,,,,,,,,,,,,,1,0,1,
104,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Peru,"Tropical, terrestrial",Not stated,1,1,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",NATURE / Deforestation & degradation,positive,1,"Positive for deforestation; degradation varies by study (Positive, neutral, positive)",Not stated,,,,,,,,GQL / Poverty: per capita income and expenditure,neutral,0,Neutral,Not stated,Not stated,,,,,,,2,1,0,1
105,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Ecuador,"Tropical, terrestrial",Not stated,1,0,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",NATURE / Deforestation,positive,1,Positive,Not stated,,,,,,,,,,,,,,,,,,,,1,1,0,
106,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Paraguay,"Tropical, terrestrial",Not stated,1,0,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",NATURE / Deforestation,positive,1,Positive,Not stated,,,,,,,,,,,,,,,,,,,,1,1,0,
107,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Chile,"Tropical, terrestrial",Not stated,1,0,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",NATURE / Deforestation,positive,1,Positive,Not stated,,,,,,,,,,,,,,,,,,,,1,1,0,
108,44,"Schleicher, J. (2018). The environmental and social impacts of protected areas and conservation concessions in South America. Current Opinion in Environmental Sustainability, 32, 1–8. https://doi.org/10.1016/j.cosust.2018.01.001",The environmental and social impacts of protected areas and conservation concessions in South America,https://doi.org/10.1016/j.cosust.2018.01.001,2018,(Review),Bolivia,"Tropical, terrestrial",Not stated,0,1,1,(review paper),,"(Review paper) Matching, matching & regression, Inside-outside, Regression, Before-after",,,,,,,,,,,,,GQL / Poverty,neutral,0,Neutral,Not stated,,,,,,,,1,0,0,
81,45,"Schleicher, J., Peres, C. A., & Leader‐Williams, N. (2019). Conservation performance of tropical protected areas: How important is management? Conservation Letters, 12(5), e12650. https://doi.org/10.1111/conl.12650",Conservation performance of tropical protected areas: How important is management?,https://doi.org/10.1111/conl.12650,2019,(43 PAs in Peruvian Amazon),Peru,"Tropical, terrestrial",43,1,0,1,Matching,,We applied nonparametric Wilcoxon and Spearman rank correlation tests to assess whether PA management indicators were associated with avoided deforestation and forest degradation,NATURE / Deforestation & forest degradation,positive,1,"while increasing PA management input to Amazonian PAs tended to reduce likelihoods of forest degradation and deforestation, the associations were weak. ",Remote sensing,"PA management was not strongly associated with impacts. Challenges identified by stakeholders: lack of law enforcement, corruption, overlaps with other land titles, and a general lack of respect attributed to land titles",,,,,,,,,,,,,,,,,,,1,1,0,
109,46,"Schleicher, J., Peres, C. A., Amano, T., Llactayo, W., & Leader-Williams, N. (2017). Conservation performance of different conservation governance regimes in the Peruvian Amazon. Scientific Reports, 7(1), 1–10. https://doi.org/10.1038/s41598-017-10736-w",Conservation performance of different conservation governance regimes in the Peruvian Amazon,https://doi.org/10.1038/s41598-017-10736-w,2017,(30 PAs),Peru,"Tropical, terrestrial",30,1,0,1,"Compares PAs to Indigenous Territories, Conservation Concessions, logging and mining concessions, and unprotected forest",,nearest neighbour matching without replacement using propensity scores,NATURE / 1) Deforestation 2) Forest degradation,positive,1,"1) PAs reduced deforestation by 100% (median) relative to matched unprotected landscape, and also significantly lower deforestation than matched areas in logging and mining concessions
2) PAs reduced forest degradation by 85.2% (median) relative to matched unprotected landscape","1) Remote sensing (Landsat / CLASlite system) with manual post-processing, error correction using imagery, and ground truthing
2) Degradation as classified by CLASlite system (Remote sensing (Landsat / CLASlite system) with manual post-processing, error correction using imagery, and ground truthing)",,,,,,,,,,,,,,,,,,,,1,1,0,
