Mexican Common Property Forestry: Collective Choice Rules a

Mexican Common Property Forestry: Collective Choice Rules and Outcomes March 23, 2009

Figure: Forests in Mexico

Table: Forest Cover by Vertical Integration Type Mean SD N forest ha. (linearized) No sale 3377 351 1994 Stumpage 4820 359 483 Roundwood 6574 592 309 Lumber 16944 2396 126 Total 4543 281 2912 Source: Permit and NFI Data

Focus of talk

Focus of talk Internal organization and decisionmaking variables patterns other than VI

Focus of talk Internal organization and decisionmaking variables patterns other than VI Apply theory to explain internal organization and decisionmaking patterns across a number of decisions integral to community forestry. Examine relationships between internal production organization and economic and environmental and social measures.

Outline Separation of Ownership and Control Constitutional Choice Theory Hypotheses Governance Characteristics Choice Assignments Measured Outcomes Local public goods Reinvestment in forestry Institutional measure Environment Future work

Origin of project

Origin of project Phase 1 Motivation: Lack of data on CP forestry in Mexico Method: Permit data from 10 states in Mexico

Origin of project Phase 1 Motivation: Lack of data on CP forestry in Mexico Method: Permit data from 10 states in Mexico Phase 2 Motivation: Role of CP forestry institutions in market, economic development and environmental protection. Agrarian sector symbolic of sovereignty and power Political, social and economic role of forest tenure Method: Survey data with institutional economics approach

Fundamental questions of survey

Fundamental questions of survey 1. Q1: What explains vertical integration (VI) of communities with commercially valuable forestland in the formal wood products markets?

Fundamental questions of survey 1. Q1: What explains vertical integration (VI) of communities with commercially valuable forestland in the formal wood products markets? 2. Q2: What are the internal community governance (IG) characteristics? 3. Q3: Do VI and IG affect the generation and distribution of ecological, social and economic forest benefits?

Diagram of Analytic Approach

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Internal Production Organization - Separation of Ownership and Control

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Internal Production Organization - Separation of Ownership and Control Organization described by degree of separation of ownership and control (Fama and Jensen 1983) Three main roles in a political-productive organization: Risk bearers (RB) Decision managers (DM) Decision controllers/monitors (DC)

2 main hypotheses (FJ 1983): Separation of Ownership and Control Constitutional Choice Theory Hypotheses

2 main hypotheses (FJ 1983): Separation of Ownership and Control Constitutional Choice Theory Hypotheses H1: Separation in an organization of residual claims and decision management leads to a separation of decision management and decision control.

Separation of Ownership and Control Constitutional Choice Theory Hypotheses 2 main hypotheses (FJ 1983): H1: Separation in an organization of residual claims and decision management leads to a separation of decision management and decision control. Usually applies to more complex operations. H2: Combination of decision management and decision control leads to residual claims restricted to these same persons.

Typology of decisionmaking Separation of Ownership and Control Constitutional Choice Theory Hypotheses

Separation of Ownership and Control Constitutional Choice Theory Hypotheses How apply to Mexican common property forestry? Differentiation of roles among local actors and bodies: RB: community as a whole; society in a broader sense DM: CBC, JM DC: General Assembly, Jefe de Vigilancia, government agencies Control mechanisms: things that provide oversight, credible commitment, enforcement, coordination and trust among actors.

Control mechanisms Separation of Ownership and Control Constitutional Choice Theory Hypotheses Explore both conventional and nonconventional mechanisms for transparency and accountability:

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Control mechanisms Explore both conventional and nonconventional mechanisms for transparency and accountability: democratic participation: meetings, attendance reporting and monitoring practices ability to sanction of rule infractions exit opportunities: alternative income sources peer group sanctioning social capital : trust and networks third party engagement for information and oversight

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Decisionmaking - Constitutional Choice Theory Look at how decisions made and in what context. what are important decisions made in the community who is making them? who authorized that entity to make them? what are mechanisms, internal/external social or otherwise, to hold decisionmakers accountable?

