Abstract
Selected Issues
Indonesia’s Inflation Dynamic During the Covid-19 Pandemic1
This note takes stock of Indonesia’s recent inflation dynamics, analyzing the main drivers behind the recent disinflation and their implications for the outlook. The analysis suggests that the pandemic has strongly reinforced a disinflationary trend that was already underway. The disinflationary effects of the pandemic appear to stem mainly from aggregate and sectoral demand shocks, which have added to disinflationary pressures from positive supply shocks and idiosyncratic sectoral shocks. Staff projects inflation to stabilize in the coming months before picking up later in 2021.
A. Introduction
1. Indonesia has experienced significant disinflation over the past year, reflected in both consumer and producer price indices (left chart below). Headline CPI inflation fell to 1.7 percent y/y in December 2020 from 3 percent in March, below Bank Indonesia’s target band of 2 to 4 percent (right chart below). Core inflation declined to 1.6 percent from 2.9 percent over the same period. Much of the downward shift in inflation rates occurred against the backdrop of the COVID-19 pandemic, which buffeted the Indonesian economy through a multitude of channels.
CPI and PPI Inflation
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); CEIC Data Ltd.; IMF staff estimates.Headline and Core Inflation
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Bank Indonesia: CEIC Data Ltd.; IMF staff estimates.2. The pandemic has affected most other emerging market (EM) economies around the same time and via similar channels as Indonesia (Ebrahimy and others 2020). Therefore, one could expect that inflation across a broad range of EMs may display similar patterns. Among a sample of 22 EMs, most countries experienced some disinflation in the initial lockdown phase, but generally saw a reacceleration once the economy began to recover (see chart).2 This pattern is particularly pronounced among other ASEAN countries like Malaysia and Thailand (though not the Philippines), which saw 12-month inflation rates turn negative amid the lockdown.
CPI Indonesia and Emerging Markets 1/
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia(BPS): Haver Analytics; IMF staff estimates.1/ Darker shade indicates 25th to 75th percentiles, lighter shade indicates 10th to 25th and 75th to 90th percentiles.3. The U-shaped pattern generally also holds for producer prices of ASEAN countries, which also saw a notable drop during the early phase of the pandemic (see chart). Of note, Indonesia’s producer price inflation had been on a firm downward path since well before the pandemic, falling from an average of 3 percent in 2016–18 to 0.5 percent at end-2019 and 0.3 percent in September 2020.
PPI Indonesia and Emerging Markets 1/
(In percent, year-an-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Source: Statistics Indonesia (BPS), Hauer Analytics, IMF Staff Calculation.1/ Darker shade indicates 25th to 75th percentiles, lighter shade indicates 10th to 25th and 75th to 90th percentiles.B. Inflation During the Pandemic: Phases and Drivers
4. The COVID-19 pandemic has severely disrupted economic activity in Indonesia and the rest of the world. Lockdowns and other social restrictions have affected supply and demand through several channels that interact in complex manners. For the Indonesian economy as a whole, the onset of the pandemic sharply reduced aggregate demand, as agents were unable or unwilling to consume and/or invest. At the same time, the pandemic affected different sectors in very different ways, depending on how contact-intensive particular sectors are and whether workers are able to work from home (Del Rio-Chanona and others 2020). For example, the hotel and accommodation industry has been among the most affected, while agriculture has been among the least affected. The asymmetric impact on different sectors is mirrored in asymmetric impacts across geographic regions, as some sectors are important for certain regions but not for others. For example, Jakarta has a high share of business services in its GDP while Kalimantan is mining-oriented and Bali is heavily reliant on tourism.
5. In order to assess the likely impact on inflation, it is useful to distinguish several phases of the epidemic in Indonesia and consider how different drivers are likely to have affected inflation in each of these phases. Table 1 presents a stylized illustration of the phases and drivers considered to be most relevant.