110,47,"Sims, K. R. E., & Alix-Garcia, J. M. (2017). Parks versus PES: Evaluating direct and incentive-based land conservation in Mexico. Journal of Environmental Economics and Management, 86, 8–28. https://doi.org/10.1016/j.jeem.2016.11.010",Parks versus PES: Evaluating direct and incentive-based land conservation in Mexico,https://doi.org/10.1016/j.jeem.2016.11.010,2017,All PAs in Mexico created before 2010 (names not listed),Mexico,"Tropical, terrestrial",Not stated,1,1,1,Compares PAs to PES (unsure which dropdown option this corresponds to),,"""our empirical strategy relies on comparisons of changes in outcomes in the 2000s between localities with different shares of land protected. Specifically, we model changes over time in the outcomes in the past decade as a function of the share of land treated during this period, controlling for state fixed effects, pre-trends, and geographic characteristics that determined selection criteria. This identifies impacts based on a comparison of changes in outcomes between localities with similar baseline characteristics and pre-trends but with greater versus less share of area with protected status during the past decade. Given remaining concerns about unobservable confounders, we check for potential differences in pre-trends and run multiple robustness checks, including estimating bounds under the assumption of remaining omitted variables.""",NATURE / net change in forest cover,positive,1,24% increase in net forest cover 2000-2012,Hansen et al 2013,,,,,,,,GQL / change in local poverty alleviation index,mixed,-1,no significant affect overall: there is a significant decrease in poverty alleviation for localities with a greater share in strict protected areas; a positive but not significant increase in poverty alleviation for localities with greater share in biosphere reserves; and a negative but not statistically significant change in poverty alleviation for mixed-use area,"A weighted average of indicators including rates of literacy, primary schooling, availability of potable water, sanitation and electricity, and housing characteristics (source: CONAPO)",Not stated,DRIVERS / population,mixed,-1,"Average estimates of both types of protection on population trends suggest that PES and mixed-use protected areas have led to decreases in population (-0.042 and -0.049) while strict protected areas and biosphere reserves have not. However, the significance of the population result on PES is not robust to several of the specification checks.","hundreds of people per square km (CONAPO) --  not sure whether an increase (or decrease) in population would be considered ""positive"" or ""negative"" anyway - the authors don't seem to have an opinion. ""We include population density growth as an outcome because of heated prior debates about the effects of parks on population trends (e.g. see Wittemyer et al., 2008 and response letters) and to test whether poverty alleviation impacts might be explained by migration.""",Not stated,3,1,2,-1
29,48,"Sims, KRE",Conservation and development: Evidence from Thai protected areas,dx.doi.org/10.1016/j.jeem.2010.05.003,2010,# Subnational region level study (31 wildlife sanctuaries and 57 national parks),North and Northeastern Thailand,"Tropical, terrestrial ",88,1,1,1,"Intervention, across sites",Regression frameworks,see outcome 1,NATURE / Forest cover,positive,1,"""the results demonstrate that protected areas did significantly increase forest cover, preventing clearing that otherwise would have taken place and imposing a binding constraint on land use » (moderate magnitude "" The results show that protected areas did significantly constrain forest clearing, reducing the amount of land available for agriculture by approximately 11% for a change from no protection to the median share protected (one-third)."") ; 
« The national forest reserves also appear to have had smaller effects on forest cover than the national parks or wildlife sanctuaries »","wildlife sanctuaries and national parks (cat. I and II, strictly PA); data on socioeconomic outcomes from a poverty mapping study with satellite-based estimates of forest cover; «poverty mapping analysis for the year 2000 by Healy and Jitsuchon (combines data from household consumption or expenditure surveys and census survey, 4113 subdistricts)"" + ""The poverty headcount ratio, poverty gap and squared poverty gap are part of the Foster–Greer–Thorbecke (FGT) family of poverty measures"" ; two regression frameworks = standard ordinary least squares regression model and  instrumental variable model (with priority watershed status)",,,,,,,,GQL / Poverty  « Did Thai protected areas exacerbate poverty? »,positive,1,"« protected areas increased average consumption and lowered poverty rates, despite imposing binding constraints on agricultural land availability. Socioeconomic gains are likely explained by increased tourism in and around protected areas. However, net impacts are largest at intermediate distances from major cities, » ; (""The OLS estimates suggest that an increase in the share of locality land protected from zero to the median share among those protected (one-third) corresponds to a 4.5% increase in monthly household consumption and a 10.3% decrease in the poverty headcount ratio. "")
