Cigarette Production Is Falling Where Are Consumers Going?

Line value. Data for March 2020 through March 2024 is not available in this compilation.

The figures do fluctuate, but they stay within a range of roughly 9 to 12 percent. In every period recorded in the table, filtered kretek cigarettes hold second place, in both urban and rural areas.

A caveat needs stating, though. Data from March 2020 through March 2024 is not available in this compilation. So we can say cigarettes held second place at the data points that were recorded, but we cannot confirm that position never shifted in between.

There has also been a change in the rupiah values involved. The national Poverty Line rose from Rp440,538 per capita per month in September 2019 to Rp669,235 in March 2026. That means a similar-looking percentage at two different points in time actually represents rupiah values that are quite far apart.

The question then is not just whether cigarettes still hold second place in the Poverty Line basket, but whether that same pattern shows up in household spending.

BPS's Rokok dan Kemiskinan ("Cigarettes and Poverty") presentation includes a Susenas tabulation by expenditure quintile. All households are divided into five equal-sized groups, from the lowest-spending group to the highest.

The result differs depending on what you compare against. Measured against food spending alone, the proportion of cigarette consumption is relatively similar across all groups. But measured against total spending, the lowest group allocates roughly twice the proportion the highest group does.

This difference relates to Engel's law: as household income rises, the overall share of spending allocated to food tends to shrink.

So the pattern does hold at the household level. There are two important caveats, though. The data is from March 2015, making it eleven years old, and the calculation covers every household in each group, including households that don't smoke at all.

The drop in production alongside cigarettes' persistently high position in the Poverty Line looks contradictory if consumer choice is read as only two possibilities: keep smoking or quit.

Production data shows a more specific kind of change. The decline has not been even across all segments.

According to Antara (2025), Director General of Customs and Excise Askolani explained that the production decline through March 2025 occurred mainly in the segments with the highest excise rates:

• Tier I: down 10 percent

• Tier II: up 1.3 percent

• Tier III: up 7 percent

The shift is also visible in product type. Customs and Excise recorded a move from machine-rolled kretek cigarettes toward hand-rolled kretek cigarettes and cheaper products (CNBC Indonesia, 2025).

Production of labor-intensive hand-rolled cigarettes (SKT) rose 13.3 percent in 2024. At the same time, the number of cigarette factory workers rose to 301,113 — the highest in ten years — even as total production volume fell.

In the industry, this shift by consumers from more expensive products to cheaper ones is known as downtrading.

For that reason, the decline in national production volume is not enough on its own to conclude that consumers are leaving the cigarette category. Part of what we're seeing could be a shift from one price segment to another rather than exit from the category altogether.

At this point, the two sets of data need to be kept separate.

Production data covers the entire national market across all income levels. The Poverty Line, meanwhile, reflects the consumption pattern of the reference population BPS uses to construct that line. The populations being measured are not the same.

That means the segment-shifting visible in production data describes the market as a whole. Whether low-income households do the same thing when prices move cannot be answered from that data.

Askolani himself said the production decline is not entirely caused by downtrading, but is also related to purchasing power and health policy (Tempo, 2025).

There's another possibility worth factoring in: part of the official decline in production may not be a decline in consumption at all, but a shift toward products outside the excise system. The trouble is that estimates of how large that illegal consumption actually is vary widely.

The range spans 7 to 47 percent. A gap that wide doesn't automatically mean one figure is correct and the others are wrong, since their definitions, scope, and sampling methods all differ.

What's worth noting is how far the highest estimate sits from the others. That figure is also frequently used in arguments against excise-tax increases. Because of that, whichever number gets used should always be read alongside its methodology.

Price also needs to be distinguished from the excise rate. The government did not raise the Tobacco Product Excise rate in either 2025 or 2026. The decision for 2026 was announced by Finance Minister Purbaya Yudhi Sadewa in September 2025 — the fourth time in the past 15 years the rate has stayed flat, after 2014, 2019, and 2025 (Kumparan, 2025).

Even though the excise rate did not rise, the Retail Selling Price of nearly all tobacco products still rose by an average of around 10 percent starting 1 January 2025, through Minister of Finance Regulations 96/2024 and 97/2024 (Kompas, 2025).

So the excise rate and the retail price are two different instruments. They cannot be treated as the same thing when reading changes in consumption.

Culture is often the easiest explanation offered for why cigarette consumption in Indonesia stays high. Kretek has long been embedded in everyday life, so when consumption persists, the explanation tends to get pinned back on cultural factors.

Provincial BPS data for March 2026 shows a more varied picture.

That share ranges from 6.15 percent in rural Yogyakarta to 15.30 percent in urban Bengkulu — a gap of more than double.

