Background

Debate over the effects of different property regimes on natural resource systems has long been controversial, incited by Hardin’s [1] thesis that common pool resources will inevitably suffer from overexploitation and degradation. Moreover, the dominant paradigm long held that government or private property was required for conservation and sustainable resource use. In response, a large body of scholarship has demonstrated that widening the breadth of property rights held by local-level actors in common property regimes can lead to more efficient and effective outcomes for resource sustainability [2–4]. Devolution of property rights to community and local level actors has since been used as an instrument for achieving goals as disparate as poverty alleviation [5], gender equity [6], resource conservation [7], and climate change mitigation [8]. Of course, states have also retained or claimed new property rights or allocated them to private sector actors in the name of these same goals [9].

A large, diverse, and rapidly growing body of literature has investigated the links between these property regime transitions and their environmental outcomes. A significant portion of the literature assesses recent decentralization policies, broadly described as tenure reforms, that transfer decision-making rights and authority from central to local governments or formally recognized existing de facto rights at the local level. Another branch of literature assesses the outcomes of initiatives for community-based natural resource management and community-based conservation. Thus far, the literature has yielded mixed findings on resource conditions and sustainability such as biodiversity loss, forest cover change [10], fisheries decline [11], and rangeland degradation [12]. Despite the expanding literature, little has been done to account for the variation in environmental impacts which limits advances in policy making and management intervention. Moreover existing syntheses and reviews on property regimes focus mostly on community based management [13–15], although natural resources are governed through state, private and common property regimes across diverse ecological and political systems. Broadening the scope to examine outcomes in state and private property regimes can give us valuable theoretical and policy insights on similarities and systematic differences within and across resource systems. The increasing emphasis on landscape approaches and thinking beyond individual resource systems makes informing policy and practice at multiple scales of governance even more crucial [16].

This systematic review will synthesize extant empirical evidence of the impacts of different property rights regimes on environmental outcomes in three resource systems at local to regional scales in developing countries: forests, fisheries and rangelands. Although this review will limit itself to the assessment of environmental outcomes, it will also consider context and mediating factors and will aim to determine more systematically which contextual elements matter most decisively. Accounting for the context is especially important since property regime transitions are not always unidirectional nor fully realized, leaving ample room for discrepancies between existing de facto and newly inscribed de jure regimes and conflicts between recognized and unrecognized actors [17].

Objectives of the review

The review seeks to answer the following primary question:

  • What are the environmental impacts of different property regimes in forests, fisheries, and rangelands in developing countries?

It also poses two secondary questions:

  • Which property regimes are associated with positive, negative or neutral environmental outcomes?

  • How do those environmental outcomes compare within and across resource systems and world regions?

Assessment of the impacts of property rights regimes across a range of different resource systems and ecosystem types enables exploration of commonalities and differences across these systems. As the review is interested in looking both short term and long term results of property rights interventions, both terms impact as well as outcome will feature in the review. Appendix explains in detail the definitions of other key terms used in this review.

This review adopts a PICO (Population-Intervention-Comparator-Outcomes) framework to structure analysis of these research questions [18], summarized in Table 1.

Table 1 Research framework for Population-Intervention-Comparator-Outcomes (PICO)
Full size table

Population

The population refers to the three resource systems: forests, fisheries and rangelands (see Appendix for operational definitions of each of these systems). We have chosen these systems due to the importance of the ecosystem services they provide as well as their broad geographical coverage, which includes much of the range of ecosystem types found across the globe. These three resource systems also represent differing levels of resource mobility and variability, thus introducing important variation in the biophysical nature of the resources they provide. We will exclude other natural systems as well as heavily human-modified systems such as irrigation systems and cities.

Intervention

The intervention refers to the introduction of a particular property rights regime, whether state, private, and community or some combination of these (mixed regimes). The intervention could also be the establishment of a protected area for the explicit objective of resource conservation, although these cases will be treated separately. The review uses a bundle of rights approach, introduced by Schlager and Ostrom [19], to examine how the distribution of access, withdrawal, management, exclusion, and alienation rights in state, private, community, and mixed property regimes affects resource outcomes. In addition, this review considers the right to income from resource use as part of the bundle of rights that comprise a property regime, defined as a system of rules governing access to and control over resources [20], and specifying permissible and forbidden actions in relation to use and management, responsibilities and obligations [2, 21]. Also in the case of protected areas, the bundle of rights approach will be used as it captures well the different joint/mixed property rights arrangements present in protected areas.