Schematic: Drivers of Inflation during the COVID-19 Pandemic
Schematic: Drivers of Inflation during the COVID-19 Pandemic
| Stage of Pandemic | Pre-COVID-19 | Lockdown | Early recovery | Transition to new normal | |
|---|---|---|---|---|---|
| Aggregate demand | below trend | sharply lower | gradually recovering | approaching new equilibrium | |
| Sectoral demand shocks | -- | sharp drop and shifts | partial reversal/recovery | approaching new equilibrium | |
| Output/employment gap | small negative gap | gapwidening sharply | large negative gap | gap gradually closing | |
| Food supply | fluctuating | strong harvest | food supply normalizing | uncertain | |
| Administered prices | small increasees | small price increases | small price decreases | price increases likely small but growing | |
|
|
|||||
Schematic: Drivers of Inflation during the COVID-19 Pandemic
| Stage of Pandemic | Pre-COVID-19 | Lockdown | Early recovery | Transition to new normal | |
|---|---|---|---|---|---|
| Aggregate demand | below trend | sharply lower | gradually recovering | approaching new equilibrium | |
| Sectoral demand shocks | -- | sharp drop and shifts | partial reversal/recovery | approaching new equilibrium | |
| Output/employment gap | small negative gap | gapwidening sharply | large negative gap | gap gradually closing | |
| Food supply | fluctuating | strong harvest | food supply normalizing | uncertain | |
| Administered prices | small increasees | small price increases | small price decreases | price increases likely small but growing | |
|
|
|||||
Pre-COVID-19 phase: In the period leading up to the pandemic, inflation in Indonesia was on a downward trend, likely reflecting a slowdown in aggregate demand and some degree of slack in the economy. In addition, Bank Indonesia’s preannounced lowering of the inflation target by ½ percentage point at end-2017 and end-2019 with increased policy credibility likely also contributed to the lower inflation readings. Looking at specific CPI components, prices of volatile food (16 percent of CPI basket) were fluctuating but had little net effect on headline inflation until early 2020 (see chart). Administered price increases (18 percent of CPI basket) were relatively subdued.
Lockdown phase: When the spread of COVID-19 in Indonesia began to escalate, the Indonesian authorities imposed tight social restrictions to mitigate the epidemic. Aggregate demand plummeted, exerting downward pressure on prices, while supply bottlenecks appear to have been limited. Some sectors were more severely affected than others, and a few sectors even saw increased demand, such as personal care and hygiene products. Administered prices were little changed during the lockdown phase and into the early recovery phase, contributing to disinflationary pressure.3
Early recovery: The economy began to recover around June/July 2020, with Q3 real GDP growing at 12.4 percent (q/q s.a.a.r). Month-on-month changes in consumer prices continued to fall, however, dipping into negative territory. These declines would be consistent with the view that the rebound in demand has not yet significantly reduced substantial economic slack. Nevertheless, sectoral shocks began to reverse during this period, and downward pressure on prices at the sectoral level likely started to subside.
Transition to new normal: As the pandemic eases, demand is expected to recover and the output gap should narrow over time. Inflation is expected to pick up gradually as a result, with the speed depending on the pace of the economic recovery and several idiosyncratic drivers, including food supply dynamics and the path of administered prices.
Component of CPI Inflation
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); and IMF staff estimates.6. Beyond the disruptions to economic activity, it is worth noting that the COVID-19 pandemic also introduces complications regarding the measurement of inflation. An emerging literature finds that consumption baskets changed substantially during the pandemic, and the lockdown phase in particular (e.g., Cavallo, 2020; Reinsdorf, 2020; and Seiler, 2020). This is in contrast to the standard fixed basket weights typically used to measure inflation. As a result, standard measures of inflation may understate “true” price increases during the pandemic because the items that are in high demand during the pandemic will generally see higher price increases and vice versa (Reinsdorf, 2020). A separate study by Jaravel and O’Connel (2020) finds evidence that there was indeed an initial spike in consumer prices during the lockdown, based on high-frequency data for advanced economies. International evidence suggests that the mismeasurement from fixed inflation baskets can be significant but mostly temporary, arguing against nonstandard changes to inflation baskets (Reinsdorf, 2020).
C. A Disaggregate Analysis of Consumer and Producer Prices
7. In order to investigate why Indonesia’s inflation rates have continued to decelerate in recent months, we analyze component and subcomponent data of CPI and PPI indices.4 Most components of the CPI contributed to the disinflationary trend, though food prices accounted for 35 subcomponents (such as “food crops”). The Indonesian statistics office does not publish more disaggregated price data.
