Can industrial policy drive economic diversification? Early evidence from Saudi Arabia

Many resource-rich economies face the same structural challenge: how to build a diversified economy when commodity revenues have long dominated government spending and investment. Industrial policy has emerged as a common response — targeted interventions designed to foster new sectors and integrate domestic producers into global value chains. But does it work? And how can governments know whether their choices are having the intended effect? Saudi Arabia offers a rare opportunity to study this challenge in real time.
 

The Challenge

With hydrocarbons accounting for around 60 percent of government revenues in 2024, Saudi Arabia's dependence on oil has left the economy exposed to commodity price swings and the structural effects of the Dutch Disease. In response, the Kingdom launched an ambitious diversification agenda in 2016 with Vision 2030, deploying industrial policy at scale through state interventions designed to reshape the structure of economic activity by selectively channeling support toward specific sectors. 

Yet no systematic evidence existed on the landscape of these interventions, let alone their effects. Without that, it is difficult to know what is working, what needs adjustment, and where course corrections are needed.
 

The Approach

The new working paper helps fill that gap, combining three complementary approaches.

1. A machine-learning-based industrial policy database was used to map the scale, instruments, and sectoral focus of Saudi Arabia's industrial policy activity.
2. Staggered difference-in-differences methods was applied to UN Comtrade data (2019–2023) to assess the impact of industrial policy interventions on exports.
3. The Orbis firm-level database (2010–2022) was used to capture how industrial policy interventions affected firms, including employment, capital investment, and earnings, in the years following implementation.
4. Together, these produce a layered picture, from macro policy patterns down to how individual firms responded.
    

What the Research Found

The scale and focus of industrial policy in Saudi Arabia has been substantial and evolving. Industrial policy activity surged around 2018–2020, exceeding GCC and global averages (Figure 1). The toolkit is broad, with a comparatively heavy emphasis on investment-oriented instruments (Figure 2). Sectoral targeting has shifted up the value chain — from iron, steel, and chemicals in 2010 toward vehicles, furniture, and consumer goods by 2022 (Figure 3).
 

Figure 1: Saudi Arabia vs. GCC average vs. world average industrial policy over time

^{Source: Yang and Chattha 2026}

Figure 2: Saudi Arabia's industrial policy by policy instruments

^{Source: Yang and Chattha 2026}

Figure 3: Evolution of industry distribution of industrial policy

^{Source: Yang and Chattha 2026}

On exports, no significant effect has been detected over the observed period (Figure 4). Export diversification gains have not yet materialized. This is not necessarily a verdict on the policies — structural transformation takes time, and five years is a short window for measuring shifts of this magnitude. But the absence of early export signals is worth watching, and it underscores why ongoing evaluation matters more than a single snapshot.

Figure 4: Saudi Arabia industrial policy event study on exports

^{Source: Yang and Chattha 2026}

On domestic firm outcomes, early adjustment is visible. Industrial policy is associated with a decline in the number of active firms in the first two to three years following implementation, suggesting shifts in firm composition. Local content requirements are linked to positive employment responses alongside declining capital investment, while domestic subsidies are associated with transitory earnings declines.

Figure 5: Firm level impact of local content requirements on employment, capital and earnings

^{Source: Yang and Chattha 2026}

These results capture short-run adjustment dynamics, not the deeper productivity gains industrial policy is ultimately designed to generate. They should be read as early evidence of where the transformation currently stands.

What This Suggests for Policy

Two lessons emerge that are relevant for Saudi Arabia and countries across the world.

1. Design matters as much as scale. When public support flows toward incumbents or less productive firms, it can suppress the competitive pressure that drives productivity growth. Support that is conditional, performance-based, and time-bound is more likely to catalyze genuine capability building. The question is not just how much to spend, but to whom and under what conditions.
2. Industrial policy cannot work in isolation. Effectiveness depends heavily on the surrounding policy ecosystem - skills, infrastructure, access to finance, and regulatory quality all determine whether firms can respond to the opportunities created. Targeting one constraint without addressing others risks generating short-run compliance without the structural change needed for lasting growth.
    

Beyond Saudi Arabia

Resource-rich economies across the world face similar diversification challenges and are deploying similar policy tools and the questions this paper asks of Saudi Arabia are the same ones those governments should be asking of themselves. The analytical approach developed here, combining machine-learning-based policy databases with trade data and firm-level microdata, offers a replicable template for building the evidence base governments need to adjust their policies in real time.

Perhaps the most important insight from this research is about the value of evaluation itself. Governments that deploy industrial policy at scale without the capacity to measure its effects are flying blind. Building evaluative infrastructure alongside the policies is not a secondary concern. It is foundational to the success of the transformation effort, and a priority for every government serious about making industrial policy work.