Why we don’t trust Indonesia’s GDP data - Capital Economics
Emerging Asia Economics

Why we don’t trust Indonesia’s GDP data

Emerging Asia Economics Focus
Written by Gareth Leather

Indonesia’s GDP figures lack credibility – economic growth has been far too stable over the past few years to be believable. We think a better guide to the health of the economy is our own Indonesia Activity Tracker (IAT). This shows that economic growth has slowed sharply since the start of the year, and that the economy is now growing at a slower pace than the official figures suggest.

  • Indonesia’s GDP figures lack credibility – economic growth has been far too stable over the past few years to be believable. We think a better guide to the health of the economy is our own Indonesia Activity Tracker (IAT). This shows that economic growth has slowed sharply since the start of the year, and that the economy is now growing at a slower pace than the official figures suggest.
  • It was no surprise when last week Indonesia reported another quarter of 5% growth. Since the start of 2014, GDP growth has always been reported to be within a range of 4.7% and 5.3% y/y.
  • This stability is hard to square with the number of shocks that have hit the economy over this period, including volatile commodity prices, a sharp fall in the rupiah and an aggressive rate-tightening cycle which has subsequently been partly reversed. In recent years growth in Indonesia according to the official figures has been much more stable than in any other major economy.
  • In order to construct the GDP figures, Indonesia’s statistical agency (BPS) starts with production data to calculate gross value added (GVA), before adding “taxes less subsidies” to get GDP. The BPS separately estimates an expenditure breakdown and introduces a statistical discrepancy to make the GDP figures align with GDP as estimated via the production method. In other words, the BPS regard the production method as the most reliable estimate of GDP, which is quite common for emerging markets.
  • Part of the problem lies in the “taxes less subsidies” on products (TLS) component, which makes the GDP growth figures exceptionally stable. Since 2014 the TLS component has had an almost perfect inverse relationship with GVA. In other countries, the relationship is positive.
  • It is of course possible that all of this is just a coincidence. But it is striking that the unusual pattern first emerged at the time that growth in GVA slowed. It’s surely possible that the statistical agency has been trying to massage the figures.
  • The expenditure data have their own problems. The biggest issue lies with the household consumption figures, which account for 55% of GDP. Rather than deriving the figures from household spending surveys, consumption growth is simply extrapolated from the GDP figures. With reported GDP growth stable, household consumption growth has been almost completely flat at 5.0% over the past few years. The figures are of little help as a guide to the strength of consumption.
  • To provide an alternative (and better) guide to the health of the economy, we have developed our own Tracker. The IAT is constructed using low-profile monthly indicators, that are less likely to have been meddled with by the statistical authorities. The Tracker suggests that, not only has the economy slowed sharply in recent quarters, but that it is growing at a much weaker pace than the official figures suggest.
  • Looking ahead, the official figures, for what they are worth, will probably continue to show the economy growing at around 5% over the next few years regardless of how the economy is actually performing. In contrast, we think the IAT will show the economy growing at a much weaker rate.

Why we don’t trust Indonesia’s GDP data

Indonesia’s GDP figures lack credibility – economic growth in recent years has been far too stable to be believable. We think a better guide to the health of the economy is our own Indonesia Activity Tracker (IAT). This shows that economic growth has slowed sharply since the start of the year, but that the economy is now growing at a slower pace than the official figures suggest.

GDP growth has been too stable to believe

We have long held the view that Indonesia’s GDP figures lack credibility, and it was no surprise when last week Indonesia reported another quarter of 5% growth. (See our Data Response, “Official figures overstate health of economy”, 5th November.)

Since the start of 2014 GDP growth has been reported within a range of 4.7% and 5.3% y/y. (See Chart 1.) This stability is hard to square with the number of shocks that have hit the economy over this period, including volatile commodity prices, a sharp fall in the rupiah and an aggressive rate-tightening cycle which has subsequently been partly reversed.

Chart 1: Indonesia GDP (% y/y)

Sources: Refinitiv, Capital Economics

Over this period growth in Indonesia according to the official figures has been much more stable than in any other economy.

The best way of measuring volatility is by calculating the coefficient of variation, which is the ratio of the standard deviation to the mean. A lower coefficient of variation corresponds to a less volatile series. The official figures make Indonesia the most stable economy in the world. (See Chart 2.)