Comparator

This review compares environmental outcomes based on analysis of studies using the following three methodologies: change from before to after the intervention (temporal change), case control studies (with-without comparison), or a combination of both (a BACI—before-after-control-impact—design). This latter design is based on data from before and after the intervention of interest and in sites where the intervention took place and matched control sites that are similar as possible to the intervention sites except that there was no intervention. The BACI approach seeks to rule out potentially confounding effects and to increase confidence that outcomes observed were due to the intervention [22].

Reviewers anticipate that control sites will often be characterized by open access regimes, but they may be any of the different property rights regimes identified as long as they were present in both treatment and control sites prior to the treatment (change in property rights regime). Moreover, the socioeconomic and environmental baselines of control and treatment sites should be of sufficient similarity such that divergent environmental results, if any, are attributable to the intervention or named contextual factors. In the case of protected areas, the comparison will need to present temporal comparison (before-after) within the protected area or spatial comparison with another regime outside the protected area.

Outcomes

The outcome of interest in this analysis consists of qualitative and quantitative changes in environmental measures, which may vary by resource system. Table 1 includes illustrative outcome indicators likely to be found in relevant studies. Based on information on change and/or difference in these indicators in each study, reviewers will determine whether the environmental outcomes associated with different property regimes were positive, negative or neutral. Both the original outcome measures reported in the studies under review and the reviewers’ assessment of environmental outcomes will be recorded.

Methods

Searches

To capture as unbiased and comprehensive a set of relevant literature as possible within the constraints of the review budget, time allocation, and familiarity with languages, the search will be conducted in the databases shown in Table 2. Searches conducted on Google and Google Scholar will be limited to the first 200 hits retrieved. Any links will be followed only once from the original hit. Previous systematic reviews on the topic [13, 15, 23–25] and literature reviews assessing tenure and environmental outcomes that are identified by the search will be hand-searched to identify further relevant studies.

Table 2 The list of databases and other data sources for the systematic review
Full size table

Electronic search strategies have been tested using the ISI Web of Knowledge, CAB Abstracts and Google Scholar. This testing process has been documented and informs this protocol (see Additional file 1). The following search terms will be applied to the different databases, with search term and database specific modifications. As the operation of websites and database-specific search engines varies greatly, the reviewers will modify their search and search terms for each database in order to retrieve results that balance the needs for sensitivity and specificity to the review question [18]. If use of several search terms is impossible, the reviewers will take advantage of available topic-relevant key words and publication categorizations. Reviewers will also adjust for different word permutations or suffixes through the use of wildcard symbols, where applicable. The development and implementation of the search strategy will be recorded, including the testing process, number of hits, relevance of the results and the date of search. The reviewers will also contact individual organizations (through librarians or other information specialists) for further guidance.

To reduce language bias that may be associated with limiting the review to English language publications, the search will be also conducted in French and Spanish. Search in French and Spanish will be conducted in Google, Google Scholar as well as in primary databases where applicable. In addition, reviewers will conduct search for grey literature within institutions and NGOs whose main language is French or Spanish, e.g. CIRAD (French) and Facultad Latinoamericana de Ciencias Sociales (Spanish). In order to take benefit from all the languages mastered by the review team an additional search in Indonesian will be conducted where applicable. The databases will be searched with following English terms and with their French, Spanish and Indonesian translations.

Population terms: Forest, fish, marine, grassland, pastoralist, pasture, rangeland

Intervention terms: Collective, common, community, customary, government, public, private, small-scale, state, public, private, company, concession, participatory, collaborative, co-operative, co-manage, shared, joint

Intervention-related terms: Decentralization, tenure, reform, allocation, ownership, property right, property rights, property regime, property system, management, access, harvest, open access

Examples how search words will be combined are presented in Table 3. Searches will also be conducted using different institutional accesses (CIFOR, University of Michigan and University of Oxford) to take advantage of different subscription databases.

Table 3 List of search terms in Google Scholar(1), WOK(1)and CAB(1)
Full size table

Study inclusion and exclusion criteria

Inclusion criteria will be applied to select the relevant articles captured by the search. Inclusion criteria will be applied to the titles and abstracts of articles. Studies will be included if they fulfill the following criteria.

Type of study

Only primary empirical literature will be included, such as case studies, case–control studies and cohort studies, including quantitative and qualitative research.

Subjects studied: We will include studies that asses any property regime associated with forests, fisheries and rangelands. Studies need to assess outcomes from before/after change in a property regime (temporal dimension), assess outcomes from different regimes in a with/without setting (spatial dimension), or a combination of these (BACI). Papers reviewing environmental outcomes without a reference to a specific property regime will not be included, nor will studies of plantation forests and aquaculture. The review will exclude commentary and position papers.

Outcomes

Studies must measure and/or qualitatively assess change and/or difference in environmental outcomes as illustrated Table 1.