In Bengkulu, the urban cigarette share even approaches rice's share, at 19.38 percent. This pattern does not appear in any other province in the compilation (BPS Bengkulu Province, 2026).

In other words, cigarettes' high standing at the national level doesn't mean their importance is the same in every region. There are substantial differences from one area to the next.

This variation also matters for re-examining the downtrading hypothesis. If a shift between price tiers really is happening among low-income households, the pattern isn't necessarily the same everywhere.

Aggregate expenditure data can show that a category is holding steady. What it can't yet answer is how consumers are sustaining it.

When prices rise, for instance, a consumer could buy fewer sticks, switch to a cheaper product, switch brands within the same price tier, change where they buy, or simply keep smoking as usual while cutting spending elsewhere.

All five possibilities could produce similar-looking aggregate expenditure figures. Without individual-level data, we can't tell which one is actually happening.

The same applies when we ask why consumers keep buying cigarettes at all. Ease of access, tiered price affordability, social attachment, and habit are all factors worth examining. Differences between provinces suggest the weight of each factor probably isn't the same everywhere. For now, though, all of this remains conjecture.

This is where the limits of the available data show up. Segment-shifting is already visible in national market data. Cigarettes' position in household spending also still holds. But the data linking the two isn't available yet: when cigarette prices change, what do low-income consumers actually do?

That question matters because a decline in production doesn't automatically mean a decline in the number of smokers. It could reflect a shift between products, a change in quantity consumed, a change in where people buy, or even a shift to an unrecorded market.

Without consumer-level data, we still can't tell which of these shifts is the largest.

References

Antara. (2025, 30 April). DJBC sebut downtrading memengaruhi penurunan produksi rokok. Antara News. https://www.antaranews.com/berita/4805997/djbc-sebut-downtrading-memengaruhi-penurunan-produksi-rokok

Badan Pusat Statistik. (2015). Rokok dan kemiskinan [Paparan]. Badan Pusat Statistik Republik Indonesia. https://komnaspt.or.id/wp-content/uploads/2020/12/Paparan_Rokok-dan-Kemiskinan_BPS-RI_2015-1.pdf

Badan Pusat Statistik. (2018, 16 Juli). Persentase penduduk miskin Maret 2018 turun menjadi 9,82 persen. Badan Pusat Statistik. https://www.bps.go.id/id/pressrelease/2018/07/16/1483/Persentase-Penduduk-Miskin-maret-2018-turun-menjadi-9-82-persen.html

Badan Pusat Statistik. (2026). Profil kemiskinan di Indonesia Maret 2026. Badan Pusat Statistik.

Badan Pusat Statistik Provinsi Bengkulu. (2026, 5 Agustus). Siaran pers BPS Provinsi Bengkulu 5 Agustus 2026. https://bengkulu.bps.go.id/id/news/2026/08/05/1012/siaran-pers-bps-provinsi-bengkulu-5-agustus-2026.html

CNBC Indonesia. (2025, 28 Desember). Fenomena rokok murah & pengusaha yang bisa kembali bernapas. CNBC Indonesia. https://www.cnbcindonesia.com/news/20251228194824-4-697739/fenomena-rokok-murah-pengusaha-yang-bisa-kembali-bernapas

Kompas. (2025, 27 September). Industri tembakau tertekan, pemerintah putuskan cukai rokok tak naik 2026. Kompas.id. https://www.kompas.id/artikel/industri-tembakau-tertekan-pemerintah-pilih-tidak-naikkan-cukai-2026

Kumparan. (2025, 6 Oktober). Siasat perangi rokok ilegal yang merajalela. Kumparan. https://kumparan.com/kumparannews/siasat-perangi-rokok-ilegal-yang-merajalela-25zd1tjNikJ

Suara.com. (2026, 26 Juni). Hitung-hitungan kerugian negara dari peredaran rokok ilegal. Suara.com. https://www.suara.com/bisnis/2026/06/26/192159/hitung-hitungan-kerugian-negara-dari-peredaran-rokok-ilegal

Tempo. (2025). Apa itu downtrading yang disebut jadi pemicu penurunan produksi rokok. Tempo.co. https://www.tempo.co/ekonomi/apa-itu-downtrading-yang-disebut-jadi-pemicu-penurunan-produksi-rokok--1334476

Tribunnews. (2026, 28 Juni). INDEF ungkap produksi rokok terus turun, rokok ilegal kian marak. Tribunnews.com. https://www.tribunnews.com/bisnis/7847652/indef-ungkap-produksi-rokok-terus-turun-rokok-ilegal-kian-marak