Economic and Business Impacts of School Meal Programmes: A Cross-Country Meta-Aggregation
Description
School meal programmes are increasingly framed as economic instruments, yet the evidence on their impacts is geographically fragmented and metrically incommensurable, which impedes cross-country comparison. This study identifies the characteristics of that literature and constructs the architecture of categories, synthesised findings, and transmission pathways through which the impacts arise. A Boolean search of Scopus retrieved 3,064 records, screened under PRISMA 2020 to 50 studies published between 2016 and 2026, of which 41 contributed findings to an adapted meta-aggregative synthesis graded through ConQual. Aggregation resolved the evidence into 25 categories and seven synthesised findings across three pathways: procurement and household carry two findings each, human capital one, and two are cross-pathway. Confidence proves systematically uneven along the demand-supply division. Only maternal labour supply and bargaining position attain a High grade; budget substitution and long-run human capital returns attain Moderate; multipliers, family farm integration, and governance attain Low; and fiscal incidence, resting on two categories, Very Low. The justification most invoked for expanding local procurement therefore rests on the weakest evidence in the corpus.
Files
Steps to reproduce
To submit this systematic review dataset to Mendeley Data, use the following text for the "Steps to reproduce" field to ensure compliance with PRISMA 2020 and Scopus reproducibility standards: ------------------------------ ## 1. Database Search and Retrieval * Open the Scopus database interface using an institutional subscription. * Execute the exact Boolean query in the TITLE-ABS-KEY field: ("school meal*" OR "school lunch*" OR "school feeding" OR "school feeding program*" OR "school food program*" OR "free school meal*" OR "free school lunch*" OR "school nutrition program*" OR "school breakfast program*") AND (impact* OR effect* OR outcome* OR evaluation OR benefit* OR effectiveness) AND (student* OR child* OR adolescent* OR school*) * Apply automatic filters sequentially: Publication Year (2016–2026) $\rightarrow$ Subject Areas (Economics, Econometrics and Finance; Business, Management and Accounting) $\rightarrow$ Document Type (Article) $\rightarrow$ Language (English). * Export the resulting 158 bibliographic records as a .csv or .bib file. ## 2. Title and Abstract Screening * Import the exported records into the screening_matrix.xlsx file. * Screen records independently by two assessors using the five-dimension relevance scoring instrument (Score 0 to 2 per dimension; total score range 0–10). * Retain records scoring $\geq$ 2 (Moderate or High relevance categories) for full-text assessment. * Resolve any inter-rater disagreements via a third assessor, aiming for a Cohen's kappa threshold of $\geq$ 0.61. ## 3. Full-Text Assessment and Extraction * Retrieve full texts for the 55 eligible studies (and resolve any missing text, e.g., re-retrieving Huni et al., 2025). * Apply the predefined inclusion/exclusion criteria from Table 3 (e.g., exclude purely nutritional or clinical studies without an economic framing). * For the 50 included studies, perform data extraction into the finding_extraction_matrix.xlsx: * Layer 1: Extract study characteristics (methodology, country, design, causal strategy). * Layer 2: Extract direct finding statements, metric families, and assign JBI credibility levels (Unequivocal, Credible, or Unsupported). ## 4. Risk of Bias and Aggregation * Assess study quality using ROBINS-I for non-randomised econometric designs, the JBI checklist for qualitative research, and MMAT for mixed-methods. Record justifications in risk_of_bias_matrix.xlsx. * Pool all findings graded Unequivocal or Credible. Group identical findings inductively into 25 categories (minimum 2 studies from different authors per category). * Aggregate categories into 7 action-oriented synthesised findings across the defined transmission pathways. * Apply ConQual rules to determine final confidence grades (High, Moderate, Low, Very Low).
Institutions
- IPB UniversityWest Java, Bogor
Categories
Funders
- IPB UniversityWest Java, Bogor