📈 economics

Fuel Demand, Spatial Competition, and Local Market Structure: Evidence from Gasohol Retailing in Lima and Callao

This study utilizes district-level data from Lima and Callao to estimate that the price elasticity of demand for 90- and 95-octane gasohol ranges from -1.50 to -3.17, indicating a significant negative response of fuel demand to price changes despite the exceptional economic context of the pandemic and tax exemptions.

Original authors: David Ricardo Gonzales Peña, Jorge Ricardo Gonzales Peña

Published 2026-09-12

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: David Ricardo Gonzales Peña, Jorge Ricardo Gonzales Peña

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the bustling streets of Lima and Callao, Peru, the daily rhythm of life depends heavily on the fuel that powers cars, taxis, and motorcycles. For decades, policymakers have tried to balance the need for affordable transportation with the urgent goal of reducing air pollution. One tool they use is a tax on fuel, designed to make polluting options more expensive and encourage people to switch to cleaner alternatives. However, a fundamental question has long lingered without a clear answer: how much will people actually change their driving habits if the price goes up? In economics, this sensitivity is known as price elasticity. If a small price hike causes drivers to stop filling up their tanks, demand is considered "elastic," and the tax works well to reduce pollution. If drivers keep buying the same amount regardless of the cost, demand is "inelastic," and the tax fails to change behavior. Understanding this relationship is crucial for governments trying to design effective environmental policies without causing undue hardship for their citizens.

A recent study by researchers David Ricardo Gonzales Peña and Jorge Ricardo Gonzales Peña dives into this question by examining the local fuel market in Metropolitan Lima and Callao. The team focused on two specific types of fuel, known as gasohol, which is a blend of gasoline and alcohol. They looked at the 90-octane and 95-octane varieties, which are the most common fuels used by private vehicles in the region. The researchers gathered a massive amount of data from the Supervisory Agency for Investment in Energy and Mining, covering every month from March 2020 to December 2022. This period was unique and challenging, as it began right at the start of the global pandemic, a time when lockdowns and mobility restrictions drastically altered how people moved around the city. The study also had to account for various government interventions, such as a fund that stabilized fuel prices and temporary tax exemptions, which made the market behave differently than usual.

To find the answer, the researchers compared fuel prices with the actual amount of fuel sold in each district of the city. They used a statistical approach that allowed them to isolate the effect of price changes from other factors, such as the general economic situation or the specific characteristics of each neighborhood. Their analysis revealed a clear and strong pattern: when the real price of fuel went up, the amount people bought went down. This confirms that drivers in Lima are sensitive to price changes. Specifically, the study found that for every one percent increase in the price of 90-octane gasohol, the quantity demanded dropped by about one and a half percent. For the higher-quality 95-octane gasohol, the reaction was even stronger; a one percent price increase led to a drop in demand of more than two percent.

The researchers noticed that these effects became even more pronounced when they looked only at the years 2021 and 2022, after the initial shock of the pandemic had passed and people began returning to their normal routines. In this later period, the demand for 90-octane fuel became nearly two and a half times more sensitive to price, while the demand for 95-octane fuel jumped to over three times as sensitive. This suggests that once the immediate chaos of the pandemic subsided, drivers had more freedom to adjust their behavior, perhaps by combining trips, switching to public transport, or choosing cheaper fuel options when prices rose. The fact that the more expensive 95-octane fuel showed a higher sensitivity makes sense, as it is often used by wealthier drivers or in newer cars who have more flexibility to change their habits compared to those who rely on older vehicles or have fewer alternatives.

The study also explored whether the number of gas stations in a neighborhood influenced these price reactions, testing if more competition between stations changed how prices were passed on to consumers. While the researchers tried to use advanced methods to isolate the pure cause-and-effect relationship between fuel costs and sales, the complex mix of pandemic restrictions and government price controls made it difficult to draw a single, definitive conclusion from those specific tests. Instead, they relied on the clear patterns observed in the main data. The results indicate that while fuel demand in Lima is responsive to price, the exact magnitude of that response depends heavily on the context. The findings serve as a vital reality check for policymakers, showing that taxes can influence behavior, but the effect is not a simple, fixed number. It varies based on the type of fuel, the economic environment, and the specific circumstances of the time, reminding us that the relationship between a price tag and a consumer's decision is a dynamic and complex dance of real-world choices.

