Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Staatz MSU CIRAD 2010


Published on

  • Be the first to comment

  • Be the first to like this

Staatz MSU CIRAD 2010

  1. 1. Pathways by Which Improved Information Can Affect Market Performance: Implications for Evaluation of MIS Impacts John Staatz Michigan State University Workshop on Agricultural Market Information Systems in Africa: Renewal & Impact Montpellier, France, 29 31 March 2010 29-31 1
  2. 2. Key messages Complementarity between public and private activities to improve market information MIS can improve market performance both through Direct impacts on decisions of private actors: Spatial and temporal arbitrage Production decisions P d ti d i i Indirect effects via: Complementarity with other hard and soft infrastructure ( f (e.g., roads, credit programs), d d ) Impacts on improving government policies 2
  3. 3. Key messages Yet as one moves towards these more round- about impacts of MIS, it becomes increasingly difficult to diffi lt t quantify th i tif the impact of MIS d t t f due to: Problems of attribution of impacts to the MIS actions vis a vis other complementary and often p y contemporaneous actions (e.g., market reforms) The difficulty of specifying the counterfactual (in terms of the policy environment) for the without- without MIS situation. 3
  4. 4. Why try to estimate impacts? Conceptual links between better information and improved market performance Reasons why private investment may lead to yp y insufficient provision of market information Indivisibility Non-excludabilityy Uncertainty Lack of effective demand among the very poor New ICT may change some of these reasons (e.g., non-excludabiltity), increasing role for private info. provision Q Question remains: At margin, how much to invest g , in public provision of market information relative to other programs? 4
  5. 5. Pathways from better information to better market performance: implications for evaluation 5
  6. 6. Potential Direct Impacts of Better Market Information & Ways of Evaluating Them Rent redistribution in short run– e.g., reduction in intra-market margins following introduction of MIS E.g., early work on Malian MIS Value consumer savings per kg times market volumes Does not examine longer-run growth implications of such redistribution 6
  7. 7. Potential Direct Impacts of Better Market Information & Ways of Evaluating Them Improved spatial arbitrage of existing production more stable and possibly higher average prices Examples: Jensen’s (2007) analysis of introduction of cell phones on J ’ l i f i t d ti f ll h Kerala fishermen’s prices Jenny Aker’s (2009) analysis of impact of cell phones on marketing ma gins Nige ma keting margins in Niger Need good data; this approach captures 1st-round effects only—not longer-term growth or supply y g g pp y response 7
  8. 8. Example: Fish Prices & Mobile Phone Service in Kerala Robert Jensen. The Digital Provide: Information (Technology), Market Performance, and Welfare in the South Indian Fisheries Sector. 8 Quarterly Journal of Economics, CXXIII, 3 (Aug. 2007): 879-924.
  9. 9. Potential Impacts of Reducing Actors’ Price Forecast Error D Aim is to model (using J S social surplus models) N H the costs reductions of being off of the R W I equilibrium price and Z quantity L M E Constraints: How to know reduction in forecast error? K T F D Quasi-static O Estimates of D & S elasticities Kizito (2008): Adopted from Hayami and Peterson, 1972 9
  10. 10. Estimates of with/without / improved information over time Problems of attribution given the complementarity of information with other market improvements Better information as an input into better policies, which then improves payoffs to market information (both public and private) Example of OMA in Mali E l f i M li Problem for estimation of impacts everything is endogenous! 10
  11. 11. So how to measure impacts of more dynamic effects? User satisfaction surveys? Identifying what information comes from MIS vs other sources, including the market itself? Contingent valuation? Problems of stated preference vs revealed vs. preference? 11
  12. 12. So how to measure impacts of more dynamic effects? Revealed preference of funding by public sector or through a users’ organization? May be the most reliable indicator but does this help you indicator, ex-ante? Can you duplicate successful models from elsewhere? Donor funding crowding out national funding? Need an explicit transition strategy to national funding Are we pursuing an impossible objective in trying to precisely quantify the payoffs to improve market information? Quantify what you can, but then build credible stories on more dynamic effects 12
  13. 13. Your suggestions? Thanks for your attention and your help! 13