Wednesday, October 31, 2012

Contingent Valuation Methodology - a means to the capture and analyze willingness to pay

If people merely ‘want’ something, it may not be backed with a willingness to pay for that service. Hence one needs to ascertain people’s maximum willingness to pay for service options under consideration. Willingness-To-Pay (WTP) is the maximum amount an individual is ready to pay for a particular goods/service. Consumer surveys are carried out to estimate the WTP for goods or service under consideration.

The WTP can be estimated using three different ways:
  1. By observing prices that people pay for similar goods in various other markets
  2. By observing individual expenditures on money, time and labor, etc. to obtain goods, OR to avoid their loss. This method might involve an assessment of coping strategies and involve observations, focus group discussions and even house-hold surveys.
  3. By directly asking people what they are willing to pay for goods and services in the future
The first two approaches are based on observations of behavior and are called Revealed Preference Techniques. The third technique is based upon Stated Preferences and includes Contingent Valuation Method.

Contingent Valuation Methodology (CVM) creates a hypothetical market scenario and tries to obtain the value for particular goods, contingent to the scenario. The economic concept that the CV Surveys are trying to capture is the maximum amount that the individual would be willing to pay for certain goods/services.

The most important part of the CV Survey is to simulate a realistic contingent valuation scenario, which has accurately priced options that reflect the levels of prices the goods/service provider will have to charge.

The advantages and disadvantages of the CVM are presented below:


CVM captures a fuller range of benefits of service improvements by investigating people’s maximum willingness to pay for different levels of service that are currently not available
The cost of CVM analysis for a smaller project in terms of time and money are significant. However, incremental costs are relatively modest
Consumers can bid on a range of different service options, thereby defining project designs and technology choices
The results of CVM are often not transferable between locations. This creates a special focus requirement on sample planning
The CVM generates information on household ability and willingness to pay for on-going services, thereby guiding tariff and cost-recovery policy
Sample size needs to be substantial to avoid problems involved with aggregation of responses
If stated clearly, the results of CVM Survey are conceptually easy for non-specialists
Individual biases can cause misleading results

The CVM has proven to be the most popular of available methods for monetary valuation of environment.[1]

The application of Willingness-to-Pay Survey for a utility roll-out is depicted in the following figure. 

How Willingness-to-Pay Study fits into the operational roll-out for utilities?
The chart shows how the study integrates with the other functions of the utility and interacts with them to create an eco-system based on optimum approximations.

When a respondent does not answer a survey question truthfully, it is said to introduce a bias into the survey that undermines the validity of the survey. Therefore, the basic design consideration of the CV Survey is to avoid the many biases that might occur.

The different types of biases and the ways to avoid the biases in your CVM study are shown in the table below. These biases also need to be controlled during the interview conducted for the CVM Study.

Main CVM Biases and Errors
How to avoid
Low Strategic Bid
Respondent lowers their bid assuming that the state, or others, will pay more
Emphasize on policy of state that if the community is not willing to pay sufficiently, the project might not take off
High Strategic Bid
Respondent raises their bid above the real WTP to ensure that the project goes ahead
Make it clear that there will be no subsidies, this is the real amount. Choose correct bidding model
Hypothetical Bias
Respondent does not understand or believe in the options
Explain options clearly.
Poor Sampling
Non-random sample selected which might result in poor quality of data collected.
Ensure accurate mapping of survey area and an appropriate random sampling methodology
Starting Point Bias
Starting price for bidding games influences the final WTP
Vary the starting prices within the sampling frame
Interview and Compliance Bias
Enumerator influenced biases
Analyze responses by enumerator and discard biased responses
Payment Method Bias
Payment method does somehow affect the responses
This might be a realistic bias revealing preferences to certain payment method. This can be safely ignored.

Best practices for Designing WTP Survey
1.     ·  Survey Design should involve easily understood and pre-tested language, taking feedback from all focus group ·   Data Collection should be planned with appropriate attention to sample size, collection methods, sample representation of general population, and randomized selection ·     Correct statistical tests need to be applied for accurate interpretation of the results ·  Regression results for CVM bids should be conducted for validating the data ·  Question Design and interview process should be developed to reduce the bias introduced in the survey

For a case study on WTP for Water, visit our blog here.

