VALUE CHAINS FOR JOBS ESTIMATION Lets Work Pillar
VALUE CHAINS FOR JOBS ESTIMATION Let’s Work - Pillar 2: Measuring the Impact of Private sector Investments on Jobs London, 16 -17 September, 2015
Where are the jobs in value chains? Source: http: //www. qstockinventory. com/wp-content/uploads/2013/09/Value-Chain-Word-Cloud. jpg
Outline 1. Context and introduction to the tool 2. Sampling 3. The survey 4. Analyzing results for jobs estimation 5. Key questions and next steps
CONTEXT AND INTRODUCTION TO THE TOOL 3
Why value chains for development? Source: https: //nextgenafricanfarmers. files. wordpress. com/2013/06/agri-value-chain. gif § Job creation leverage: VCs offer link between competitiveness and jobs competitive lead firms can drive large-scale job creation and productivity growth through the chain § Specificity: The critical interventions to unlock competitiveness are often industry (VC) specific § Mobilization potential: VCs offer a platform to build coalitions of mutual interest, increasing potential for
Why value chains for jobs estimation? – example Value chains help us understand better the scale, location, and nature of jobs Example: Ethiopia leather footwear value chain Exports ~ 10, 000 formal jobs (v 600, 000+ in Vietnam, 80% female) 6 firms Local market FDI 9 firms Local formal Imports 2 nd hand 1000+ firms Informal Footwear manufacturers 9 firms FDI Imports Local formal 15 firms Leather processing manufacturers Hide and skin collectors Cattle farmers Imports 1, 500 collectors ~ 15 million Source: Brautigam, Tang, and Mc. Millan (2011); Ethiopia Central Statistics Agency 1. 6 m commercial (just 0. 7 m dedicated livestock) 13. 2 m subsistence (2. 2 m dedicated livestock)
Why value chains for jobs estimation? – the tool The proposed tool: value chain surveys, complemented (where needed) by semistructured research approaches (interviews, focus groups) § ‘Real-time’ § Highly sector (VC) specific § Documents extent and nature of relationships among actors in the chain § Captures nuance around firm decisions to employ capital and labor and to source inputs § Exposes differences across segments of firms in the chain § Highlights factors that open / shut off opportunities for
Not just about estimating jobs #s Job creation effects - MORE - Job quality effects (productivity, earnings, skills) - BETTER - Jobs access effects - INCLUSIVE -
The value chain tool: 3 inter-related objectives Pillar 1 Pillar 2 VALUE CHAIN DEVELOPMENT JOBS ESTIMATION JOBS FOR VALUE ESTIMATIO CHAINN FOR FIRM -SPECIFIC WIDE INVESTMENT TS
Combining approaches Pillar 2 (Measurement) VALUE CHAIN JOBS ESTIMATION Pillar 1 (Sector deep dives) COMPREHENSI VE VALUE CHAIN DEVELOPMENT • Relatively narrow exercise to estimate job creation effects • Dependent on underlying assumptions on how the value chain is / will be structured and dynamics of its operation • Looks at potential for value chain expansion • Looks at potential for increasing links and inclusivity in value chain • Aims to identify constraints and interventions to address growth and inclusivity PILLAR 2 COUNTRIES MAY INCLUDE ESTIMATION + MORE COMPREHENSIVE DEVELOPMENT APPROACH
FIELD RESEARCH 3 -4 weeks 6 -16 weeks Kick-off mission • Database analysis • Generic VC (global) mapping • Review existing sector info • Initial sector screening Initial counterpart and stakeholders consultation VC selection / mobilization Identification of potential firms to survey ANALYSIS / CONCLUSIONS 3 -4 weeks Field research missions (optional) Concluding mission FIRM SURVEYS (50 -100+ sample, context-dependent) FIRM INTERVIEWS AND FOCUS GROUPS Analysis and recomme ndations (20 -40 sample) IMPLEMENTATION PLANNING SCOPING AND PRELIMINARY ANALYSIS Develop and implement sampling strategy ACTIVITY PHAS E Broad overview of the process and tools TOOL PUBLIC PRIVATE DIALOGUE RFP / TOR template Terms of reference for hiring external experts to lead the VC analysis Screening template Survey instruments Criteria and data sources for screening and selecting value chains General and sector-based survey templates Sampling guidelines Strategies for sampling depending on data and context Interview guidelines Guidelines for semi-structured interviews Completion checklist List of what should be understood by the end of the analytical process
SAMPLING 11
Sampling for value chains – why it is critical… and hard
First challenge – defining and mapping the actors in the chain § Tracing the existing supply chain of an individual investee is relatively straightforward (if not necessarily simple)… but it is enough to understand jobs potential? § Challenge of identifying firms outside the existing chain § Value chain map as a starting point – Structure – Approximate number of market players in each part of the chain – Identification of main companies in the chain and the nature of firms (SME v large, formal v informal, domestic v FDI) § Where to draw the line in the value / supply chain? – Vertically: forward past processing and into distribution / recycling?