greater impact in national parks than wildlife sanctuaries","wildlife sanctuaries and national parks (cat. I and II, strictly PA); data on socioeconomic outcomes from a poverty mapping study with satellite-based estimates of forest cover; «poverty mapping analysis for the year 2000 by Healy and Jitsuchon (combines data from household consumption or expenditure surveys and census survey, 4113 subdistricts)"" + ""The poverty headcount ratio, poverty gap and squared poverty gap are part of the Foster–Greer–Thorbecke (FGT) family of poverty measures"" ; two regression frameworks = standard ordinary least squares regression model and  instrumental variable model (with priority watershed status)","tourism offsetting opportunity costs ; migration (outruled) and spillovers (negligible) ; « The history of designation suggests that locations were more likely to be protected if they had higher historical forest cover, were important for watershed protection, were further from high quality agricultural land, further from mineral and timber resources, and were closer to national borders. »",,,,,,,2,2,0,2
111,49,"Smallhorn-West, P. F., Weeks, R., Gurney, G., & Pressey, R. L. (2020). Ecological and socioeconomic impacts of marine protected areas in the South Pacific: Assessing the evidence base. Biodiversity and Conservation, 29(2), 349–380. https://doi.org/10.1007/s10531-019-01918-1",Ecological and socioeconomic impacts of marine protected areas in the South Pacific: Assessing the evidence base,https://doi.org/10.1007/s10531-019-01918-1,2020,(65 MPAs),"(6 countries in the South Pacific): 
Fiji, Vanuatu, Tuvalu, New Caledonia, French Polynesia, Tonga","Tropical, marine",65,1,1,(varies by study),(varies by study),,(synthesis),"NATURE / Ecological outcomes: Habitat - coral cover, Habitat - algal cover, Total fish biomass, Total fish density, Total fish diversity, Target fish biomass, Target fish density, Herbivore biomass, Target invertebrate density, ",mixed,-1,"Based on 52 identified studies, 42% of measured ecological impacts were positive.","The most frequently measured variables were total fish diversity and target fish biomass, other measured variables included habitat (coral cover, algal cover), total fish biomass, total fish density, total fish diversity, herbivore biomass, and target invertebrate density (NOTE: ecological and socioeconomic outcomes were broken up into categories, so we could report on each but there are 9 ecological and 5 socioecomic so that would add a lot of outcomes)","No-take MPAs had a greater number of positive ecological impacts than periodic closures and there was little evidence of any long-term ecological recovery within periodic closures following harvesting. Both centrally governed and community-based MPAs had similar percentages of positive ecological impacts (48% and 43% respectively). The greatest percentage of neutral and negative ecological impacts was for periodic closures (71%), which were implemented only under community-based governance approaches. When centrally governed MPAs failed to achieve positive impacts, it was generally suggested that the reasons were environmental (e.g. sediment discharge from a river mouth) or biological (e.g. changing predator dynamics). In contrast, when community-based MPAs failed to achieve positive impacts, factors most often suggested were related to reserve design (e.g. close to human populations), management (e.g. lack of compliance), or social constraints (e.g. poacher aggression).",,,,,,,"GQL / Socioeconomic outcomes: Catch, Economic outcomes, Resource management decision making, Perceptions of ecological change, Perceptions of socioeconomic change",positive,1,"72% of socioeconomic impacts were positive, these were from only eight studies. All five categories (see next column) had generally positive impacts, most frequently for catch, economic impacts, and perceived socioeconomic benefits.","Socioeconomic variables were grouped into five categories for summary analysis: (i) catch (e.g. CPUE, maximum catch