8. A look at subcomponent data confirms that there has been a downward shift in inflation rates for most goods and services (see chart). The downward shift is particularly pronounced on the upper end of the distribution, i.e., for CPI subcomponents that had previously seen relatively elevated inflation rates. Some examples include education, restaurants, and other services, which saw declines in year-on-year inflation rates of 2 to 5 percentage points. Notably, each of these components relate to contact-intensive sectors, consistent with the notion that these sectors have been disproportionately affected by the pandemic.
Contribution to Headline Inflation
(In percent, of year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); Data Ltd; IMF staff estimates.CPI Subcomponents 1/
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); Haver Analytics; IMF staff estimates.1/ Darker shade indicates 25th to 75th percentiles, lighter shade indicates 10th to 25th and 75th to 90th percentiles.9. Subcomponent data shed further light on inflation dynamics during the lockdown and subsequent recovery. We conduct an analysis similar in spirit to the work by Banerjee and others (2020), which looks at “inflation at risk” from COVID-19, using a quantile regression approach to estimate the probability distribution of inflation outcomes during the pandemic (see also Lopez-Salindo and Loria 2020). We use monthly data on the 90 subcomponents of CPI data and fit a skewed-t probability distribution—proposed by Azzalini and Capitanio (2003)—which is characterized by 4 moments: mean, variance, skewness, and kurtosis. The skewed-t distribution is a flexible function that nests both normal and standard t-distribution. Thus, it allows us to stay broadly agnostic about the shape of the distribution of inflation densities.
10. The analysis suggests that inflation densities shifted to the left in 2020:Q2, the quarter when the lockdown took place. The inflation density in Q2 (see chart) narrowed somewhat compared to the previous quarter, suggesting that the initial shock was relatively uniform across inflation components (and thus economic sectors). Subsequently, the density of inflation readings widened in 2020:Q3, suggesting that the early recovery phase was marked by greater dispersion in how different economic sectors were affected. Looking ahead, it will be important to monitor whether this greater dispersion of inflation outcomes at the sectoral level continues.
Fitted Densities of CPI Inflation Subcomponents
(Annualized, in percent, m/m)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); and IMF staff estimatesD. Regional CPI Data Point to Broad-Based Disinflation with Some Geographic Differentiation
11. Regional inflation data suggest that the COVID-19 shock has had broad-based effects across Indonesia while also pointing to the importance of asymmetric shocks (see chart). CPI data for 90 cities show that disinflation has been observed in most regions of Indonesia. At the same time, larger cities have generally seen more elevated inflation rates than smaller cities, notably the five cities with biggest CPI weights: Jakarta, Bekasi, Surabaya, Depok, and Tangerang.
CPI in 90 Cities 1/
(In percent, year on year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); Haver Analytics; IMF staff estimates.1/ Darker shade indicates 25th to 75th percentiles, lighter shade indicates 10th to 25th and 75th to 90th percentiles.12. Several other characteristics appear to have shaped regional inflation dynamics. For example, the cities below the 10th percentile in the probability distribution were mostly located outside of Java, particularly in Sumatra island. The common cause of the low inflation (or deflation) in those cities appears to have been a significant decline in the prices of air transport and the food, beverages and tobacco component, likely due to restricted economic activities and mobility amid the pandemic. Meanwhile, the cities above the 90th percentile in the probability distribution include the East Nusa Tenggara province and the Papua province. The more elevated inflation rates appear to have been driven by food commodities, given that supplies to these regions largely come from other parts of Indonesia. Furthermore, in tourism-dependent Bali, the city of Denpasar saw inflation fall from 2.2 percent in June to 0.8 percent in September, reflecting the sharp decline in economic activity, with regional GDP contracting by 12 percent y/y in 2020:Q3.
E. Producer Prices: Mining and Utilities in Deflation
13. Data on producer price inflation shed additional light on inflation dynamics as they allow for a more targeted analysis of different economic sectors. Indonesia’s producer price index shows inflation falling from 4.2 percent y/y in 2018:Q3 to 0.7 percent in 2020:Q1 (at the onset of the pandemic) and -0.7 percent 2020:Q2, suggesting that the pandemic has reinforced a disinflationary trend that was already underway. In 2020:Q3, PPI inflation recovered to 0.3 percent. Sector-level data show that mining and utilities played a key role in driving down the overall PPI index, with the two sectors registering deflation of 10.8 percent and 0.6 percent y/y, respectively, as of September 2020 (see chart). Downward pressure in the mining sector was mainly driven by falling coal prices (coal exports account for 20.9 percent of total exports), which dropped 51 percent from their peak in September 2018.