Choice Rules and Levels of Analysis Levels of decisionmaking (Ostrom 1990): Separation of Ownership and Control Constitutional Choice Theory Hypotheses Constitutional level: rules by which rules are made set in our sample; change in 1992 Agrarian Reform opened scope for local organization Collective level: Policy-making, adjudication and management GA decides (by majority, consensus, voting cum bargaining) internal organization GA also retains other responsibilities assigned to it by constitution Operational level: provision, approriation, monitoring, enforcement often by CBC, JV or JM can still be the GA

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Selection of Collective Choice Rules (Zusman 1992) Rules (r i ) for making decisions:

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Selection of Collective Choice Rules (Zusman 1992) Rules (r i ) for making decisions: Do you decide by majority, consensus, sole decisionmaker? Expected utility of decision under that rule, e.g. B(r i ) Transaction costs of decisionmaking using that rule, C(r i ) Risk of not being part of group which decides (i.e. winning coalition) using that rule = spread(r i ) Maybe B(consensus) > B(soledecisionmaker) but C(consensus) > C(soledecisionmaker). Spread increases when size of winning coalition decreases.

Separation of Ownership and Control Constitutional Choice Theory Hypotheses What are the objectives? Performance measures?

Separation of Ownership and Control Constitutional Choice Theory Hypotheses What are the objectives? Performance measures? Table: Forestry Benefits Yes Rival? Yes Private: Dividends, wages, income smoothing Excludable? No Common: timber and nontimber goods. No Club: Investment in forestry operations, Investments in local public goods Public goods: Environmental conservation practices, income/employment smoothing

Diagram of Analytic Approach Separation of Ownership and Control Constitutional Choice Theory Hypotheses

Some Hypotheses Outline Separation of Ownership and Control Constitutional Choice Theory Hypotheses

Separation of Ownership and Control Constitutional Choice Theory Hypotheses Some Hypotheses (FJ) More delegation of authority as the number of productive activities increase (FJ) As production activities increase and diversify, agents separate from the decision managers to monitor the managers. (FJ) Internal governance and not vertical integration distinguishes decisionmaking processes. (FJ) Institutional innovations a response to need for accountability mechanisms. (FJ) Integration of outside expertise into local decisionmaking leads to higher socially preferable performance measures. (ZUS) Distribution of profits is decided at General Assembly level. (ZUS) Long-term investment decreases when decision assigned to smaller coalition. (ZUS) Operational decisions made by CBC or work group leaders or parcel holders.

Sample Statistics Outline Governance Characteristics Choice Assignments Measured Outcomes Table: Survey Sample by State State VI Level Durango Michoacan Total Total Sample Total Sample Sample % No sale 136 4 143 3 7 Stumpage 140 10 120 6 16 Roundwood 68 9 15 3 12 Lumber 42 5 12 1 6 Total 290 28 147 13 41 Source: Survey data

Governance Characteristics Choice Assignments Measured Outcomes Figure: Michoacan sample communities by type

Governance Characteristics Choice Assignments Measured Outcomes Figure: Durango sample communities by type

Governance Characteristics Choice Assignments Measured Outcomes Degree of Delegation Table: Number of Forestry Positions 0-1 2-3 4-5 6-7 None 4 2 0 0 Stumpage 3 12 2 0 Roundwood 1 7 3 1 Lumber 0 1 4 1 Source: Survey data. χ2 sig. at 1%.

Governance Characteristics Choice Assignments Measured Outcomes Degree of Delegation Table: Number of Effective Forestry Positions 0-1 2-3 4-5 6-7 None 4 2 0 0 Stumpage 4 12 1 0 Roundwood 4 5 3 0 Lumber 0 2 3 1 Source: Survey data. χ2 sig. at 10%.