Technical Summary: Fuel Demand, Spatial Competition, and Local Market Structure

Problem Statement and Context
This study addresses a gap in the empirical evaluation of Peru's fuel policy, specifically the Selective Consumption Tax (ISC) introduced in 2006 via Law No. 28694. The law assigned higher taxes to more polluting fuels to encourage environmental objectives, yet its implementation was not accompanied by explicit ex ante estimates of demand elasticity. Consequently, the expected impact of the policy on local market quantities remained unknown.

The research focuses on the market for vehicular gasohol (a blend of 7.8% fuel alcohol and 92.2% gasoline) in Metropolitan Lima and Callao (LMC). This region is critical as it concentrates 65.9% of the national vehicle fleet and accounts for 87% of carbon monoxide emissions from mobile sources. The analysis covers the period from March 2020 to December 2022, a timeframe characterized by exceptional conditions: the COVID-19 pandemic, mobility restrictions, the operation of the Fuel Price Stabilization Fund (FEPC), and a temporary exemption of the ISC. The study ends in December 2022 due to a 2023 regulatory reform that reclassified fuel octane levels, creating a structural break in the data series.

Methodology
The authors utilize monthly district-level data on prices and quantities sold, sourced from the Supervisory Agency for Investment in Energy and Mining (Osinergmin). The dataset covers 803 service stations across LMC districts.

Main Specification: The primary empirical strategy employs a log-log demand specification with district and month fixed effects:
$\ln Q_{ipt} = \beta \ln P_{ipt} + \alpha_i + \gamma_t + \varepsilon_{ipt}$
Where $Q$ is the quantity of gasohol sold, $P$ is the real price (deflated by the vehicular fuel CPI), $\alpha_i$ represents district fixed effects (controlling for urban density, infrastructure, and socioeconomic composition), and $\gamma_t$ represents month fixed effects (absorbing aggregate shocks like pandemic restrictions and macroeconomic variations). Standard errors are clustered at the district level.
Instrumental Variable (IV) Strategy: To address potential price endogeneity, the authors construct a shift-share instrument inspired by the Bartik logic. This instrument interacts:
1. Exogenous Shocks: Lagged logarithmic variations in Osinergmin reference prices for 90- and 95-octane gasohol.
2. Local Exposure: The initial consumption share of each district and the level of spatial competition (measured by the number of service stations within 455m, 800m, and 1,000m radii).
  The authors explicitly note that the validity of this instrument relies on the assumption that spatial competition affects demand only through price transmission, acknowledging that spatial competition may be correlated with unobserved socioeconomic factors.
Robustness Checks: The study includes specifications with aggregate controls (GDP, vehicle flow, stringency index) and interactions between district income levels and time blocks to test for heterogeneous recovery trajectories across socioeconomic groups.

Key Results
The analysis yields the following estimates for the price elasticity of demand:

Full Sample (2020–2022):
- 90-octane gasohol: -1.50
- 95-octane gasohol: -2.20
Restricted Sample (2021–2022): Excluding the initial pandemic shock months, the magnitude of elasticity increases:
- 90-octane gasohol: -2.46
- 95-octane gasohol: -3.17

The results indicate a statistically significant negative relationship between real prices and quantity demanded. The higher elasticity for 95-octane gasohol suggests it is more sensitive to price changes, potentially reflecting differences in vehicle types (e.g., higher-end cars) or consumer income levels.

Regarding the instrumental variable exercises, the authors report that while the first-stage diagnostics show relevance, the IV estimates do not consistently produce a negative elasticity interpretable within the standard economic model. Consequently, the authors treat the IV results as methodological diagnostics rather than definitive causal estimates, relying primarily on the fixed-effects specifications.

Significance and Contributions
The paper makes two primary contributions to the literature:

Policy Evaluation: It provides the first specific ex post analysis of demand elasticity for the Peruvian gasohol market, offering empirical evidence to evaluate a policy (Law No. 28694) that was implemented without quantitative estimates of its expected consumption impact.
Methodological Application: It proposes and tests an instrumental variable application to isolate exogenous price variation in the Peruvian fuel market. The study highlights the challenges of identifying causal effects in this sector due to the interplay of pandemic shocks, price stabilization funds, and tax exemptions.

Limitations and Cautions
The authors emphasize that the estimated elasticities should be interpreted with caution due to the "exceptional context" of the sample period. The results are conditional on the specific institutional framework (FEPC, ISC exemptions) and the unique mobility patterns of the pandemic era. Furthermore, the study is limited by the use of district-level aggregates rather than transaction-level data, and the 2023 fuel classification reform prevents the direct extension of these specific product definitions beyond the study period. The authors conclude that while gasohol demand responds to price, the magnitude of these responses in normal market conditions may differ from the estimates provided.

More like this