Second challenge – finding all the actors in the chain § Data sources: –Previous VC and sector studies (global for the generic structure; national for specific) –National and sectoral statistics –Industry associations (websites, studies) –Consultations with stakeholders Best case scenario: Worst case scenario: + full list of sampling units – no list of + information on sampling units structural – no information relationship on structural e. g. all formal, firm relationship census/register, e. g. all informal supported by and no data on detailed I-O firmsvalue chain participants can involve Techniques to identify current / potential geographical sampling techniques, snowball sampling, and other social network -related approaches
Sample sizes for robustness and budget § Precision: 7. 5% with 90% confidence interval? § Uniformity of underlying population: 50/50 proportion? § Stratification: ensuring the sample is representative for different strata has big implications on sample size – what are the strata that we think will matter? Small v medium/large firms? Or firms inside v outside the existing chain? § Also need to inflate for sample loss and non-response 99 % 95 % 90 %
So how many surveys would we need to do? § Below is an estimate based on the case of the Ethiopian leather value chain discussed earlier (note: most value chains we study will have smaller populations) § It may be practical to under-sample where uniformity is reasonably expected (e. g. cattle farmers, especially informal)
THE SURVEY 17
Key elements and coverage of the survey tool DYNAMIC EFFECTS DIRECT AND INDIRECT STATIC EFFECTS (e. g. productivity gains, tech / knowledge spillovers, changes in formalization, etc) (e. g. investment, growth, etc. ) Survey modules 1. Markets and competitive environment 2. Workforce 3. Production and cost 4. Supply chain ? Sector-specific versions? - Agriculture - Services - Tourism - Manufacturing (default) Firm-type versions? structure - Medium and large versus microenterprise Surveys estimated to require approximately 45 minutes for completion (face-to-face)
Module 1: Firm background, markets, competitive environment • Firm background: ownership, legal status, registration, age • Profile of main business – products, services and share of sales • # of customers; nature of customer relationship • Selling price • Share of sales by main product and market (local v national v export) • Barriers to increasing sales More detailed survey for value chain development projects includes additional questions on: • Competitors • • Access to market information Transport to markets Relationships with key buyers (including technical assistance) Strategic planning
Module 2: Workforce • Profile of current workforce by type/skills: #, gender, share foreign, share casual • Reasons for using casual / part-time labor • Wages • Labor-related barriers to growth (regulation, skills, availability, turnover, etc) • Challenges to finding skills / new hires • Specific occupations and technical skills needed More detailed survey for value chain development projects includes additional questions on: • Non-wage benefits • Non-cognitive skills requirements • Skills gaps • Training practices
Module 3: Production structure and costs Production capacity and output Cost structure – factory gate and delivered cost Relative use of capital and labor Hypothetical reaction to output growth (use of K and L) • Factors influencing labor force expansion • • More detailed survey for value chain development projects includes additional questions on: • Technologies employed • Challenges to improving productivity • Other constraints in operations
Module 4: Supply chain • Top goods and services inputs and sourcing pattern • Hypothetical: response in input levels and sourcing with output growth • # of active local and foreign suppliers for top inputs • Sourcing from informal suppliers • Share of local sourcing – current and planned • Obstacles to sourcing from local firms More detailed survey for value chain development projects includes additional questions on: • Input implications of technology changes • Provision of technical support to local suppliers
More detailed / comprehensive survey covers additional issues on structure and dynamics of value chain • Access to market information • Dynamics of relationships at vertical and horizontal levels in value chain • Existence of cooperative arrangements in the value chain • Formal and informal institutional relationships in the value chain
ANALYZING RESULTS FOR JOBS ESTIMATION 24
Basic approach DEFINING A JOB: For measuring direct employment, we use the definitions agreed by IFI’s in the recent ‘indicator harmonization’ effort § Establishing a relationship between output and jobs (multiplier, or ‘jobs coefficient’) – Investment and output relationship taken as exogenous § Drawing on survey results of: – Relationship between lead firm (or overall value chain) growth and growth of different strata of firms in the chain – e. g. medium/large v microenterprise – # of jobs in the existing supply chain (and skills/wage profile) – will need to consider market shares of participants and how these may change with VC dynamics – Links between firm growth an hiring – combination of hypothetical questions in survey with survey results and relationship between firm output and workforce profile (and K/L allocation) – Local supply expansion / contraction – combination of hypothetical
KEY QUESTIONS AND NEXT STEPS 26
Moving forward § We aim to finalize the pilot survey instrument and guidelines by the end of the month – outstanding questions / issues for input – Managing the estimation v VC development objectives – how to most effectively use the instrument – Short v longer version of the survey – Sector-specific versions? – Stratification strategy – Are there other issues we should probe in the survey? Other ways to get at the issues we are probing? § Survey guidelines to be developed § Initial pilots starting October / November § Revisit following initial pilots
THANK YOU Thomas Farole Email: tfarole@worldbank. org
- Slides: 29