Indonesia Investment Report 2012 — Spending the Windfall on the Wrong Thing

Indonesia in 2012 was, by consensus, one of the more attractive structural stories in emerging markets, and the case rested on genuine strengths:

• A large and young population, with the working-age share rising for two more decades.
• A domestic-demand-driven economy, which the Asia-Pacific Investment Report 2008 shows was a substantial advantage in a trade shock.
• Abundant natural resources in strong demand.
• And political stability after a difficult transition.

The structural case was correct and the execution was the problem, in a way that is common enough to be worth generalising.

The commodity boom of the preceding decade — coal, palm oil, minerals — had generated large export revenues and substantial fiscal receipts. This created exactly the fiscal room an emerging economy needs to build the infrastructure that raises its potential growth rate.

Most of it was spent on fuel subsidies instead.

The subsidy's fiscal cost was enormous, in several years exceeding capital spending outright. And it was worse than merely unproductive, per the analysis in the Asia-Pacific Investment Report 2014:

• It is regressive, delivering most of its value to higher-consumption households.
• It encourages fuel consumption in an economy that had become a net oil importer, so it widened the current account deficit directly.
• And it is open-ended, growing with world prices — so the cost was largest exactly when the budget was most stressed.

Meanwhile the infrastructure deficit was the binding constraint on growth. Ports, roads, power and rail were insufficient, and the cost appeared as logistics expense rather than as a visible shortage — which made it easy to defer.

The 2013 external episode described in the Asia-Pacific Investment Report 2013 found Indonesia among the most exposed economies, and the exposure was a direct consequence of these choices: a current account deficit widened by subsidised fuel consumption, financed by the portfolio flows that reversed.

A dividend that has to be collected

The demographic argument was central to the investment case and is more conditional than it was usually presented.

The mechanism: as birth rates fall, the working-age share of the population rises for several decades. More workers per dependant raises output per person, and high savings during working years fund investment.

Why it is a window rather than a trend:

• It is temporary by construction. The same cohort that produces the dividend eventually retires, reversing it — the schedule the Asia-Pacific Investment Report 2012 describes.
• It is conditional on employment. A large working-age population is a dividend only if those people find productive work. Unemployed or underemployed, they are a cost and a source of instability.
• And productive work requires capital. A worker with equipment, power, transport and functioning institutions produces multiples of what the same worker produces without them.

A demographic dividend is not delivered by demography. It is an opportunity to employ more people productively, and whether that happens depends on whether the capital exists to employ them with.

This is what made the infrastructure choice consequential rather than merely suboptimal. The dividend's window was open, and the capital required to convert it into output was the thing not being built.

The employment composition tells the story: a large share of the workforce remained in agriculture and informal services, which is precisely the low-productivity employment that the structural transformation described in the Asia-Pacific Investment Report 2012 is supposed to move people out of. That transition requires somewhere for them to go — factories, formal services, cities that function — and building those is a capital problem.

The subsidy that crowded out the future

The trade-off deserves stating plainly because the two sides of it are rarely presented together.

What a fuel subsidy delivers:

• Immediate, visible relief to every consumer of fuel.
• Political support, since the benefit is tangible and universally experienced.
• And protection from world price volatility for households and firms.

What the same money delivers as infrastructure:

• A permanent increase in productive capacity, since a port or road serves for decades.
• Lower logistics costs for every firm, which improves competitiveness broadly.
• Higher potential growth, compounding over time.
• And no ongoing cost once built, unlike a subsidy that must be paid again every year.

The asymmetry is stark:

A subsidy is an annual expense that buys the same thing every year. Infrastructure is a one-time expense that keeps paying. Spending a windfall on the first rather than the second converts a permanent gain into a temporary one.

Why the choice went the way it did is the concentrated-versus-diffuse politics from the Asia-Pacific Investment Report 2012, with the signs reversed:

• Removing the subsidy imposes visible, immediate costs on everyone.
• Building infrastructure delivers diffuse benefits over years, credited to nobody in particular.
• And previous removal attempts had produced serious unrest, making governments reasonably cautious.

The additional trap is that the subsidy's cost rises with world prices, so the fiscal room to build infrastructure shrinks exactly when commodity revenues are highest — the subsidy consumes the windfall automatically.

The Asia-Pacific Investment Report 2014 describes the window that eventually opened when world prices collapsed, and the reform was executed then. The intervening decade's spending was not recoverable.

A constraint that shows up as a cost

Infrastructure deficits are unusual among economic constraints in that they rarely produce a visible shortage. They produce a price.

How the deficit manifested:

• Logistics costs as a share of output were high by regional comparison — the cost of moving goods within the country exceeding the cost of shipping them abroad in some cases.
• Port congestion added days to shipping times, which is working capital tied up.
• Power supply was unreliable, so firms bought generators — a private substitute for public infrastructure, at much higher cost per unit.
• And road transport was slow, which limits how far a firm can economically sell.

Why the cost form makes it easy to defer:

• There is no queue and no visible crisis. Firms adapt, absorb the cost, and continue.
• The cost is distributed across thousands of firms, none of which experiences it as an emergency.
• And the counterfactual is invisible. Nobody observes the factories that were not built because logistics made them uneconomic.

The geographic dimension compounded it. An archipelago has structurally higher infrastructure requirements than a contiguous landmass — the same population served by ports and shipping rather than roads, with no economies of scale from a single national network.

The investment consequence:

• Firms with their own logistics capability had a durable advantage over those dependent on public infrastructure — which favoured large incumbents and foreign firms with capital.
• Domestic market integration was weaker than the national statistics suggest, since regions were more separate economically than politically.
• And the return on infrastructure investment was correspondingly high, which is why it was the obvious use of the windfall and why the failure to make it was so costly.

