Indonesia’s GDP revision: a crisper snapshot

Indonesia’s national statistics agency (Badan Pusat Statistik, BPS) released quarterly national accounts statistics on February 5. Any quarterly data release creates a flurry of interest (well, at least amongst macroeconomists and economy-watchers hungry for the latest update on near-term growth trends). But this is a particularly important release because, as well as providing data for the final quarter of 2014, it also incorporates two significant revisions to Indonesia’s GDP statistics: (1) it  shifts the basis of the computation from the year 2000 to 2010, and (2) it adopts a significantly updated methodology and presentation of the statistics (updating Indonesia’s national accounts from the 1993 System of National Accounts [SNA] to SNA 2008).[1]

What do these revisions tell us about Indonesia’s economy that we didn’t know before? One change immediately stands out: total output in current prices is about 4.4 percent larger than previously estimated in 2014 (and 5.2 percent larger on average over 2010-2014). This is a significant change, adding IDR 448 trillion, or about USD 35.5 billion at the current market exchange rate, to the estimated size of the economy as of 2014. Roughly a third of the extra measured output is due to the incorporation of new kinds of economic activity under SNA 2008, and about two-thirds comes from more accurate measurements of previously-measured kinds of output, according to BPS.  

Although the improved measurement of output results in a higher level of GDP, it also results in the measured rate of growth of the economy since 2011 being lower, by a significant 0.3 percentage points in 2011, about 0.2 percentage points in 2012 and 2013, and a marginal 0.04 percentage points in 2014.

In short, Indonesia’s economy looks rather significantly bigger, and just a little slower-growing, than previously believed (Table 1).



These changes are especially noteworthy because annual economic output is an essential yardstick against which many other economic variables are gauged. All ratios with the revised GDP in the denominator shrink. Here are some important instances, showing the impact of the revision, and in the case of fiscal data also the new availability of Q4 2014 GDP:

• Indonesia’s current account deficit (or surplus of investment spending over savings): at a cumulative USD 20 billion over Q1-Q3 2014, was equivalent to 3.1 percent of cumulative GDP over those quarters, now a smaller 2.9 percent of GDP.
• Indonesia’s external debt stock: at USD 294.4 billion as of November 2014 (as defined by Bank Indonesia), was equivalent to 34.7 percent of GDP for 2014, now a smaller 33.2 percent
• Indonesia’s fiscal deficit, at IDR 227.4 trillion in 2014 (unaudited), was 2.3 percent of GDP, now a narrower 2.2 percent.
• Indonesia’s tax revenues, at IDR 1,143 trillion in 2014, were equivalent to 11.4 percent of GDP, now a lower 10.8 percent.

As the above examples show, the magnitude of the changes to GDP ratios for Indonesia are not large enough to prompt a rethink of economic conditions and risks. This makes sense, because the best estimate now shows somewhat bigger annual economic output in recent years, but not dramatically so, while nothing else has changed. After all, Indonesia still recorded USD 20 billion worth of net current account outflows in 2014 through the third quarter, and had to finance the same fiscal deficit, as it did before, and the government faces the same, well-recognized challenge to raise more revenues, even if this is now starting from a lower-than-previous tax-to-GDP ratio. [2] In contrast, some recent historical and methodological revisions in other economies have resulted in far greater differences. In April 2014, for example, Nigeria’s GDP revisions approximately doubled the estimated size of the economy, causing it to overtake South Africa’s as Africa’s largest!

Beyond the change to “top line” GDP and hence some major ratios, the revised figures also mark an important step forward in the continual process of improving the statistics tracking Indonesia’s large and rapidly evolving economy. Thanks to the new methodology, more detailed sectoral data are now available, while the historically large discrepancy between production- and expenditure-side GDP shrinks. Economists and policymakers will be chewing over these new data and teasing out their additional implications, including for GDP forecasts, for some time to come.

In the meantime, Indonesia’s GDP revision offers us a useful reminder: economic statistics are complex, and are subject to ongoing methodological improvement and revision. GDP statistics provide a vital series of “snapshots” for tracking the economy and benchmarking key economic stocks and flows, but the economy is so big, complex and dynamic that these snapshots are inevitably a bit blurry and incomplete. [3] You just wouldn’t always think so, to see how financial markets and the media can react to each release, or to comparatively small adjustments or differences in forecasts, including for Indonesia (while often ignoring large subsequent revisions!). Can economy-watchers do more to understand, acknowledge and inform stakeholders about the inevitable limitations of major data, even as these get better over time?