SciELO - Scientific Electronic Library Online

vol.59 issue1Human papillomavirus vaccine efficacy in the prevention of anogenital warts: systematic review and meta-analysisExperiences of street harassment and associations with perceptions of social cohesion among women in Mexico City author indexsubject indexsearch form
Home Pagealphabetic serial listing  

Services on Demand




Related links

  • Have no similar articlesSimilars in SciELO


Salud Pública de México

Print version ISSN 0036-3634

Salud pública Méx vol.59 n.1 Cuernavaca Jan./Feb. 2017 

Artículos originales

Analysis of cigarette demand in Argentina: the impact of price changes on consumption and government revenues

Análisis de la demanda de cigarrillos en Argentina: impacto de variaciones de precios en el consumo y recaudación

Germán Rodríguez-Iglesias, MSc1  * 

Verónica Schoj, MD1 

Frank Chaloupka, PhD2 

Beatriz Champagne, PhD3 

Martín González-Rozada, PhD4 

1 Fundación InterAmericana del Corazón. Argentina.

2 University of Illinois. Chicago, USA.

3InterAmerican Heart Foundation. Dallas, USA.

4Universidad Torcuato Di Tella. Buenos Aires, Argentina.



To estimate cigarette demand and to simulate a tax policy targeted to reduce tobacco consumption.

Materials and methods:

Demand was estimated using a vector error correction model. Simulation exercises present the impact of a tax increase on consumption and revenues.


Changes in real income and the real price of cigarettes affect the demand for cigarettes in Argentina. The long term price elasticity is 0.279 (a 10% increase in real prices reduces cigarette consumption by 2.79% per quarter) and the long term income elasticity is 0.411 (a 10% increase in real income raises consumption by 4.11% per quarter). Even in a conservative scenario, simulations show that increasing the price of cigarettes by 100% using excise taxes would maximize revenues and reduce cigarette consumption.


There is sufficient room to increase taxes, reducing cigarette consumption, while still increasing tax revenues.

Keywords: tobacco; taxes; elasticity; health policy; Argentina



Estimar la demanda de cigarrillos y simular una política fiscal dirigida a reducir el consumo de tabaco.

Material y métodos:

Se estima la demanda mediante el modelo de corrección de errores. Se simula el impacto del incremento de los impuestos en el consumo y la recaudación.


Las variaciones en ingreso y precio real de los cigarrillos afectan la demanda. La elasticidad precio de la demanda de largo plazo es de 0.279 (10% de aumento en los precios reales reduce el consumo de cigarrillos en 2.79% en un trimestre) y la elasticidad ingreso de largo plazo es 0.411 (10% de aumento en el ingreso real aumenta el consumo en 4.11% en un trimestre). Aun en un escenario conservador, un incremento del precio de los cigarrillos de 100% vía impuestos maximizaría la recaudación y reduciría el consumo de cigarrillos.


Es posible incrementar los impuestos reduciendo el consumo de cigarrillos e incrementando la recaudación.

Palabras clave: tabaco; impuestos; elasticidad; política de salud; Argentina

A growing literature shows that the overall demand for tobacco products is significantly affected by c hanges in tobacco prices.1,2,3,4 Empirical evidence for high-income countries shows a demand price elasticity in the range of -0.25 to -0.50.5 In the last years there has been evidence from low and middle-income countries, showing elasticity to be at least as responsive, and often more responsive, to price than it is in high-income countries.6,7,8,9,10,11

According to the National Risk Factor Survey 2013, 25% of the Argentinean population smokes,12 and cigarette smoking contributes to over 40 000 deaths per year, with an estimated 13.5% of all deaths attributable to tobacco.13 Argentina did not ratify the World Health Organization’s (WHO) Framework Convention on Tobacco Control (FCTC), as a result of tobacco industry influence, mostly through tobacco growers.14

In June 2011, a comprehensive National Tobacco Control Law (Act 26687) was passed. However it did not include mention of cigarette taxes or pricing.