Components of PPI Inflation
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); IMF staff estimates.14. The disproportionate role of the mining and utilities sectors is evident when looking at the dispersion of inflation readings (see chart). Most PPI sectors have seen only limited disinflation in recent quarters and the overall PPI index is well below the median of the distribution, reflecting the heavy weights of sectors in that are experiencing deflation.
PPI Subcomponents, All Sectors 1/
(In percent year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (EPS); Haver Analytics; IMF Staff estimates.1/ Darker shade indicates 25th to 75th percentiles, lighter shade indicates 10th to 25th and 75th to 90th percentiles.F. Property Prices
15. Property price data show limited disinflation in recent quarters. Property price increases averaged around 2 percent for much of 2019, before dipping to 1.2 percent in 2020:Q3 (see chart). The largest contractions in 3Q were seen in Batam at the border of Singapore, as well as Balikpapan (East Kalimantan), and Banjarmasin (South Kalimantan). Larger contractions were mostly occurred in large and medium properties.
Batam was severely affected by declining tourism and trade due to the pandemic, reflecting its close proximity to Singapore. Residential property price inflation fell 3.8 percent in March and 1.6 percent in September.
Balikpapan, the largest city of East Kalimantan, is the home of the largest coal and oil producer in Indonesia. Economic activity of this region was affected by weaker coal demand and global coal prices, which dragged down the mining sector (33 percent of regional GDP) to contract by 6.2 percent y/y in Q2. Residential property prices deflation in Balikpapan stood at 0.4 percent y/y in September, down from 0.1 percent deflation in March.
Banjarmasin, the capital city of South Kalimantan, is one of the largest crude palm oil and coal producers in Indonesia. Its economy contracted 2.6 percent y/y in Q2, dragged down by weaker demand in the mining sector amid the pandemic. Residential property price inflation fell from 0.5 percent in March to -0.4 percent in September.
Residential Property Prices in 18 Cities 1/
(In percent, year-on-year)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); Haver Analytics; IMF staff estimates.1/ Darker shade indicates 25th to 75th percentiles, lighter shade indicates 10th to 25th and 75th to 90th percentiles.G. Takeaways and Outlook
16. Overall, the analysis suggests that disinflationary pressures in Indonesia have resulted from the complex interplay of a multitude of shocks, many of which relate to the pandemic while others do not. Most of these shocks are expected to dissipate gradually over time, which is likely to result in inflation returning back towards its pre-COVID-19 level (see also the discussion in Goodhart and Pradhan, 2020). This is reflected in staff projections for a gradual return of inflation over the course of 2021 (see chart), with headline inflation is projected to climb from 1.7 percent in December 2020 to 3 percent next year, while core inflation is forecasted to climb from 1.6 percent in December 2020 to 2.6 percent at end-2021. These projections are broadly consistent with Bank Indonesia’s forecast, which also envisions inflation to gradually return to the 3.0%±1% target range in 2021 (Bank Indonesia 2020).
Headline Inflation
(12-month percent change)
Citation: IMF Staff Country Reports 2021, 047; 10.5089/9781513570860.002.A002
Sources: Statistics Indonesia (BPS); and IMF staff estimates.17. Factors affecting the outlook for staff inflation projections include:
Aggregate demand has started to recover, reflected in the expansion of economic activity in 2020:Q3 relative to the prior quarter. While high frequency data suggest that growth has slowed in 2020:Q4, the recovery is expected to regain some traction in coming quarters.
Sectoral demand shocks have begun to unwind, with sectors that saw the sharpest contractions in Q2 being among those with the strongest increases in economic activity in Q3, such as the transportation and the travel and accommodation sectors. Similarly, these sectors are likely to see inflation recover as the COVID-19 shock eases.
The mining sector stands out as having been subject to the largest idiosyncratic shocks, stemming from the sharp drop in the prices of coal and other commodities since late 2018, reinforced by the pandemic early in 2020. These commodity prices have generally recovered from their trough in spring, reflected in a pickup in producer prices in Q3 that should continue in the period ahead.
