420 likes | 513 Views
Paulo Correa Lead Economist, World Bank Arianna Legovini Head, Development Impact Evaluation Initiative (DIME), World Bank. Smart specialization: opportunities and tools for iterative learning . Objectives of presentation.
E N D
Paulo Correa Lead Economist, World Bank Arianna Legovini Head, Development Impact Evaluation Initiative (DIME), World Bank Smart specialization: opportunities and tools for iterative learning
Objectives of presentation • Provide motivation for the use of iterative/experimental learning in innovation strategies • Develop a framework for identifying bottlenecks and market failures to justify interventions • Highlight how the use of impact evaluations can maximize R&D investment impact through rigorous measurement and learning
Background • Europe 2020: future of EC Cohesion Policy 2014-2020 dedicated to objectives of smart, sustainable and inclusive growth • Thematic concentration on research and innovation to maximize the impact of investment of structural funds • Research and innovation strategies for smart specialization (RIS3) to ensure effective implementation
Information and implications Enabling specialization Discovery through experimentation HIGH MED LOW Unleash latent comparative advantage Risk level for policy makers Modernization NO INFO PARTIAL INFO FULL INFO
Developing your RIS3: Where to specialize? • Where market information is lacking we may not know ex ante where to specialize • Policies targeting sectors/products not desirable due to incomplete and asymmetric information for policymakers • Ex ante decision-making process will likely be biased towards "incumbents"
Developing your RIS3:Finding one’s niche, together • Create enabling environment for efficient market selection • Promote entrepreneurship across the board • Successful companies will constitute the new specialization of the country/region (self-discovery) • Develop a flexible strategy and integrate iterative learning • Focus on measurable intermediate goals • Identify bottlenecks and market failures • Experiment, learn, adapt
Measurable goals • Overall objective • Increase the impact of R&D expenditure on national/regional development • Measurable goals (examples): • Increase private R&D and innovation investment • Accelerate R&D commercialization • Improve technology adoption by SMEs • Research excellence
Identifying bottlenecks • Constraints to increasing private R&D and innovation investment • Credit? • Incentives? • IP? • Constraints to improving technology adoption by SMEs • Awareness? • Skills? • Critical mass? • Constraints to research excellence • Skilled labor? • Partnerships/coordination? • Let’s consider acceleration of R&D in more detail…
Strengthening the innovation chain R&D Commercialization Identify the weakest links in the chain to target, test, and learn from
Why would we assume this? Our tendency is to start with the interventions and make assumptions about how they will help meet our goals But, what is the basis for this assumption? Is the intervention justified?
Who? Start with the objective and identify who the bottleneck lies with
Why? Identify why there is a need for public intervention (the market failure)
Why? Identify why there is a need for public intervention (the market failure)
What? What intervention can overcome this market failure?
And experiment With alternative instruments to learn and adapt
A practical guide to real time impact evaluation and experimentation • Identify bottlenecks to determine possible interventions to be tested • Put down hypotheses for the bottlenecks • But how can we rigorously test which interventions work best?
Why impact evaluation? • Public investment to focus on changes not levels • Targeting areas that would have done just as well even without public support misses the point • Public funds to be used to catalyze growth or improvements that would not have occurred through market forces alone • So, just seeing improvements related to investments is not enough • We need to know what would have happened if the support had not been provided • This can improve public resource allocation
Use the right tools to get the right answers Let’s look at R&D investment before and after the intervention. Before After Treatment group Treatment group According to monitoring, the amount of R&D diminished: the intervention had a negative effect?
Use the right tools to get the right answers Let’s look at R&D investment before and after the intervention. Before After Recession Impact of the intervention Treatment group Control group Treatment group Control group According to evaluation, the intervention had a positive effect: without it, R&D reduction would have been larger!
Use them to learn (compare alternatives) Compare the impact of different mechanisms before and after the intervention. Before After Compare impact Treatment 1 group Control group Treatment 1 group Treatment 2 group Treatment 2 group Control group Treatment 2 has a larger effect
How can impact evaluation help? • We are fairly clear on the WHAT is needed but not on the HOW to provide it • Use what we already know to improve policy design • IE evidence points to behavioral mechanisms that are common across sectors—need to operationalize evidence in the specific context • Identify and learn how to make it work • Careful critical thinking on what you need to know to take better decisions in the future
Use what we know from behavioral finance People … • forget, remind them • text messaging • procrastinate, give them deadlines • announce one time calls for matching grants • shy away from choosing, provide default options • default sign up in training, incentive schemes • say they will but won’t, pre-commit them • precommitment for investments
401(k) Default increase take up [Madrian and Shea, 2001]
Public recognition may be more powerful than financial incentives [Ashrafet al., 2011]
Precommitment increases investment & technology adoption • Farmers offered precommitment savings accounts increase: • Land under cultivation by 10% • Agricultural inputs by 26% • Crop output by 22% Brune, Gine, Golberg, Yang (2011)
And think about what else we need to learn • What information and incentives will facilitate academia-private sector partnerships? • What level of technical assistance is needed to turn start ups into viable businesses? • Is finance more important than technical support? • And test it…
Mimicking venture investor: addressing incomplete information
DIME support develop communities of practice generate local knowledge expandglobal policy adoption Large research teams Train and design evaluation of programs at scale Identify regularities Networks of policy-makers Experiments to test possible solutions Use networks to discuss results Regular IE thematic events Find and adopt solution Stimulate policy-makers to consider adoption
Task • Starting from your goals, identify • Implementation and behavioral issues • Possible instruments and mechanisms • 3-4 ideas about possible things to test in your strategy