Argentina has one of the cheapest cigarettes in the world. Cigarettes have become twice as affordable in 2014 that in 2004.15

If the government increases tax rates, prices will increase, and consumption will decrease. Price elasticity allows us to quantify changes in consumption as a result of price variations.

To obtain price elasticity we estimated a cigarette demand function. Using a similar methodology but different data sets, González-Rozada7 and Martinez, Mejia and Perez-Stable16 have found Argentina’s demand price elasticity ranging from -0.26 to -0.31. The present study updates González-Rozada’s analysis and develops simulations to analyze the impact of price changes on consumption and government revenues. This study introduces a cigarette tax structure analysis in the simulation of the impact of a tax increase. Understanding of the tax structure is necessary in the definition of a cigarette tax increase. Findings in elasticity estimations, cigarette tax structure, and simulation estimations can be used in tobacco control policies, as demonstrated in other countries.

Materials and methods

This paper focuses on cigarette consumption because 98.2% of tobacco consumers in Argentina smoke cigarettes.17 We used monthly data on cigarette sales and average weighted prices published by the Secretary of Agriculture (SAGPyA) from January 1996 to February 2004.18 From March 2004, these data were published by the Ministry of Agriculture (MINAGRI).19 The average nominal income of the private sector (computed by the Ministry of Economics, MECON)20 was used as a proxy for the nominal income of the population. Variables were expressed in real terms, dividing nominal data by the consumer price index (CPI) computed by the National Statistics Agency (Instituto Nacional de Estadísticas y Censos, INDEC)21 for the period of January 1996 to December 2006 linked with the estimated monthly inflation rate computed by independent private consultants from January 2007 onwards.15

To establish the functional form of the demand for cigarettes, it is necessary to determine the statistical properties of the individual series. The order of integration of each series was determined by the augmented Dickey-Fuller (ADF) test.22

As seen in table I, consumption and price series do not separately reject the hypothesis of a unit root. Consequently, the price and consumption variables individually have a stochastic trend.

Table I Augmented Dickey-Fuller and Perron unit root tests. Argentina, 2015 

Note: The number of lags was selected according to the Schwarz Information Criterion. The probability of rejection was calculated using the critical values for the test tabulated in. log denotes the natural logarithm

Source: Authors’ estimation using national data.19,21,32 Real price base for January 1996=100

We used the Perron formal procedure to contrast unit root in the presence of a structural break.23,24 As Perron test statistic (table I) is lower than the critical value, we cannot reject the null hypothesis of unit root in the presence of a structural break, then real income is a non-stationary series.

In sum, cigarettes consumption, real average retail price and real income individually follow a random walk. The only possibility to estimate a demand function for cigarettes was if the three variables were cointegrated.

We tested cointegration using two procedures; Engle-Granger25 and Johansen tests.26 Table II presents the Engle-Granger procedure, showing that residuals are stationary. To confirm this result, we applied the Johansen cointegration test 26 that rejects the null hypothesis of no cointegration vector and cannot reject null hypothesis of one cointegration vector.

Table II.Engle-Granger and Johansen Cointegration tests. Argentina, 2015 

Note: The number of lags was selected according to the Schwarz Information Criterion. The Engle-Granger Test probability of rejection was calculated using the critical values for the test tabulated in.33 The Johansen Cointegration test probability of rejection was calculated using the34 critical values

Source: Authors’ estimation

Empirical evidence suggests cointegration among the variables.

Cigarette demand function

Cointegration implies that the cigarette demand function can be specified with an error correction model (ECM) that takes into account both the short-run dynamics and the long-run relationship between the variables.

The long-run equilibrium function is:


where ct is the natural logarithm of cigarettes consumption, pt is the natural logarithm of real average retail price of cigarettes, and yt is the natural logarithm of real income of the population at time t.λ1 and λ2 are parameters and ut is a stationary error with zero mean. ki includes coefficients on the control variables, λ1 can be interpreted as the long-run price elasticity of demand, while λ2 is the long-run income elasticity of demand.