[1] “The Contingent Valuation Method: Retrospect and Prospect” (2008), Clive L. Spash

Tuesday, October 30, 2012

DF Attractiveness Matrix: Revised 4 RFPs for Bihar Distribution Franchisees

Recently, BSEB has released revised RFPs for Power Distribution Franchisee for 4 districts in Bihar - PESU, Muzaffarpur, Gaya and Bhagalpur.

pManifold's DF Attractiveness Matrix, provides a quick comparative study on key decision parameters for Bihar 4 DFs with reference to other states DFs like for Nagpur, Agra, Gwalior and Jamshedpur. (click on the image to see enlarged view)

Some of the key excerpts are highlighted below:
  • The current DF scope is at District Level for all the areas. Adjoining areas (approx within 15-20 km) are also in scope of DF.   
  • Geographical Area (Sq.Km.) is not mentioned in the RFP. However, a rough map (not to scale) is provided in RFP for indicative purpose.
  • Patna city and its adjoining areas comes under PESU, which has the highest consumer base compared to other areas/regions. Others are in the range of consumer base of 1.15 lakhs.
  • Connected Load (in KWs) is highest in PESU area, followed by Agra, Jamshedpur and Nagpur in that order. Connected Load for Muzaffarpur and Bhagalpur is provided only at urban (i.e. town) level and not at district level.
  • Electricity Sales (LUs) is highest in PESU area, followed by Jamshedpur and Agra in that order
  • Collection Efficiency is second highest for PESU area, after Nagpur, while it is comparatively low for other towns (i.e Gaya, Muzaffarpur and Bhagalpur in that order) creating good opportunity there for improving commercial losses from effective DF operations.
  • Distribution losses is highest for Gaya, followed by Bhagalpur, Gwalior in that order. PESU area has least Distribution losses in Bihar regions.
  • AT&C losses is highest for Gaya, followed by Gwalior. In Bihar, AT&C losses from high to low rank from Gaya, Bhagalpur, PESU and Muzaffarpur. 
  • Average Billing Rate (ABR) is among the highest for all the areas in Bihar (Avg. Rs. 5.58/kWh), which is highest when compared to others like Gwalior, Nagpur, etc. (This raises some concerns on the reliability of the shared ABR rates)
  • BSEB has specified minimum Benchmark Input Price's for bidders. The minimum benchmark price is for Gaya, followed by Bhagalpur, and highest for PESU area. 
Clearly, across all regions, PESU area will likely attract more bidders because of its volume. The other 3 regions will attract small and new players to develop their base in fast emerging DF landscape in India. The baseline shared in RFP is still not very strong, and will attract many questions during the pre-bid meeting, that is planned for 16th Nov, 2012. Emerging Bihar calls again and this time for success of its first DF model, which will hopefully see increased participation.

Reference: BSEB RFP

Posted by: Kunjan Bagdia @ pManifold

Thursday, October 25, 2012

Part 2 of 2: Interview at Reuters on future of Power Distribution Franchisee model in India

This is sequel to "Part 1 of 2: Interview at Reuters on future of Power Distribution Franchisee model in India". 

pManifold: Some suggestions on the same are as follows:

Needed Intervention
Inadequate and mis-represented
Baseline information for bidding
·      Irrational bidding
·      Delayed bid decision, because of revisions, and litigation's
·      Financial and Non performance risk from DF operator
·      DISCOM to invest in right Technical and Commercial due-diligence for forming the RFP baseline and have it audited by an independent agency
·      DISCOM taking responsibility of  wrong baseline
Poor stakeholder engagement during the bid process
·      Poor final bid participation
·      Risk of re-tendering to mitigate poorer competition
·      Non-optimal DF terms and conditions, leading to later contractual conflicts, and non performance
·      Engage State Govt., DISCOM, Regulator, Bidders and Utility employees well into the DF conceptualization
·      Increase transparency of processes and decisions
Constrictive, open       
ended and non-optimal contract design
·      Minimum benchmark bid prices, disallow financing creativity from Bidders
·      Constrictive elements like improper Escrow design, unclear asset ownership etc. creates difficulty for financing
·      Constrictive qualification criterion (like asking for end-to-end distribution experience, not allowing consortium bidding) brings poorer participation and hence poorer bid price discovery
·      No clearer SLAs led to poorer performance monitoring 
·      There is need for clearer Exit options, to make the model attractive for private Developers and PE investors
·      Better design of Escrow mechanism to be favourable for bank financing
·      Strong SLAs commitment from both DISCOM & DF:
o  Discom: Committed power supply and quality
o  DF: Meeting AT&C loss reduction targets; making power purchase payments regularly; meeting customer satisfaction and other SLAs.       