size); (ii) economic impacts (e.g. income growth, revenue from tourism); (iii) resource management decision-making (e.g. participation, inclusion of marginalised groups); (iv) perceptions of ecological change (e.g. perception of coral cover, fish biomass); and (v) perceptions of socioeconomic change (e.g. perceived change in remittance, change to income from fishing) (NOTE: ecological and socioeconomic outcomes were broken up into categories, so we could report on each but there are 9 ecological and 5 socioecomic so that would add a lot of outcomes)",Community managed/no-take reserves were more often associated with positive socioeconomic outcomes. Centrally-managed/no-take reserves and Community managed/periodic closure were also more often associated with positive socioeconomic outcomes but sample sizes were small. ,,,,,,,2,1,1,0
83,50,"Tesfaw, A. T., Pfaff, A., Kroner, R. E. G., Qin, S., Medeiros, R., & Mascia, M. B. (2018). Land-use and land-cover change shape the sustainability and impacts of protected areas. Proceedings of the National Academy of Sciences, 115(9), 2084–2089. https://doi.org/10.1073/pnas.1716462115",Land-use and land-cover change shape the sustainability and impacts of protected areas,https://doi.org/10.1073/pnas.1716462115,2018,"(PAs in Rondonia, Brazil)",Brazil,"Tropical, terrestrial",Not stated,1,0,0,,,"To analyze PADDD risk for patterns consistent with bargaining, we fit linear probability regressions to study key covariates’ influences on the risk of degazettement",NATURE / Deforestation,mixed,-1,"PAs that were ineffective in stemming deforestation were more likely to be degazetted or downsized, whereas the PAs (or portions of PAs) that were effective were more likely to see protections maintained or, in some cases, strengthened. We did not find statistically significant impacts of PA degazettement on deforestation for either ineffective or effective PAs... consistent with our hypothesis about a lack of additional effect of the PADDD itself, given that these particular PAs had already failed to effectively block nearby deforestation pressures.",,"PAs that were ineffective in stemming deforestation were more likely to be degazetted or downsized, whereas the PAs (or portions of PAs) that were effective were more likely to see protections maintained or, in some cases, strengthened. ",,,,,,,,,,,,,,,,,,,1,0,1,
115,51,"Ford, S. A., Jepsen, M. R., Kingston, N., Lewis, E., Brooks, T. M., MacSharry, B., & Mertz, O. (n.d.). Deforestation leakage undermines conservation value of tropical and subtropical forest protected areas. Global Ecology and Biogeography, n/a(n/a). https://doi.org/10.1111/geb.13172",Deforestation leakage undermines conservation value of tropical and subtropical forest protected areas. ,https://doi.org/10.1111/geb.13172,2020,,Global,"Tropical/subtropical, terrestrial",120,1,0,1,"protected area buffer zones and matched, unprotected areas",Matching,"We used Global Forest Change data to assess the average yearly rate of deforestation in protected areas, protected area buffer zones and statistically matched, unprotected control areas. ",1) Deforestation leakage 2) Irreplaceability of species,negative,-2,"1) In 55 cases, deforestation rates were higher in buffer zones than in protected and control areas, suggesting a relatively high prevalence of deforestation leakage stemming from protected areas. In 78.2% of documented leakage cases, reduced deforestation in protected areas was not sufficient to offset the amount of deforestation in 10 km buffer zones to a level that would be expected without protection.  The results suggest that protected areas are generally effective at preventing deforestation within their jurisdiction; however, leakage patterns can undermine conservation success because buffer zones often contain habitat for threatened species. 2) In 90.9% of leakage cases, the irreplaceability of species in the 10 km buffer zone was greater than that of the protected area, implying a negative impact of leakage on threatened species.",remote sensing,,,,,,,,,,,,,,,,,,,,1,0,1,