In the short-run, the variables may not be in the steady state (long-run equilibrium). Outside the steady state, the lag structure of the model cannot be specified with certainty. Therefore, short term is part of a general model with r lags (months),


where β0 is the short-run price elasticity and λ0 is the short-run income elasticity of consumption. εt is a stationary error term. The term in the curly brackets represents the long-run relation (1), while all variables in first differences measure the short-run dynamics of the model. The long-run equilibrium is reached when the braces term, called error-correction term is equal to zero.


Elasticity estimations

Table III shows the results of the long-run and short-run estimations. The long-run cigarette price elasticity is -0.28, while the long-run cigarette income elasticity is 0.41. Thus, 10% increase in real prices will reduce long-run total cigarette consumption by 2.79%, and 10% increase in real income will raise long-run consumption by 4.11%. The long-run is defined by estimating coefficient r in equation (2). The Johansen test states that the value of r-1 is equal to 2. This implies that the long-run is three months (a quarter) (table III).

Table III Long-Run and Short-Run Elasticity Estimates. Argentina, 2015 

* p < 0.05

‡ p < 0.01

Notes: 13th Month Pay – June Half and 13th Month Pay – December Half are dummy variables that take on the value 1 for June and December, respectively, to capture the effect of the split 13th month pay and any seasonal effects

D captures March 2002 structural break

Short-run lags are determined by the Johansen test. D is the finite difference operator

Log denotes the natural logarithm

Dependent Variable: Log (Total Cigarette Consumption)

Estimation Method: Least Squares

Sample: 1996:01 2014:12

Source: Authors’ estimation using national data.21,32,35 Real price base for January 1996=100

The short-run price elasticity of demand - β0 - and the short-run income elasticity - γ 0 ,- are not statistically significant. The coefficient associated with the error correction term -( α-1)- is negative and statistically significant, which means that the model is stable.

Cigarette tax structure

In December 2014, the average price of a pack of cigarettes was AR$17.42, and the federal tax percentage collected on each pack was 69.25%. Taxes levied on tobacco products are, in general, ad valorem. There is a consumption tax (Value Added Tax (VAT) and three excises taxes: Additional Emergency Tax (AET); Special Tobacco Fund (FET) and Internal tax (IT). The only tax that is not ad valorem is a small component of the FET that is specific.

In nominal terms, IT account for 60%, IAE for 7%, VAT for 21% and FET for 8.35%. However, the taxes are applied over different tax bases differing nominal rates from the share of the price that each tax accounts. The total tax burden accounts for 68.18% of the average price of a pack of cigarettes in Argentina (AR$11.88 / AR$17.42 x 100 = 68.18%).Other tobacco products (cigars, roll-your-own, etc.) have significantly lower taxes (VAT 21% and IT from16 to 20%) table IV shows the tobacco tax structure and the percentage of tax as a share of the price.

Table IV Components of price of a pack of cigarettes. December 2014 

AET: Additional Emergency Tax

FET: Special Tobacco Fund

VAT: Value Added Tax

Source: Constructed by the authors using information from the tax laws of Argentina

Simulation of the impact of a cigarette tax increase

Considering December 2014 values [annual cigarette sales 2.1 billion packs; average retail price AR$17.42 per pack; total tax on cigarettes (excise duty plus VAT) AR$11.88; cigarette excise tax AR$10.75 per pack; federal cigarette tax revenue AR$ 27.9 billion (US$ 2 billion); exchange rate: US$1=AR$$13.73; population 42 202 935 (more than 15 years old 31 452 302); 27 per capita consumption 48.53 packs/year and per capita consumption (more than 15 years old) 64.97 packs/year. The simulation exercise was conducted for three different cigarette price elasticities. The “neutral scenario” is based on the -0.28 estimated value; the “high revenues scenario” considered a -0.18 price elasticity obtained from the -0.28 estimated value plus twice its standard error (0.047); the “low revenues scenario” is a -0.37, obtained from the -0.28 estimated value less twice its standard error.

We assume that the tax increase passes through totally to the retail price. This is probably a conservative scenario for public health impact considering that the tobacco industry will likely reduce total profits as tobacco consumption falls.

