Mapping Developing Countries’ SMART Internationalisation Policy for Inward Foreign Direct Investment: A Scoping Review Protocol

2.1. Review Question

The primary review question is: “What is known about SMART (Sustainable, Measurable, Accountable, Responsive, Transparent) policies in the context of competition for inward foreign direct investment (FDI) in emerging economies, including definitions, instruments, institutional moderators, and social outcomes?”

The sub-questions include the following:

• How are SMART policies defined and operationalised in relation to inward FDI?

2.2. Eligibility Criteria

Eligibility will follow the JBI Population/Participants, Concept, Context (PCC) framework (Peters et al., 2020):

• Participants/Population: Policymakers, investment promotion agencies (IPAs), multinational enterprises (MNEs), or populations in emerging economies affected by inward FDI (e.g., workers, communities). There are no restrictions on age, gender, or sector, but the focus is on aggregate-level impacts rather than individual cases.

Inclusion Criteria:

• (i1) Empirical or conceptual studies on inward FDI in emerging economies (UNCTAD, 2025).

Exclusion Criteria:

• (e1) Outward FDI-focused research (Sattar et al., 2022).

2.3. Search Strategy

The search strategy was developed using the PCC framework. To minimise researcher bias and to systematically identify equivalent terms, we used the litsearchr package in R, following the methodology proposed by Grames et al. (2019) and Tudge (2021). The procedure comprises the following steps: (1) naïve search; (2) construction of a keyword co-occurrence network; (3) identification of change points in keyword importance; (4) creation of a candidate terms list; (5) grouping of terms into PCC-based concept categories; (6) generation of the master search string; and (7) iterative checking and refinement of results.

The process began with a naïve search in Scopus, Web of Science, and EconLit, using simple expressions representing the core elements of PCC (Population, Concept, Context). This initial search, conducted without filters, aimed to identify a set of clearly relevant articles to serve as seed articles. The results were exported in.ris format and imported into R software, where duplicates were automatically removed using litsearchr’s deduplication algorithm. Next, the titles and abstracts of the seed articles were processed using the Rapid Automatic Keyword Extraction (RAKE) algorithm, which generated an initial list of potential keywords. These terms were then analysed in “litsearchr” through the construction of a co-occurrence network, from which measures of term importance (especially frequency and degree centrality) were calculated. Based on these indicators, the (R) package identified a cut-off point to automatically select the most informative terms to represent the review topic. The resulting list was then manually screened by confirming, term by term, their conceptual relevance (TRUE/FALSE), removing irrelevant or overly broad items, and grouping synonyms. After this screening stage, the terms were organised into three PCC-aligned conceptual blocks:

(1) Population/Participants (e.g., “investment promotion agenc*”, “IPA*”, “policymaker*”),

After this step, the write_search function in litsearchr was used to generate the master search string, combining Boolean operators (AND/OR), truncation (e.g., invest*), and quoted compound terms. The resulting search string was as follows: ((“investment promotion agenc*” OR IPA* OR policymaker* OR “policy maker*” OR “government agenc*”) AND (“bilateral invest*” OR “bilateral investment treat*” OR determinant* OR “direct invest*” OR “direct investment inflow*” OR flow* OR “foreign direct investment*” OR FDI OR inflow* OR invest* OR “investment inflow*” OR “investment treat*” OR “outward foreign direct invest*” OR promot* OR “investment promotion” OR “investment facilitation” OR “investment incentiv*” OR “tax incentiv*” OR “fiscal incentiv*” OR subsid* OR “one-stop shop*” OR “single window” OR “special economic zone*” OR SEZ* OR “competition for FDI” OR “location competition” OR “tax competition” OR “corporate tax*” OR “regulator* qualit*” OR “corruption index*” OR “institutional qualit*” OR “ease of doing business” OR “investment climate” OR governance OR “digital FDI” OR “online portal*” OR “e-government” OR “e-governance” OR “foreign invest*”) AND (countr* OR “developing econom*” OR “developing countr*” OR “emerging econom*” OR “global south” OR BRICS OR “Latin America” OR Asia OR Africa))

After validation, the search string was adapted to each database’s syntax (e.g., TITLE-ABS-KEY in Scopus). Filters permitted under PRISMA-S, such as language and document type, were applied only after the searches were executed, without altering the logic of the initial strategy. The results were exported in.ris format, deduplicated in R, and subsequently imported into Rayyan for title/abstract and full-text screening. All searches were documented in accordance with PRISMA-S, including dates, databases, full search strings, and hit counts.