116,53,"Haruna, A., Pfaff, A., Ende, S. van den, & Joppa, L. (2014). Evolving protected-area impacts in Panama: Impact shifts show that plans require anticipation. Environmental Research Letters, 9(3), 035007. https://doi.org/10.1088/1748-9326/9/3/035007",Evolving protected-area impacts in Panama: Impact shifts show that plans require anticipation.,https://doi.org/10.1088/1748-9326/9/3/035007,2014,,Panama,"Tropica, terrestrial",Not stated (all PAs established before 1992),1,0,1,"protected areas, unprotected areas","simple (naive) comparison, OLS, matching","Simple means (by time period), OLS Regression (by time period), Matching (propensity score matching, covariate matching)",Deforestation,positive,1,"(magnitude of effect varied based on method used): ""controlling for differences in characteristics between protected and unprotected locations reduces estimated impacts by something approaching one half"" (but effects of PAs on avoiding deforestation were still overall positive)  Also, effects varied based on location of PAs and time period: ""for 1992–2000 deforestation, PAs farther from urban areas had higher impact. However, for the 2000–2008 land-cover change, we no longer see this spatial impacts difference""",remote sensing,,,,,,,,,,,,,,,,,,,,1,1,0,
117,54,"Blackman, A., Pfaff, A., & Robalino, J. (2015). Paper park performance: Mexico’s natural protected areas in the 1990s. Global Environmental Change, 31, 50–61. https://doi.org/10.1016/j.gloenvcha.2014.12.004",Paper park performance: Mexico’s natural protected areas in the 1990s,https://doi.org/10.1016/j.gloenvcha.2014.12.004,2015,,Mexico,"Tropical, terrestrial",56,1,0,1,"protected areas, unprotected areas",Matching,We use high-resolution satellite data to measure deforestation and (covariate and propensity score) matching to control for NPAs’ nonrandom siting and for spillovers.,Deforestation,mixed,-1,"At the national-level, the average PA had no discernable effect inside or outside its borders. However, regional subgroups of PAs had heterogeneous effects. Also, large, new, mixed use, well-funded PAs cut more deforestation inside their borders.
NPAs were effective in stemming deforestation in one of the nine CONANP regions: Region 6, which accounts for 36 percent of all NPA plots in our national sample.
The only other region in which our preferred covariate estimators generate significant ATTs is Region 3, which accounts for 27 percent of all NPA plots in our national sample. Here, the covariate matching estimators, which are significant at the 1 and 10 percent levels, suggest NPAs increased deforestation by 131–156 percent. We note, however, that neither of the two propensity score matching estimators for Region 3 is statistically significant, which suggests these results are not as robust as those for Region 6.
The anomalous effect of the NPAs in Region 3 begs the question of whether these protected areas drive our finding that at the national-level, the average NPA does not have a statistically significant effect on deforestation. To address that question, we drop Region 3 from the national sample and re-estimate ATTs (Table 2, last column). For this subsample, the matching analysis suggests that on average, pre-1993 NPAs cut deforestation by 43–51 percent.

Spillovers: At the national-level, we are not able to reject the null hypothesis of a zero average effect of pre-1993 NPAs on clearing within 20 km of their borders. Spillovers in Region 6 were positive; that is, NPAs had the same type of effect on deforestation both inside and outside their borders. They reduced deforestation inside their borders by 56–72 percent, and reduced it by 23–29 percent outside. In Region 9, by contrast, spillovers were, loosely speaking, negative. The implications is that if NPAs had any effect on deforestation inside their borders, it was to reduce it. The matching estimators clearly indicate that NPAs increased deforestation outside their borders, however. ",remote sensing,,,,,,,,,,,,,,,,,,,,1,0,1,
118,55,"Pfaff, A., Santiago-Ávila, F., & Joppa, L. (2017). Evolving Protected-Area Impacts in Mexico: Political Shifts as Suggested by Impact Evaluations. Forests, 8(1), 17. https://doi.org/10.3390/f8010017",Evolving Protected-Area Impacts in Mexico: Political Shifts as Suggested by Impact Evaluations,https://doi.org/10.3390/f8010017,2017,,Mexico,"Tropical, terrestrial",Not stated (PAs created by 2000),1,0,1,"protected areas, unprotected areas",Matching,"We rigorously analyze the impacts of Mexican PAs on 2000–2005 loss of natural land cover, using matching to reduce location bias caused by typical land-use economics and politics. ",Deforestation,positive,1,"We find a 3.2% lower loss, on average, due to PAs. Since politics often vary by type of PA, we also show that in Mexico stricter PAs are closer to cities and have greater impact than mixed-use PAs.",remote sensing,,,,,,,,,,,,,,,,,,,,1,1,0,