Regional focus

Studies will only be included if the research focuses on the developing countries of Latin America and the Caribbean, Africa, and Asia and the Pacific (see Appendix for complete list of countries). Developing countries are those defined as either low or middle income according to the World Bank. We will use this classification rather than others (such as OECD/non-OECD), as the division into low and middle income countries enables synthesis that includes economic context.

Timeframe

Studies need to have been published between 1990-present.

Based on the inclusion criteria described above, the review will use a three-step process to identify studies for inclusion.

  1. (1)

    Studies clearly not relevant will be excluded on the basis of titles only

  2. (2)

    Studies with potentially relevant titles will be assessed using their abstracts.

  3. (3)

    Any potentially relevant study that gets through stages 1 and 2 will be collected and assessed for inclusion as full text.

A kappa statistic – the standard measurement used in previous systematic reviews [15],[13] to check for consistency - will be calculated for all reviewers involved in screening prior to work on inclusion. This has been trialed during protocol development and moderate to high levels of consistency have been achieved during these trials. Once satisfactory kappa statistics has been achieved for our final set of screeners, reviewers will determine separately which papers fulfill the inclusion criteria at each step for batches of publications.

To check for consistency of the selection at each stage, two authors will review a 10% random subsample of abstracts and full texts at these screening stages. If too many differences between inclusion and exclusion are perceived, further discussion on interpretations and possible revision of the criteria will be done iteratively until a satisfactory kappa figure is achieved, following best practice with other published systematic reviews. We will also record the reasons for exclusion at full text for each article and provide this information as an appendix in the final systematic review.

Potential effect modifiers and reasons for heterogeneity

The following potential effect modifiers related to the environmental, socio-economic, and political context of the intervention will be considered and recorded:

Environmental context

  • Location

  • Ecosystem type

  • Spatial extent of resource area

  • Elevation

  • Accessibility

  • Baseline resource condition

  • Existence of external environmental management intervention

Socio-economic context

  • Population density in study/resource area

  • Change in population in study/resource area

  • Local and external market demand on resource

  • Economic inequality (information stated in study, GINI coefficient, etc.)

  • Presence of education initiatives

  • Presence of infrastructure

Political context

  • Nature of political regime (democracy, authoritarian, totalitarian)

  • Decentralization (whether decentralized or decentralizing; year decentralization process began; extent of decentralization: advanced; not advanced)

  • Corruption (no corruption, low corruption, high corruption according to study; other measures of corruption, e.g. WGI, Transparency International)

The following additional characteristics of property regime interventions will also be noted where information is available in the study:

  • Clarity of rights

  • Stability of rights

  • Level of enforcement

  • Legitimacy of decision-making authority over rights

  • Gender equality of property rights

  • External support: whether the regime is supported by external actors, such as NGOs, donors, or companies

  • Formal protected area: Whether property rights regime applies to a legally protected area

The variables listed above were identified based on consultation with experts in the field of property rights and natural resource governance and knowledge among the review team of the empirical and theoretical literature relevant to this review. The reviewers have winnowed the number of potentially relevant variables to a manageable set that addresses especially salient hypotheses in this area of inquiry.

Study quality assessment

Once all relevant articles have been identified, full texts will be reviewed to assess study quality according to the questions below. These questions are based on recommendations by the Cochrane Collaboration [26] as well as previous reviews [13], but have been modified to account for the realities of available research on the impacts of property rights, which is characterized by an extensive number of qualitative case studies [27]. Two researchers will code a 10% random sample of articles to test the coding protocol and intercoder reliability. Kappa values will be calculated to assess agreement and if agreement is less than 50% the researchers will adjust the coding protocol to increase clarity and agreement.

Questions and coding system used to guide the quality assessment

  1. 1.

    Clarity and replicability of methods: Are the research methods clearly presented so that the research could be repeated? [clear and repeatable =1, not clear and repeatable = 0]

  2. 2.

    Appropriateness of methods: Are the research methods appropriate for addressing the research question(s)? [appropriate = 1, not appropriate = 0]

  3. 3.

    Study design category: Which of the following categories is most appropriate to describe this study? [cross sectional study or time series study =0; case control study = 1; controlled before-and-after (BACI) study = 2]

  4. 4.

    Sample size: Is sample size explained and well justified? [yes = 1, no = 0]

  5. 5.

    Confounding factors: Did the study account for and seek to minimize the effects of potential confounding factors in its design and analysis? [yes = 1, no = 0]

Studies will be assessed based on the above quality categories. The quality of each study will be scored based on the above questions, with results recorded in a separate Excel spreadsheet. Explanations for each decision will be recorded in order to keep the process transparent and repeatable.