The grey literature was searched via Google Scholar (the first 200 results), institutional repositories (UNCTAD IRIS, OECD iLibrary, and the World Bank Open Knowledge Repository), and through hand-searching the reference lists of the included studies. Forward/backward citation tracking was performed with Litmaps and Google Scholar alerts. Experts from the Global Investment Competitiveness Network were contacted for unpublished reports. Searches were documented in accordance with PRISMA-S, including full search strings, dates and hit counts. For grey literature, a simplified and adapted search string was used: (“investment promotion agency” OR “IPA” OR “policy maker” OR “government agency” “foreign direct investment” OR FDI OR “investment promotion” OR “investment facilitation” OR “tax incentive” OR “special economic zone” OR “competition for FDI” OR “institutional quality” OR “investment climate” “developing country” OR “developing economy” OR “emerging economy” OR “global south” OR BRICS).

The search strategy was developed iteratively with a research librarian and piloted in Scopus to refine terms for sensitivity and specificity (Peters et al., 2020). No date or language limits will be applied initially, but the results will be filtered post hoc. The master search string combines keywords from key references (e.g., “investment promotion”, “competition policy”, “digital FDI”) using Boolean operators (AND/OR), proximity searches (e.g., “FDI NEAR/5 policy”), and truncation (e.g., “invest*”). Synonyms were derived from thesauri (e.g., MeSH for EconLit) and existing reviews on investment facilitation (Peters et al., 2020).

The strategy will be adapted for each database (e.g., subject headings in EconLit). Grey literature will be searched using Google Scholar (the first 200 results), institutional repositories (UNCTAD IRIS, OECD iLibrary, and the World Bank Open Knowledge Repository), and hand searches of the reference lists of the included studies. Forward/backward citation tracking will use Litmaps and Google Scholar alerts. Experts from the Global Investment Competitiveness Network will be contacted for unpublished reports. Searches, including full search strings, dates, and hit counts, will be documented in accordance with PRISMA-S.

The search strategy will be iteratively piloted and refined several times in accordance with established JBI methodological guidance for scoping reviews, which emphasises the importance of preliminary testing to enhance sensitivity, specificity, and overall relevance while minimising researcher bias and ensuring comprehensive coverage of heterogeneous literature (Peters et al., 2020). This iterative process involves initial testing in a primary database such as Scopus, where the draft search string is applied to retrieve a sample of records; these are then reviewed for alignment with the PCC framework, relevance to SMART policies and FDI outcomes, and identification of additional synonyms, missed concepts, or overly broad terms that may introduce noise. Adjustments—such as refining Boolean operators, incorporating proximity operators, adding or removing truncation/wildcards, or incorporating subject headings—are made based on the results, with subsequent re-piloting to validate improvements until the strategy achieves an optimal balance of recall and precision.

The full development process, including pilot iterations, rationale for modifications, and final search strings across all databases, will be transparently documented and reported in its entirety in an appendix (S1 Appendix: Full Search Strategy), drawing on best-practice recommendations for search transparency and reproducibility in scoping review protocols (Moran et al., 2008). This rigorous approach not only strengthens the methodological robustness of the review but also facilitates future replication and supports the identification of emerging patterns in the fragmented evidence base on SMART FDI policies in emerging economies (see: Tables 1 and 2)

OPEN IN VIEWER

Table 1.

Search Strategy Using PCC Framework

PCC keyword block	Search terms
#1 Population (P):	“emerging econom*” OR “developing countr*” OR LMIC* OR “global south” OR “transition econom*”
#2 Concept (C):	“smart polic*” OR “smart regulat*” OR “evidence-based polic*” OR “digital governance” OR “e-government” OR “one-stop shop*” OR “investment promotion agenc*” OR “investment incentive*” OR “regulator* transparenc*” OR “ESG polic*” OR “monitor* compliance”
#3 Context (C/FDI):	“foreign direct investment” OR FDI
#4 Outcomes (social):	“social impact*” OR employment OR “labour standard*” OR wages OR “income inequalit*” OR education OR health OR gender OR environment* OR sustainab*
Master Boolean string (example to adapt in each database syntax):	(#1) AND (#2) AND (#3) AND (#4)

OPEN IN VIEWER

Table 2.

Preliminary Secondary Sources

Organisation	Website	Description
UNCTAD	unctad.org	World Investment Reports, policy monitors, FDI statistics
World Bank	worldbank.org	Governance indicators, investment climate, digital gov.
OECD	oecd-ilibrary.org	Regulatory governance, investment policy reviews
IMF	imf.org	Policy advice, macro context
WTO	wto.org	Trade/investment policy environment
ILO	ilo.org	Labour standards/outcomes
IPA portals (e.g., Invest India, InvestSA)	(various)	One-stop shop/e-gov examples

Source: Authors.

2.4. Source Selection

References will be exported to Rayyan software (https://www.rayyan.ai/) for deduplication via automated algorithms supplemented by manual checks (Ouzzani et al., 2016). Two independent reviewers will screen titles/abstracts using Rayyan’s blinding feature. Inter-rater agreement will be calculated using Cohen’s Kappa (target > 0.7 for substantial agreement) (Aguinis et al., 2023). Disagreements will be resolved through discussion or by consulting a third reviewer. A pilot study of 50 records will calibrate the criteria and refine the process (Smith & Duncan, 2022). Full-text screening will follow, with reasons for exclusion logged (e.g., “no social outcomes” or “no relevant outcomes”). The PRISMA 2020 flow diagram will visualise the process (Page et al., 2021) (see: Figure 1).OPEN IN VIEWER

2.5. Data Extraction

The data will be charted using a standardised, piloted form in Rayyan (or Excel), and tested on 5-10 studies for consistency (Pollock et al., 2023). Two reviewers will independently extract the data, with 20% of the data double-checked for accuracy. The extracted variables include the following:

• Bibliographic: Author, year, title, journal, country/region, publication type.

Missing values will not be formally imputed; instead, disagreements will be settled through consensus.

2.6. Data Analysis and Synthesis

The analysis will be descriptive and narrative, without a meta-analysis. Qualitative thematic analysis will identify patterns in SMART definitions, instruments, and outcomes using inductive coding in NVivo software (initial codes piloted on 10% of the data). Descriptive statistics (e.g., frequencies by year, region, sector) will be used in R (or Excel). Bibliometric analysis using i) VantagePoint (version 17.0), UCINET (version 6.758) and Gephi (version 0.10.1) and alternatively VOSviewer (version 1.6.20) to map keyword co-occurrences (threshold: minimum five occurrences), author networks, and citation clusters, with parameters such as normalisation method (association strength) reported (Aguinis et al., 2023). The synthesis will integrate findings narratively, accompanied by tables/figures that highlight themes (e.g., policy-outcome matrices) and gaps (e.g., underrepresented regions). The enabling/constraining factors will be linked to dependency theory frameworks.

2.7. Quality Assessment

Although not required in JBI scoping reviews (Peters et al., 2020), an optional quality assessment will enhance credibility without excluding studies, thereby allowing for a comprehensive mapping of the diverse and heterogeneous evidence base on SMART policies for inward FDI. This approach aligns with the exploratory nature of scoping reviews, which prioritise breadth over stringent methodological filtering, while adding rigour by systematically evaluating the strengths and limitations of included sources. For qualitative studies, we will employ the Critical Appraisal Skills Programme (CASP) checklist, comprising 10 questions scored on a 0-2 scale per item, with thresholds defining high (≥13/20), moderate (7-12/20), and low (<7/20) quality (Long et al., 2020). The quantitative sources will be assessed using a modified Downs and Black (1998) checklist (27 items, with percentage scores categorizing high ≥70% and low <70% quality), while mixed-methods studies will follow an integrated approach as outlined by Noyes et al. (2019), which combines quantitative and qualitative evidence to explore factors such as effectiveness, acceptability, and contextual influences. By contextualising findings in the discussion—such as highlighting potential biases or methodological gaps in lower-quality studies-this assessment will strengthen the reliability of our thematic synthesis and policy recommendations, ultimately supporting more informed insights into FDI impacts in emerging economies without compromising the review’s inclusive scope.

For qualitative sources, we will utilise the Critical Appraisal Skills Programme (CASP) checklist, a widely recognised tool for evaluating the rigour, credibility, and relevance of qualitative research (Long et al., 2020). This checklist consists of 10 targeted questions that address key aspects, including the clarity of research aims, the appropriateness of the methodology, the recruitment strategy, the data collection methods, the researchers’ reflexivity, the ethical considerations, the rigour of data analysis, the clarity of findings, and the overall value of the research. Each question will be scored on a 0-2 scale, where 0 indicates “no” or “can’t tell,” 1 signifies partial fulfilment (“yes, but with limitations”), and 2 denotes strong alignment (“yes”), yielding a total score out of 20. Based on these scores, studies will be categorised as high quality (≥13/20), reflecting robust methodological practices and trustworthy insights; moderate quality (7-12/20), indicating acceptable but improvable elements; or low quality (<7/20), highlighting significant shortcomings that may affect interpretability. This structured appraisal will not lead to exclusion but will inform our interpretation of the findings, allowing us to weigh the contributions of qualitative evidence more judiciously within the broader scoping review framework, particularly when exploring nuanced social outcomes such as equity and sustainability in FDI contexts.

For quantitative sources, we will apply a modified version of the Downs and Black checklist (Downs & Black, 1998), a validated and widely used instrument comprising 27 items that appraises the methodological quality of both randomised and non-randomised studies of healthcare interventions. This checklist evaluates key domains including reporting clarity, external validity, internal validity (bias and confounding), and statistical power, with most items scored dichotomously (yes/no, awarding 1 point for “yes”) and item 27 on power potentially scaled (e.g., 0–5 points in some adaptations, although often simplified in reviews to 0 or 1). In line with common modifications seen in evidence synthesis (e.g., to enhance applicability across diverse study designs or to address power calculation ambiguities), we will derive an overall percentage score by dividing the total achieved points by the maximum possible (typically 27 or adjusted for non-applicable items), categorizing studies as high quality (≥70%, indicating strong methodological robustness, low risk of bias, and reliable findings) or low quality (<70%, signalling notable limitations such as inadequate reporting, selection bias, or insufficient power that may undermine confidence in results). This structured evaluation will not result in exclusion of studies but will enrich our interpretation of quantitative evidence—particularly empirical analyses of FDI impacts, spillovers, and institutional moderators—by enabling us to highlight potential biases or strengths in the discussion, thereby supporting more nuanced thematic synthesis and policy-relevant insights into SMART policy effectiveness within emerging economies.

This study adopts a mixed-methods design via an integrated approach following Noyes et al. (2019). This involves deliberately combining and synthesising quantitative evidence (e.g., on intervention effectiveness) with qualitative evidence (e.g., on acceptability, feasibility, context, and implementation factors) within a single review framework. The approach clarifies purposes and designs for mixed-method syntheses, particularly for complex interventions and health systems, enabling a more comprehensive understanding of how and why effects occur, while addressing complexity through methods such as juxtaposing findings in matrices, using logic models, or applying frameworks like WHO-INTEGRATE. It supports robust guideline development or evidence-informed decision-making by integrating the two evidence streams at key stages rather than treating them separately. Quality will contextualise findings in the discussion (e.g., noting biases in low-quality studies).

2.8. Ethical Considerations and Dissemination

Ethical approval is not required for this scoping review, as it relies exclusively on secondary data from publicly available peer-reviewed and grey literature, with no involvement of human participants, no primary data collection, and no identifiable personal information. Conflicts of interest have been declared as none by all the authors. To maximise accessibility and impact, the final results of the review will be published open access in a reputable peer-reviewed journal, presented at relevant international conferences (e.g., UNCTAD or World Investment Forum), and disseminated through concise policy briefs targeted at key stakeholders, including policymakers, investment promotion agencies, and development practitioners, in emerging economies. Recognising the rapidly evolving nature of global FDI trends, digitalisation, and sustainability frameworks (e.g., recent UNCTAD reports), the review will be updated biennially if substantial new evidence emerges, ensuring that the conceptual framework and policy recommendations remain current and responsive to ongoing shifts in the field. This commitment to periodic updates aligns with the dynamic policy landscape and supports sustained contributions to evidence-informed decision-making for inclusive and sustainable FDI attraction in developing contexts.