Figure 1 shows the average annual per capita consumption of cigarette packs under the tax rate increase for the three scenarios defined above. The horizontal axis shows the increase in government collection per pack. The vertical axis shows the average annual per capita consumption for the population over age 15. An increase in the tax rate is interpreted as an increase in the AR$11.88 price, which is the amount that the government collects per pack. In the neutral scenario, a 10% increase in prices, from AR $17.42 to AR$19.16 per pack (resulting from a 14.7% increment in government revenue per pack), causes a 2.8% drop in average annual per capita consumption (from 64.97 to 63.02 packs per year).

Source: Authors’ estimations.

Figure 1 Relationship between Increases on government collection per pack, Cigarette Consumption Change in tax revenue. 

Figure 1 also shows the Laffer curves, which represents graphically the relationship between changes in federal tax revenue (vertical axis) and the increase in government collection per pack. A 10% increase in the amount that the government collects per pack represents a retail price increase from AR$17.42 to AR$18.61 per pack. Thus, to obtain a 10% increase in cigarette price, government revenue per pack should rise by 14.7%. This 10% tax increase results in AR$2.21 billion increase in tobacco revenues (a 160 million US dollars increase).

The “Laffer point,” i.e., the peak point at which cigarette tax revenue starts to fall as the tax rate rises, is reached in the low-revenue scenario for a 146% increment on government collection per pack, or at AR$34.77 price per pack, that is a 100% real price increment per pack. In the other two scenarios, the Laffer point is only reached after a 212% increase (more than a 145% increase in real prices for the neutral scenario and more than 237% price increase in the favorable scenario).


This study updates the results obtained by González-Rozada in 2006 7 and its estimates are consistent with those by previous authors in Argentina. 16 The results of this study had direct bearing on a significant tax increase on cigarettes that was approved in May 2016.

Argentina’s elasticities are closer to those observed in high-income countries likely below 5% like other Latin American countries (LAC) such as Chile, Brazil, Uruguay and Mexico.28 This relatively low elasticity could be attributable to the high affordability of cigarettes in these countries.29

This study confirms that the rise in taxes results in a reduction of cigarette consumption as well as an increase in government revenues even in the most conservative scenario analysis. Tax increase simulations support international evidence on the effectiveness of increasing taxes as an tobacco control measure.3 In absolute value, income elasticity is about 47.1% greater than the price of elasticity of demand. This means that to offset 10% increase in income and bring about a fall in consumption, the real price should be raised by 14.71% in a quarter. Policymakers must ensure periodic tax increases to progressively reduce cigarette affordability. Combining price and income elasticities estimations, policymakers, particularly in rapidly growing economies, may obtain a tax increase benchmark indicator that determines how much tax should increase so that cigarette affordability at least does not increase.30,31

The present study has some limitations. Our analysis considers average price and income, and does not allow us to analyze inequalities neither other tobacco products elasticities of certain groups.

This study contributed significantly to a change in fiscal policy in Argentina. In May 2016, the Argentinean federal government enacted a decree that increased the Internal Tax by 15 percentage points (from 60 to 75%) increasing the cigarettes price by 50%. According to our analysis, the total cigarette consumption would be reduced by 15%. This tax increase is an important success for public health, even though the increase in Internal Tax was only applied to cigarettes and it served to strengthen the ad valorem tax structure.

This new tax policy increases opportunities for substitution for cheaper brands and for other tobacco products like “roll-your-own”. Tax increase law should include a simpler tax structure via a specific tax and a harmonization of the taxes for all tobacco products to avoid substitution. A single increase is not sufficient and it is essential to implement periodic adjustments to offset the effects of increasing affordability in an inflationary context like Argentina’s.

Our study shows that demand analysis is necessary but not sufficient to promote successful tax policies. It is also essential to analyze the tax structure to promote effective and sustainable policies.


The authors would like to thank translator Melisa Keller for her help with the paper writing.