Forcing value-add, and where the value went

An export restriction on unprocessed minerals, legislated in this period, is a good case study in industrial policy whose incidence differs from its intent.

The rationale was coherent: exporting raw ore captures only the resource's value, while processing captures the manufacturing margin, creates jobs and builds capability. Requiring domestic processing forces the value chain onshore.

The mechanism: ban or heavily tax the export of unprocessed ore, so producers must build smelters domestically or stop exporting.

What actually happens, in sequence:

• Export volumes fall immediately, because the smelters do not exist yet. This is a direct revenue and current account loss in the short term — and it arrived as the commodity cycle was already turning.
• World prices for the affected mineral rise, since a major supplier has withdrawn. This benefits producers in other countries, who capture windfall margins.
• Domestic producers face lower realised prices, since ore that cannot be exported must be sold domestically into a market with limited processing capacity.
• Smelters get built, but slowly — they are capital-intensive with multi-year lead times, and their economics depend on power costs that the infrastructure deficit made unfavourable.

A restriction on exporting an input transfers margin to the producers of that input elsewhere in the world, immediately and reliably. Whether it eventually builds domestic capability depends on conditions the restriction itself does not create.

The conditions required for the policy to work — cheap reliable power, adequate infrastructure, available capital, and technical capability — are largely the same conditions the windfall should have been building. The policy attempted to force an outcome that the missing infrastructure made uneconomic.

The honest assessment is mixed. Processing capacity did eventually get built and the policy is defensible over a long horizon. The transition cost was large, the timing was poor, and the intervening benefit went substantially to foreign competitors.

Why the deficit widened from both ends

The external position deteriorating into 2013 had two causes operating simultaneously, which is why it moved faster than most forecasts.

On the export side:

• Commodity prices were turning, as the demand attribution error described in the Asia-Pacific Investment Report 2009 began to correct.
• Export restrictions reduced volumes in affected minerals.
• And the export base was concentrated in a few commodity categories, so there was little to offset the decline.

On the import side:

• Domestic demand was strong, driven by rising incomes and a growing consumer class — which is good news that shows up as imports in an economy that does not manufacture much of what its consumers want.
• Fuel imports were large and subsidised, so the subsidy was directly funding an import bill.
• And capital goods imports rose with investment.

The result is a current account deficit widening from both directions at once, which is a structurally different situation from a deficit caused by weak exports alone.

The funding composition was the vulnerability, per the Asia-Pacific Investment Report 2013: the deficit was financed substantially by portfolio flows rather than direct investment, and portfolio flows can leave in days.

Every element of this was published in advance — the current account trajectory, its composition, the funding mix, and the subsidy's contribution to fuel demand. The 2013 reversal was a shock in timing only.

Why the consumer story was the durable one

The commodity narrative dominated the investment case, and the more reliable story was underneath it.

The consumer economy's structural drivers, none of which depended on commodity prices:

• A large population moving through the income levels at which consumption patterns change most. The transition from subsistence to discretionary spending produces the steepest growth in categories like packaged goods, financial services, telecoms and transport.
• Urbanisation, which raises both incomes and the propensity to purchase rather than produce at home.
• Very low penetration of formal financial services, banking, insurance and credit, leaving substantial room for growth from a small base.
• And a young population adopting mobile-delivered services faster than infrastructure could deliver physical equivalents.

The last point deserves emphasis because it is a genuine leapfrog. Where physical distribution is expensive — as in an archipelago with an infrastructure deficit — digital delivery is not merely a convenience but a substitute for infrastructure that does not exist. Mobile financial services reach places a branch network cannot economically serve, which is why adoption was faster than in economies with better physical networks.

Why this was more durable than the commodity story:

• It is driven by demographics and income, which move slowly and predictably, rather than by prices set elsewhere.
• It is domestic, so it is insulated from the external channel that transmitted the 2013 reversal.
• And it compounds, since each cohort entering the consuming class stays there.

The commodity cycle determined the current account, the currency and the fiscal position. The consumer transition determined what was actually worth owning for a decade.

The qualification is that the consumer story had its own dependency on the same missing infrastructure. Physical goods still had to move, so logistics costs constrained margins in every category involving distribution. Digital services escaped this; anything requiring delivery did not — which is much of why the region's subsequent venture activity concentrated where it did, as the Southeast Asia Venture Report 2018 describes.

What an allocator could act on

Treat a demographic dividend as conditional on capital formation. A rising working-age share raises output only if the entrants find productive employment, which requires infrastructure and investment that must exist first.

Compare a subsidy's annual cost to what the same money would build once. A recurring expense that buys the same relief each year converts a windfall into a temporary gain; capital spending converts it into a permanent one.

Look for constraints that appear as costs rather than shortages. Logistics expense, private generation and port delays are the observable form of an infrastructure deficit, and they never produce the visible crisis that forces action.

Model the incidence of an export restriction, not its intent. Withdrawing supply raises world prices and transfers margin to foreign producers immediately; domestic capability arrives years later if the enabling conditions exist.

Decompose a widening current account into export and import causes. A deficit driven by strong domestic demand is a different situation from one driven by weak exports, and Indonesia had both at once.

Check deficit funding composition before assessing vulnerability. Portfolio-funded deficits reverse in days; direct-investment-funded ones do not, and the headline deficit number conceals the difference entirely.

What 2012 established

• A demographic dividend requires capital to be collected, and the window is finite.
• Fuel subsidies consumed the commodity windfall, converting a permanent opportunity into recurring consumption.
• The infrastructure deficit manifested as cost rather than shortage, which made it easy to defer indefinitely.
• Export restrictions transferred margin to foreign producers immediately while domestic capability lagged by years.
• The current account widened from both ends, and its portfolio-flow funding set up the 2013 exposure.