External linkages Outline Governance Characteristics Choice Assignments Measured Outcomes

External linkages Outline Governance Characteristics Choice Assignments Measured Outcomes Represent political, social and technical access

Governance Characteristics Choice Assignments Measured Outcomes External linkages Represent political, social and technical access Explored in our data: NGOs Associations Government program participation

NGOs Outline Governance Characteristics Choice Assignments Measured Outcomes Only 2 NGOs showed up, both in Michoacan No-sale and a stumpage (Monarch Biosphere Reserve), both for capacity training Uneven spread of NGOs and focus on environmental services.

Associational Networks Governance Characteristics Choice Assignments Measured Outcomes PST Government UNORCA Red MOCAF UNOFOC CNC CONOSIL Level 3 Unión UPSE El Salto UMAFOR El Salto? Level 2 UNECOFAEZ, Durango UCEFO El Salto, Durango UCEFO, Oaxaca Ejido ES, Dgo. Community X Community X Comunidad SJN, Michoacán Level 1 Comunidad CN, Dgo. Community Assembly (AG) Governing Board (Comisariado) President, Secretary, Oversight Council (CV) President, Secretary, Treasurer Subcoalition Household Household Household Household Level 0

Governance Characteristics Choice Assignments Measured Outcomes Table: Types of Internal Production Organization Community Work Group Parcels N=26 N=5 N=10 Durango 21 4 3 Michoacan 5 1 7 Formation dates 1962-2001 1992-2000 1929-1998 No sale 4 0 2 Stumpage 11 1 5 Roundwood 6 3 3 Lumber 5 1 0 Past maladministration 8 4 7 Source: Survey data

Profit Distribution Decision Governance Characteristics Choice Assignments Measured Outcomes Table: Choice Assignment: Profit Distribution Production level Who decides* Percent of responses Community General Assembly 62% CBC 33% Work groups General Assembly 60% WG leader 40% WG assembly 40% Parcels CBC 50% Parcel holder 50% General Assembly 25% Source: Survey data.

Price of Trade Outline Governance Characteristics Choice Assignments Measured Outcomes Table: Choice Assignment: Price of Trade Production level Who decides* Percent of responses Community General Assembly 50% Other 27% Buyer 23% CBC 14% Work groups WG leader 40% WG assembly 40% CBC 20% Parcels Parcel holder 86% Buyer 29% CBC 14% General Assembly 14% Source: Survey data.

Governance Characteristics Choice Assignments Measured Outcomes Decisionmaking on Wage/Dividend Advances Table: Choice Assignment: Wage/Reparto Advances Production level Who decides* Percent of responses Community CBC 62% General Assembly 38% Buyer 14% JV 5% Work groups WG leader 75% Buyer 50% CBC 25% Parcels Parcel holder 75% Buyer 25% CBC 25% Source: Survey data

Governance Characteristics Choice Assignments Measured Outcomes Decisionmaking on Volume to Harvest Table: Choice Assignment: Volume Production level Who decides Percent of responses Community PST 96% General Assembly 5% Work groups PST 100% General Assembly 20% WG assembly 20% Parcels PST 100% General Assembly 33% Source: Survey data

Decisionmaking on Choice of Buyer Governance Characteristics Choice Assignments Measured Outcomes Table: Choice Assignment: Select Buyer Production level Who decides* Percent of responses Community General Assembly 82% CBC 14% Work groups WG leader 60% CBC 20% WG assembly 20% PST 20% Parcels Parcel holder 86% CBC 14% Source: Survey data

Governance Characteristics Choice Assignments Measured Outcomes

Governance Characteristics Choice Assignments Measured Outcomes Considerable overlap between community and sub-community level groups.

Governance Characteristics Choice Assignments Measured Outcomes Considerable overlap between community and sub-community level groups. Intricate pattern of checks and balances represented by the 3 roles. Some consistency with own/control and choice theory hypotheses: Decision rule patterns more significant by IO than by VI. Broader decision at higher level. Operational decision (advances) at lower level but mixed. Need to cross check with other mechanisms of social sanctioning and community characteristics.