Week Governance of the Bid process & final bid evaluation     
·      Multiple revisions of RFP and DFA
·      Delayed bid closing
·      Poorer bid participation
·      Multiple extensions
·      Litigations
·      Delayed Start
·      Increasing transparency of bid process and evaluation
·      Discom taking responsibility of wrong baseline, and delayed decision making
·      Invest in proper stakeholder         engagement and online bid room/portal                                
Poorer access to finance to Operationalizing DF (both high Working Capital requirements and Capex for first 3 years)
·      Delayed start of the project
·      Hiccups and non-performance in first year, which further exaggerates opposition against DF model
·      Making Bankers and broader Finance community understand DF model better, and distinguish it from debt burden utility
·      Forming DF initial viability funding from nodal agencies like PFC, REC etc.
·      Allowing right consortium partnership with competent partners on Technical, Operational, Management and Financial side.
·      Improving constrictive clauses in contract design, to allow DF to procure easy bank financing
Weak SLAs to     Monitor Performance of DF     
·      Increase opposition to DF model, without any quantified performance assessment
·      Have clearer milestone, with right incentive/penalty mechanism to encourage performance
·      Have independent customer satisfaction assessment, to establish true metrics for end quality impact from DF model
·      Have transparent reporting mandatory for DF to Regulators
Weak Regulatory      purview of DF model
·       Weak Performance monitoring            
·      DF to stand alone report performance from baseline to ERC (together with Utility)
·      If DF is able to meet its target AT&C reductions, then its customers should benefit with tangible returns (like              either reduced tariff rates as compared to state level Tariff, or reduced load shedding, etc.)  
Weak integration     between different reform schemes (like R-APDRP,    
National Electricity Fund, DF, RGGVY,etc)
·      Poorer end performance i.e. not much AT&C loss reduction, poorer PQR, and dissatisfied customers                           
·      Ensure well integration of R-APDRP and DF projects, to guarantee DF operators with full amount and timely completion of R-APDRP project
·      Ensure interest subsidy from National Electricity Fund
·      No Service tax liability on DF model
·      Allow Electricity duty collected from customers to be used towards Working capital loan

  • At pManifold, we tend to believe the future of DF model is very strong, and infact only way to really bring more accountability and decentralization to the power distribution business, which otherwise is the weakling in the overall power value chain. The model of course needs to be strengthened, and some key improvements are shared above.
  • All our decentralized & alternative Generation efforts will not scale, until we have a strong distribution mechanism, and DF has that potential. So we are looking forward to right regulatory changes to bring integration of these models for better ‘Open Market’ with stronger Private participation:
    • Distributed Generation
    • Distribution Franchisee
    • Open Access (with net metering provisions for smaller generators as well)
    • Energy Efficiency and DSM
  • A much bigger rural market for DF is still awaiting innovation. (See our blog Rural Franchisees - Could they become pilot ground to raise next level of Distribution services?
  • India’s Telecom sector has proved how open market mechanism with supporting Regulations and increased private participation has helped increase mobile penetration, reduce tariff rates, and increase customer services. Same is now looked upon in the Power sector, and I feel we are closer now.
  • The future is bright, if all stakeholders can really work together. CUSTOMERS are the most important link for success of a DF model. The operator that can give more choices to Customers, and meet their PERCEPTION, while manage its Business PERFORMANCE, will succeed. That is the whole reason, that pManifold has developed EUCOPS (Electric Utility Customer Opinion Preferences and Satisfaction) to capture customers voice, and help DF and utility operators track their end performance, as seen by customers. We are glad that more and more DF operators have started using our customer engagement services, and we have worked at Nagpur, Gwalior, Ujjain, Sagar and Dhenkanal, interacting with 7000+ customers from urban and rural. 
  • Customers cooperation is must for DF success, and there is not much in current models to incentivize customer support, as Tariff rates are set at State level. So DF customers will continue to pay higher for inefficiencies in other circles. A tangible and good incentive model within regulatory purview to DF customers (like reduced tariff rates, higher power availability, no reliability charges, reduced Electricity duty or Demand charges etc.), can further expedite DF operationalization. This will also create pressure on local civic bodies to compete and support DF models in their regions, and faster penetration of DF model including in rural areas as envisioned by Planning Commission.


Posted by: Kunjan Bagdia @ pManifold