1. Farrelly MC, Nonnemaker JM, Watson KA. The consequences of high cigarette excise taxes for low-income smokers. PloS one 2012;7(9):e43838. [ Links ]

2. Bader P, Boisclair D, Ferrence R. Effects of tobacco taxation and pricing on smoking behavior in high risk populations: A knowledge synthesis. Int J Environ Res Public Health 2011;8(11):4118-4139. [ Links ]

3. Chaloupka FJ, Yurekli A, Fong GT. Tobacco taxes as a tobacco control strategy. Tob Control 2012;21(2):172-180. [ Links ]

4. IARC. Effectiveness of Tax and Price Policies for Tobacco Control. Ginebra: World Health Organization, 2012. [ Links ]

5. Jha P, Chaloupka FJ. The economics of global tobacco control. BMJ 2000;321(7257):358-361. [ Links ]

6. Debrott D. Economía del control del tabaco en los países del Mercosur y Estados asociados. Chile/Washington: Organización Panamericana de la Salud, 2006. [ Links ]

7. González-Rozada M. Argentina: economía del control de tabaco en los países del Mercosur y Estados asociados. Washington: Organización Panamericana de la Salud, 2006. [ Links ]

8. Ramos A, Curti D. Economía del control del tabaco en los países del Mercosur y Estados asociados . Uruguay/Washington: Organización Panamericana de la Salud , 2006. [ Links ]

9. Alcaraz VO. Economía del control del tabaco en los países del Mercosur y Estados asociados : Bolivia/Washington: Organización Panamericana de la Salud , 2006. [ Links ]

10. Ramos A, Curti D. Uruguay: Economía del control del tabaco en los paises del Mercosur y Estados asociados. Washington: Pan American Health Organization, 2005. [ Links ]

11. Nikaj S, Chaloupka FJ. The effect of prices on cigarette use among youths in the Global Youth Tobacco Survey. Nicotine Tob Res 2014;16(Suppl 1):S16-S23. [ Links ]

12. Ministerio de Salud de la Nación. Tercer Encuesta Nacional de Factores de Riesgo para Enfermedades No Transmisibles: Presentación de los principales resultados. Buenos Aires: Ministerio de Salud de la Nación, 2014. [ Links ]

13. Pichon-Riviere A, Alcaraz A, Bardach A, Augustovski F, Caporale J, Caccavo F. Carga de enfermedad atribuible al tabaquismo en Argentina. Buenos Aires: Instituto de Efectividad Clínica y Sanitaria, 2013. Documento Técnico IECS N° 7. [ Links ]

14. Mejia R, Schoj V, Barnoya J, Flores ML, Perez-Stable EJ. Tobacco Industry Strategies to Obstruct the FCTC in Argentina. CVD prev control 2008;3(4):173-179. [ Links ]

15. Rodríguez-Iglesias G, González-Rozada M, Champagne B, Schoj V. Real price and affordability as challenges for effective tobacco control policies: an analysis for Argentina. Rev Panam Salud Publica 2015;37(2):98-103. [ Links ]

16. Martinez E, Mejia R, Perez-Stable EJ. An empirical analysis of cigarette demand in Argentina. Tob Control 2015;24(1):89-93. [ Links ]

17. Konfino J, Linetzky B, Goldberg L, Ferrante D, Ondarsuhu D, Ondarsuhu D. Global Adult Tobacco Survey: Argentina 2012. Buenos Aires: Ministerio de Salud de la Nación , 2013. [ Links ]

18. Secretaría de Agricultura G, Pesca y Alimentos. Paquetes de Cigarrillos Vendidos por Rango de Precio. Buenos Aires: SAGPyA, 2004. [ Links ]

19. Paquetes de Cigarrillos Vendidos por Rango de Precio [database on the Internet]. MINAGRI, 2014 [accessed January 15, 2014]. Available from: [ Links ]

20. Indice de Salarios, valores al último día de cada mes [database on the Internet]. MECON, 2014 [accessed January 15, 2014]. Available in: ]

21. Indice de precios al consumidor [database on the Internet]. INDEC, 2014 [accessed January 15, 2014]. Available in: www.indec.mecon.arLinks ]

22. Dickey DA, Fuller WA. Distribution of the estimators for autoregressive time series with a unit root. J Am Stat Assoc 1979;74(366a):427-431. [ Links ]

23. Perron P. The great crash, the oil price shock, and the unit root hypothesis. