Methodology

How we build research from primary sources

SEAIEA's research focuses on regulations, incentives and foreign direct investment in Indonesia, Malaysia, Thailand and Vietnam. We focus particularly on advanced manufacturing, mining and natural resources, and sustainable energy.

The starting point is simple: what a policy says on paper is not always what it means in practice. We work through the underlying rules, data and institutions before comparing jurisdictions or drawing conclusions.

From paper to practice

In Southeast Asia, policies and investment figures that look comparable can mean very different things across countries. A tax incentive may apply to different activities; an investment figure may count approvals rather than money actually invested.

We define the research question first, then establish exactly what we are measuring: what a rule covers, when it takes effect, who it applies to, and what a given number actually counts.

The result is research that can be checked and compared. It is based on primary sources, explicit about uncertainty, and clear about what the evidence does and does not show.

How the evidence is built

Primary sources

We work from laws, regulations, ministry notices, investment-agency guidance and official statistics, in their original language and read in legal and institutional context.

Verification

Automated workflows help us find relevant documents across jurisdictions and sectors. An analyst then checks the source before a finding is accepted and records where it came from.

Reconciliation

Verified findings are recorded with their amendment history, effective dates, verification status and data vintage, so the record can be followed back to the underlying source.

Interpretation

We separate what the evidence directly establishes from what is calculated or inferred, and make clear where the evidence has limits.

What counts as evidence

We start with the source closest to the fact: primary documents where they exist, with credible reporting used to fill gaps.

Primary sources first

Laws, regulations, ministry notices, agency guidance and official statistical releases, read in their original language rather than in translation or summary.

Read in context

We check what a source changes, when it takes effect, who it applies to and what it means for market access or investment incentives.

Company evidence

Corporate filings and announcements are used where they provide evidence on ownership, project status or developments on the ground.

Corroborated secondary

Where primary evidence is unavailable, for example on delays, closures or implementation problems, we use credible reporting and seek independent confirmation where possible.

We search broadly because important developments are easy to miss. We verify narrowly. Automated tools help us find and organise material, but an analyst decides what the evidence means.

How we handle the numbers

Keep different measures separate

Approved, pledged and realised investment, balance-of-payments FDI, FDI stock and project counts are different measures. We don't treat them as interchangeable.

Compare like with like

Before comparing countries or years, we check what each dataset measures and whether the underlying concepts are compatible. If they aren't, we don't combine them.

Who is behind the number?

We record whether figures are preliminary or final, how countries and sectors are classified, and whether a source reflects the immediate investor or the ultimate owner.

Evidence is not causation

A pattern that is consistent with an explanation is not proof that one thing caused another. Where we derive a figure, we retain its underlying source series and denominator.

When the picture isn't complete

Ownership: where ownership cannot be traced comprehensively, we say so rather than imply a cleaner ownership chain than the evidence supports.

Sector classifications: where sector classifications are incomplete or differ between sources, we flag the difference rather than smooth it over.

Approved vs realised investment: where approved and realised investment cannot be matched project by project, we report them as separate series.

Source and date: every finding is linked to its source and date, so the underlying claim can be checked.

See it applied to a real question

Our advisory work runs on the same research base described here. Tell us what you're trying to understand.

Explore advisory →