P8 - Botswana_L Mncube ICN_ACF 11032016

Tools and methods used to
evaluate impact
ICN Agency Effectiveness Workshop
11 March 2016
Gaborone, Botswana
Liberty Mncube (Ph.D)
Chief Economist
Competition Commission of South Africa
[email protected]
Our approach
ACTIVITIES
OUTPUTS
OBJECTIVES
 Evidence of
impact on
competition
IMPROVE POLICY MAKING
Selected cases or sectors
Monitoring
 Inform prioritisation
Retrospective studies
 Learning from
past
performance
 Measure performance
 Suggest how to improve
Impact estimation
SERVE ADVOCACY PURPOSES
 Build support
 Improve awareness
Research
Wider benefits of
competition
 Better
understand
impact
 Deter anti-competitive
behaviour
Impact estimation
• For example, consumer savings
estimates focusing on direct impact
of our work on consumers.
• Estimate impact in the following
areas
– Abuse
– Cartels
– Merger control
• Example
– Cement cartel
– Estimated
overcharges of
between 7.5% & 9.5% & further
assumed the cartel would have
lasted for another 3 years.
– Consumer savings approx. R4.5
- R5.8 billion.
• The total estimates of consumer
savings are conservative in nature.
– General assumptions
– Presenting results in ranges
Monitoring
•
•
Not practical to review all cases in-depth
– Conduct ongoing monitoring of competitive dynamics in certain industries to see
how these evolve post our actions
– Involves monitoring changes in variables such as prices, output, margins and
market shares over time.
Data Required
– May vary on a case-by-case basis.
• Monthly price data per product
• Monthly cost data
• Monthly volumes sold per product
•
Incorporate a reporting requirement into conditions on decisions
•
Example
Telkom settlement has a requirement that the firm will grant the Commission
access to its transfer pricing and retail product accounts and underlying
documents as may be required by the Commission.
Retrospective studies
Method
Some advantages
Some disadvantages
Simplicity
Mostly qualitative information
Tracking evolution of markets
Expensive and time-consuming to get
representative sample
Information
requests and
Surveys
Must be carefully designed to avoid bias
Confidential information a problem
May be difficult to secure cooperation –
incentive for firms to take part
Can estimate a range of parameters Complexity
Can model different modes of
Structural
methods and competition
simulations
Changes in equilibrium outcomes
can be simulated and various
counterfactuals tested
Large amounts of accurate data required
Example,
Using
information
requests,
following the
cement cartel.
Study found
that the breakup of the cartel,
was followed
by new entry,
has generated
investment,
created jobs,
and lower
cement prices.
Research on wider benefits
• Provide links between competition,
competition
policy
and
macroeconomic outcomes such as
growth, productivity, employment,
poverty and inequality
– Research on the effects of
competition on poverty and
inequality
• WorldBank (2016)
– Estimates the potential gains to
the poor from addressing lack of
competition in key food markets.
– The study finds that the
sanctioning of the cartels helped
lower the retail prices of these
goods that form a large share of
the poor’s consumption basket.
– This lifted the disposable
incomes of the poor by 1.6
percent and lowered the
national poverty rate by 0.4
percentage points.