However, for those identified studies that fail the quality requirements (scoring 0 on our quality assessment scale), a sensitivity analysis will be conducted during the data analysis stage to determine the effect of their inclusion on the results of our systematic review. Should their results be markedly different from those of studies that met quality criteria, they will be discarded from the final synthesis. These studies will be available for analysis if the sensitivity analysis indicates that the review would be richer with their inclusion. We will in any case capture the numbers that have been assessed at different quality levels in graphical representations of the state of the evidence base.

In addition, reviewers will also record the type of the data analysis used in each case study according to the following typology:

  • qualitative analysis

  • quantitative analysis -descriptive or observational statistics

  • quantitative analysis – analysis of variance, t-test, statistical correlation or other bivariate analysis

  • quantitative analysis-multivariate regression or other multivariate analysis

Data extraction strategy

Data on individual property regime interventions and their environmental outcomes will be collected in a data extraction matrix using an Excel spreadsheet. This will include information on the resource systems in which the property rights regime is implemented, the de jure and de facto nature of the regime as determined by the specific property rights accorded under each regime, stated objectives of the property rights regime intervention, intervention year, study year, the environmental outcomes of regime interventions, and confounding factors that may explain the nature and variation of the outcomes of the regime, including baseline environmental characteristics, external regime characteristics that may further enable the outcomes of an intervention (such as the stability of rights held and level of enforcement of the regime), and characteristics of the socio-economic and political context. A coding protocol has been developed and it is presented in Table 4. As the data extraction advances, other coding systems will be developed (e.g. resource systems, countries and geographic regions).

Table 4 Initial coding protocol for data extraction
Full size table

To fully address geographic differences, we will collect not only country data but also data within subregions, and further take into account the varying ecosystem types within the broad categorizations of forests, rangelands, and fisheries by documenting individual ecosystem type. Should a paper present multiple studies of different property regimes, each of these will be recorded individually within the data extraction sheet. Thus if spatial comparisons of two regimes or comparisons of outcomes from multiple regimes are made, each regime will present an individual data entry.

Data synthesis and presentation

Our data synthesis will be based on the information categories mentioned in “Potential effect modifiers and reasons for heterogeneity” as well as “Data extraction strategy categories.” We will synthesize the results on environmental outcomes across different resource systems, ecosystem types and geographical regions. We will synthesize the environmental results considering the allocation of bundle of rights as well as the context factors. A series of matrices will be deployed to illustrate: a) how environmental outcomes may vary according to the bundle of rights allocation as well as the institutional arrangements that support (or not) the rights regimes, for example, security of rights as determined by enforcement, clarity, stability and legitimacy of authority; b) how contextual factors influence environmental outcomes; c) how environmental outcomes vary with resource type; and d) how a to c above vary in different geographical locations. The synthesis matrices will be accompanied by a narrative. We will also include a note on the performance of different methodologies in providing a nuanced understanding of the environmental effects of property regimes. A major outcome of the synthesis will be the identification of aspects that need further, in-depth inquiry as well as policy implications of current findings.

The variety of measured outcomes and possible lack of quantitative data will delimit the applicability of statistical tools. Data will be analyzed using regression and other statistical techniques (as far as possible) to complement qualitative, narrative analysis. The review will also explore whether or not to include sources for which a significant (33%-50%) portion of the data are missing as done in other relevant systematic reviews [13]. Finally, a sensitivity analysis will be conducted to synthesize the conclusions of studies that did not meet quality assessment standards, and consider differences (if any) from the results of those studies that were included in the final review. This will be done quantitatively if the data is so presented, or otherwise will be synthesized through qualitative methods using the same data extraction matrix.

It is well known that in many research areas papers are more likely to be published if they demonstrate clear, positive results (or strong negative effects), and that papers that shown little or no effect are less likely to be published than “negative”. To assess the possibility of such publication bias, we are conducting searches of ‘grey’ literature (much of it not formally published) in addition to studies in academic journals [28] will assess whether there is evidence of publication bias. If data allow, we will assess bias using funnel plots, which show effect sizes and standard error or sample sizes [29].

Appendix

Glossary of key definitions and terms used in the review protocol

Forest: Land with tree crown cover (or equivalent stocking level) of more than 10 percent and area of more than 0.5 hectares (ha). The trees should be able to reach a minimum height of 5 meters (m) at maturity. A forest may consist either of closed forest formations where trees of various storeys and undergrowth cover a high proportion of the ground or open forest formations with a continuous vegetation cover in which tree crown cover exceeds 10 percent [30].