Governance Characteristics Choice Assignments Measured Outcomes Performance measures Link governance characteristics with performance measures. Preliminary analysis supported by chi-squared tests, correlations and simple regressions.

Local public goods Outline Governance Characteristics Choice Assignments Measured Outcomes Survey: school, potable water, municipal building, church, fiestas, medical, grants, pension, other

Local public goods Outline Governance Characteristics Choice Assignments Measured Outcomes Survey: school, potable water, municipal building, church, fiestas, medical, grants, pension, other Table: Local Public Goods ρ χ2 Prob. VI schools:.42.02 potable water:.23.05 church:.31 > 0.10 range:.21 Internal org. church: -.23 > 0.10 fiestas: -.37.02 medical: <.20.03 range: -.39 Assoc. schools:.40.01 Source: Survey data.

Reinvestment in forestry Governance Characteristics Choice Assignments Measured Outcomes Survey: eqmt, market, forest or soil study, management plan, roads, tractors, trucks, and cranes, secondary processing, diversify, other Table: Forest Reinvestment ρ χ2 Prob. VI pm:.04 roads:.41.06 secondry:.41.04 any:.37.06 Internal org. transport: -.25 > 0.10 any: -.29.06 Assoc. secondry:.26.09 diversify:.29.06 any:.29.06 Source: Survey data.

Cooperation over forestry practices Governance Characteristics Choice Assignments Measured Outcomes

Cooperation over forestry practices Governance Characteristics Choice Assignments Measured Outcomes Rule conformance: T and NT contraband, severity, clearing, fire preparedness. Variations detected: VI: positive correlation. Internal org: ρ = -.50 and negative impact of division supported by preliminary reg where division was sig. ASOC: positive correlation. prelim reg not sig.

Environmental measures Governance Characteristics Choice Assignments Measured Outcomes

Environmental measures Governance Characteristics Choice Assignments Measured Outcomes Species abundance: More, equal, less, disappeared as compared with 5 yrs. ago VI: ρ = 0.44 Internal org: No patterns. ASOC: ρ =.26; χ2 Pr. = 0.06; and prelim reg. forest cover (less than half, little less, equal, more): VI: No pattern; χ2 Pr. = 0.20 Internal org.: No pattern. χ2 Pr. = 0.50 ASOC: Better; χ2 Pr. = 0.006

Outline Future work

Outline Future work Shows variations possible within each VI type.

Future work Shows variations possible within each VI type. Decisionmaking responsibilities follow not VI pattern but internal governance.

Future work Shows variations possible within each VI type. Decisionmaking responsibilities follow not VI pattern but internal governance. The choice to break up forestry activities into sub-community group levels may represent an institutional innovation to improve oversight. Shows array of approaches communities have to create collective productive organizations with all its challenges.

Future work Outline Future work Further develop and link models Further develop empirical analysis: use more details of decisionmaking and governance variations beyond VI, IO and ASOC and individual and community characteristics. Ongoing work: Cross scale linkages with federal/state/regional orgs (with G. Garcia-Lopez) Impacts of internal organizational choices (with L. Fransen) Compare VI in north v. south of Mexico; models of collective choice (with G. Rausser) 10-year panel data study in Oaxaca on institutional change

Reporting Outline Future work Typical mechanisms of control from organizational literature. Most reporting of forestry accounts are written reports More reporting as vertical integration increases. As VI increases, more likely the CBC and external accountant prepare the report. More integrated report more often (2-12 times/year). Not significant across internal organization modes. Internal modes subordinate to collective process in community. Only 3 used external auditors (a WG and parcel community - additional control mechanism?)

General Assembly Meetings Future work Constitutional core of agrarian community No differences in meeting frequencies across VI or internal organization. Most meet every 6 months or every month, as per the constitutional rules. About 1/2 to 3/4 usually attend, again per rules for having a quorum. Most inform people of the meetings by a call (convocatoria) Minutes? Almost always. Frequency of discussing forestry matters? Almost always (partly function of our sample).