18. There are significant upside and downside risks to this forecast. These risks include the possibility that the pandemic will weigh on economic activity for longer than anticipated and that disinflationary forces become more entrenched, which could result in sustained undershooting of Bank Indonesia’s inflation target. On the other hand, a more vigorous economic rebound could result in price pressures building earlier than expected, lifting inflation more quickly.
19. Finally, it bears emphasis that the COVID-19 pandemic is an exceptional event whose effects on output and inflation dynamics are highly complex and difficult to forecast. Given the protracted nature of the pandemic, its deleterious effects are likely to be with us for some time to come and much additional research will be needed to investigate the interplay of shocks that drive inflation dynamics during and after this period.
References
Azzalini, A., and A. Capitanio, 2003, “Distributions Generated by Perturbation of Symmetry with Emphasis on a Multivariate Skew t Distribution,” Journal of the Royal Statistical Society, Series B, Vol. 65, pp. 367–389.
Banerjee, Ryan Niladri, Aaron Mehrotra, and Fabrizio Zampolli, 2020, “Inflation at Risk from COVID-19,” BIS Bulletin, No 28 (Basel: Bank for International Settlements).
Bank Indonesia, 2020, “BI 7-Day Reverse Repo Rate Held at 4.00%: Synergy to Accelerate National Economic Recovery,” Press Release No. 22/75/DKom (October). Available via the Internet: https://www.bi.go.id/en/ruang-media/siaran-pers/Pages/sp_227520.aspx.
Cavallo, Alberto, 2020, “Inflation with COVID Consumption Baskets,” Harvard Business School BGIE Unit Working Paper No. 20–124 (Cambridge: Harvard Business School).
Del Rio-Chanona, R. Maria, Penny Mealy, Anton Pichler, Lafond Francois, and J. Doyne Farmer, 2020, “Supply and Demand Shocks in the COVID-19 Pandemic: An Industry and Occupation Perspective,” Oxford Review of Economic Policy, Vol. 36, Number S1, pp. S94–S137.
Ebrahimy, Ehasn, Deniz Igan, and Soledad Martinez Peria, 2020, “The Impact of COVID-19 on Inflation: Potential Drivers and Dynamics,” IMF Special Notes Series on COVID-19, September 20 (Washington: International Monetary Fund).
Goodhart, Charles, and Manoj Pradhan, 2020, “Future Imperfect after Coronavirus,” VOX CEPR Policy Portal. Available via the Internet: https://voxeu.org/article/future-imperfect-after-coronavirus.
Jaravel, Xavier, and Martin O’Connel, 2020, “Real-Time Price Indices: Inflation Spike and Falling Product Variety During the Great Lockdown,” Journal of Public Economics, Vol. 191, November, 104270. Available via the Internet: https://doi.org/10.1016/j.jpubeco.2020.104270.
Lopez-Salido, David, and Francesca Loria, 2020, “Inflation at Risk,” Finance and Economics Discussion Series 2020–013 (Washington: Board of Governors of the Federal Reserve System). Available via the Internet: https://doi.org/10.17016/FEDS.2020.013.
Reinsdorf, Marshall, 2020, “COVID-19 and the Weights of the CPI: Is Inflation Underestimated?,” paper presented at the 8th IMF Statistical Forum “Measuring the Economics of a Pandemic,” Washington, November.
Seiler, Pascal, 2020, “Weighting Bias and Inflation in the time of COVID-19: Evidence from Swiss Transaction Data,” Swiss Journal of Economics and Statistics, Vol. 156, No. 1, pp. 1–11.
Prepared by Robin Koepke and Rani Setyodewanti (APD).
The economies included in this analysis are Brazil, Chile, China, Colombia, Czech Republic, Egypt, Hungary, India, Indonesia, Lebanon, Malaysia, Mexico, Nigeria, Philippines, Poland, Russia, Saudi Arabia, South Africa, South Korea, Thailand, Turkey, and the United Arab Emirates.
There was a temporary increase in administered prices during the lockdown phase due to social distancing measures related to passenger travel.
The CPI basket includes 11 components (such as “food, beverages and tobacco”) and 43 subcomponents (such as “non-alcoholic beverages”), while the PPI basket includes 9 components (such as “agriculture” and the largest share (see chart). Two notable exceptions are personal care products (e.g., soap and other hygiene products) and health-related items, which are likely to have seen increased demand due to concerns about the pandemic.
