Consultant on Gravity Utility

Client of United Nation careers

Background ESCWA has developed the Gravity Utility for the Assessment and Reduction of Disguised Investments (G.U.A.R.D.) as a corridor-level screening approach that flags statistically anomalous patterns in recorded FDI positions involving Arab economies. The current prototype relies on the IMF Direct Investment Positions by Counterpart Economy (DIP, formerly CDIS) dataset, which reports end-year bilateral FDI positions by immediate counterpart economy on the directional principle, with equity and intercompany-debt detail, aligned with IMF Balance of Payments and International Investment Position Manual, Sixth Edition (BPM6) and the OECD Benchmark Definition of Foreign Direct Investment, Fourth Edition (BMD4). The screening logic integrates: 1) gravity-based anomaly detection (PPML with high-dimensional fixed effects) to benchmark corridor-year positions against fundamentals, 2) mirror-asymmetry diagnostics aligning inward positions with partner-reported outward mirror positions, and 3) a persistence-weighted composite index (EWMA) that prioritizes sustained patterns and incorporates a gross-to-net pipeline signal consistent with pass-through mechanics. The methodology and results are documented in two draft outputs: GUARD methodology note, and GUARD results draft paper ( The phantom footprint: a corridor-level FDI risk assessment in the Arab region ). The analysis is designed as an early-warning screening and prioritization tool to support follow-up actions such as statistical reconciliation, transparency/substance checks, and targeted review of treaty and incentive design. ESCWA seeks an independent peer review to strengthen methodological robustness, transparency, and policy-safe interpretation prior to further internal review and dissemination.

Duties and Responsibilities Under the overall guidance of the ESCWA Economic Advisor and Head of Financing for Development Office, the consultant will conduct the following tasks:

• Provide a technical and interpretive peer review of both documents (including tables, figures, boxes and annexes), ensuring internal consistency of concepts, notation, assumptions, and narrative across the methodology note and the paper draft.
• All comments must be specific, detailed, and operational, indicating exactly what changes are needed and where (e.g., wording to revise, definitions to add/adjust, tables/figures to modify, robustness checks to implement, caveat language to strengthen). Generic feedback (e.g., clarify, add robustness, check other data ) is not sufficient.
• Where proposing cross-checks or alternative datasets, the consultant must indicate whether the data exist, are accessible, and whether their dimensionality/coverage are sufficient for the intended comparison. The peer review is multi-disciplinary. The consultant is expected to provide in-depth review contributions in at least two of the thematic areas listed below, consistent with their expertise, while ensuring that comments support an integrated interpretation of findings and policy-safe messaging. Where a topic falls outside the consultant s core expertise, the consultant should flag this explicitly and provide high-level observations and/or indicate where complementary expertise may be needed.
• Data inputs and statistical perimeter (IMF DIP/CDIS, mirrors, and complementary sources for cross-checking)
• Assess whether the description and use of the IMF Direct Investment Positions by Counterpart Economy (DIP, formerly CDIS) dataset is accurate and sufficiently transparent (including: directional principle, immediate counterpart economy reporting, end-year positions, and the equity vs. intercompany debt split), and clearly state what the dataset can and cannot capture.
• Review the use and interpretation of mirror positions and associated caveats (coverage constraints, confidentiality practices, allocation differences, and reporting frictions), and assess whether the paper appropriately communicates these limitations.
• Recommend feasible cross-checks / broader validation using complementary data sources (where relevant and available), including UNCTAD and other established databases the reviewer has experience with (e.g., OECD FDI statistics and/or national IIP/BoP publications), to support interpretation and improve robustness of policy messaging. For any proposed cross-check, the reviewer must specify: o the exact dataset/source, o the unit of observation and dimensionality (e.g., corridor-year vs. country-year), o time coverage and expected sample size, o variables and matching keys required, o whether the data are available, accessible, and adequate for the intended test. o Where corridor-level validation is not feasible due to data dimensionality, the reviewer can propose an aggregation-consistent alternative (e.g., country-year plausibility checks) and clearly document conceptual differences that limit comparability.
• Econometric strategy and implementation choices (gravity/PPML).
• Review the robustness strategy (including use of alternative model variants and any corroboration/overlap logic across variants), screening logic, thresholds and score construction.
• Review the anomaly diagnostics and composite structural score (component scaling, weights, damping transformations). Validation exercises, correlation/triangulation options, and interpretive caveats
• Review any validation components (e.g., treaty-feature screens and regulatory correlation exercises).
• Propose additional feasible correlation/triangulation exercises (non-causal) that could strengthen interpretability and confidence in the screening signal, while remaining consistent with GUARD as a screening/diagnostic tool.