Wooded lands (woodland): Land either with a crown cover (or equivalent stocking level) of 5–10 percent of trees able to reach a height of 5 m at maturity, or a crown cover (or equivalent stocking level) of more than 10 percent of trees not able to reach a height of 5 m at maturity in situ (e.g. dwarf or stunted trees), or with shrub or bush cover of more than 10 percent. Wooded lands are included in this definition of forests [30].

Fishery: A geographical place, activity, or unit that is involved in raising and/or harvesting fish. As a unit, a fishery is typically defined in terms of some or all of the following: people involved, species or type of fish, area of water or seabed, method of fishing, class of boats and purpose of the activities [31].

Rangeland: Land on which the indigenous vegetation is predominantly grasses, grass-like plants, forbs, or shrubs and is managed as a natural ecosystem. If plants are introduced, they are managed similarly. Rangelands included natural grasslands, savannas, shrublands, many deserts, tundras, alpine communities, marshes and meadows” [32].

Property rights: A property right is an enforceable claim to use, control or otherwise benefit from a resource [33, 34]. Property rights is often made up of a bundle of multiple rights (and responsibilities) including [19]:

  • Access is the right to enter a defined physical property

  • Withdrawal is the right to enter a defined physical area and obtain resource units or products of a resource system (e.g., cutting firewood or timber, harvesting mushrooms, diverting water)

  • Management is the right to regulate internal use patterns and transform the resource by making improvements (e.g., planting seedlings and thinning trees)

  • Exclusion is the right to determine who will have right of withdrawal and how that right may be transferred

  • Alienation is the right to sell or lease withdrawal, management, and exclusion rights.

The bundle of rights also include the right to earn income from a resource even without using it directly and is derived from permitting others to use the resource [35, 36].

Property rights regime: a system of rules governing access to and control over resources [20]. Rules specify permissible and forbidden actions in relation to use and management, responsibilities and obligations [2, 21]. The holder of a property right can be an individual, a corporation, a group or the state/government:

  1. a)

    Private property: Individual or “legal individual” holds rights.

  2. b)

    Common property: group members hold rights (e.g. community)

  3. c)

    Public property: state holds the rights

  4. d)

    No-property or Open access: no one has rights and everyone can use the resource as they like; no effective management or regulation

List of developing countries included in the analysis

Income groups correspond to 2012 gross national income (GNI) per capita (World Bank Atlas method) [37]. We will take account of country name changes since 1990 in searching for studies from the relevant countries [38].

Low-income economies ($1,035 or less)

Afghanistan, Bangladesh, Benin, Burkina Faso, Burundi, Cambodia, Central African Republic, Chad, Comoros, Congo, Dem. Rep, Eritrea, Ethiopia, Gambia, Guinea, Guinea-Bisau, Haiti, Kenya, Kyrgyz, Liberia, Madagascar, Malawi, Mali, Mozambique, Myanmar, Nepal, Niger, Rwanda, Sierra Leone, Somalia, South Sudan, Tajikistan, Tanzania, Togo, Uganda, Zimbabwe.

Lower-middle-income economies ($1,036 to $4,085)

Armenia, Bhutan, Bolivia, Cameroon, Cape Verde, Côte d'Ivoire, Djibouti, Egypt, El Salvador, Georgia, Ghana, Guatemala, Guyana, Honduras, Indonesia, India, Kiribati, Laos, Lesotho, Mauritania, Micronesia, Mongolia, Morocco, Nicaragua, Nigeria, Pakistan, Papua, Paraguay, Philippines, Samoa, Sao Tome, Senegal, Solomon Islands, Sri Lanka, Sudan, Swaziland, Syrian Arab Republic, Timor-Leste, Uzbekistan, Vanuatu, Vietnam, West Bank and Gaza, Yemen, Zambia.

Upper-middle-income economies ($4,086 to $12,615)

Angola, Algeria, Samoa, Argentina, Azerbaijan, Belize, Botswana, Brazil, China, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, Fiji, Gabon, Grenada, Iran, Iraq, Jamaica, Jordan, Kazakhstan, Lebanon, Libya, Malaysia, Maldives, Marshall Islands, Mauritius, Mexico, Namibia, Palau, Panama, Peru, Seychelles, South Africa, St. Lucia, St. Vincent and the Grenadines, Suriname, Thailand, Tonga, Tunisia, Turkey, Turkmenistan, Tuvalu, Venezuela.

Documentation of search term testing conducted in CAB, Google Scholar and WOK

Testing process with search terms for CAB, Google Scholar and WOK. The search was conducted in the online CAB, Google Scholar and WOK databases by the authors WZ and MO, see table for detailed search terms, results, dates and comments on results.