Chart 2: Coefficient of Variation (GDP growth, 2014–present)

Sources: Refinitiv, Capital Economics

Data don’t correlate with other activity measures

We would normally expect GDP figures to correlate closely with other measures of activity, but this isn’t the case in Indonesia. For example, in most countries there is a close relationship between import growth and overall GDP growth. Chart 3 shows the example of Korea.

Chart 3: Korea GDP & Imports Volumes (% y/y)

Sources: Refinitiv, Capital Economics

The relationship is not a surprise given that an important driver of import demand is the overall strength of the economy. But in the case of Indonesia, there is no relationship at all between imports and GDP. (See Chart 4.)

Chart 4: Indonesia GDP & Imports Volumes (% y/y)

Sources: Refinitiv, Capital Economics

How are the GDP figures constructed?

In order to construct the GDP figures Indonesia’s statistical agency (BPS) starts with the production series to calculate gross value added (GVA), and then add “taxes less subsidies” to get GDP. The BPS then separately estimates an expenditure breakdown and introduces a statistical discrepancy to make the figures align with GDP as estimated via the production method. In other words, the BPS regard the production method as the most reliable estimate of GDP, which is quite common for emerging markets.

Looking at the production breakdown, while some of the components do exhibit a fair amount of volatility, most sectors are unusually stable. This is demonstrated by Chart 5 below, which shows the coefficient of variation for each major sector of the economy. More important, however, is that each of the different sectors net out, so that growth in overall GVA is much less volatile than the individual sectors.

Chart 5: Coefficient of Variation (GDP growth, 2014 – latest)

Sources: Refinitiv, Capital Economics

Another concern we have is with the “taxes less subsidies” on products (TLS) component, which makes the GDP growth figures exceptionally stable.

Chart 6: Gross Value Added & Taxes Less Subsidies Contribution to y/y GDP Growth (%-points)

Sources: Refinitiv, Capital Economics

Since 2014 the TLS component has had an almost perfect inverse relationship with GVA. Chart 6 shows that any increase or decrease in GVA’s contribution to GDP growth over this period has been offset by an opposite move in the contribution from TLS.

This inverse relationship is unusual. Normally, when GVA increases, more taxes are paid on products (and the taxes paid on products usually exceed the subsidies provided by governments on these products). In most emerging markets, there is a positive relationship between TLS and GVA. (See Chart 7.)

Chart 7: Correlation Coefficient (GVA & TLS, 2014-latest)

Sources: Refinitiv, Capital Economics

It is of course possible this is just a coincidence. But it’s striking that this issue started to occur at the time that growth in GVA slowed. It’s surely possible that the statistical agency has been trying to massage the figures.

Expenditure data have their own problems

The expenditure data have their own issues. One is that imports and exports are deflated by an exchange rate index, not the price of goods and services that are imported and exported.

Our biggest concern, however, is with the household consumption figures, which accounts for 55% of GDP. Rather than calculating the consumption figures from household spending surveys, consumption growth is simply extrapolated from the GDP figures (as measured via the production method).

With reported GDP growth stable, household consumption growth has been almost completely flat at 5.0% over the past few years and is of little use as a guide to the strength of consumption. Consumption figures from other countries show much more cyclicality. (See Chart 8.)

Chart 8: Private Consumption (% y/y)

Source: Refinitiv

An alternative measure of activity

To provide an alternative (and better) guide to the health of the economy, we have developed our Indonesia Activity Tracker (IAT), which is constructed using low-profile monthly indicators, that are less likely to have been meddled with by the statistical authorities. 

The IAT suggests that not only has the economy slowed sharply in recent quarters, but that it is growing at a much weaker pace than the official figures suggest. It also indicates that while there have been points over the past few years where growth has reached nearly 6%, for Jokowi’s first term as a whole, growth came in well below the government’s 7.0% target. (See Chart 9.)

Chart 9: Indonesia GDP & Activity Tracker (% y/y)

Sources: Refinitiv, Capital Economics

The performance of the IAT fits closely with what has happened to other key economic indicators. For example, the Tracker shows a sharp slowdown in activity in 2014-15, the period when commodity prices slumped. The IAT also points to a pickup in economic growth last year, a period when the world economy was also growing more strongly.

What to expect next for the coming year

The official figures, for what they are worth, will probably continue to show the economy growing at around 5% in 2020 regardless of how the economy is really performing. In contrast, we think growth according to the IAT will be much lower. (See our Focus, “What to expect from Jokowi 2.0?”, 15th October.)


Gareth Leather, Senior Asia Economist, +44 20 7811 3916, gareth.leather@capitaleconomics.com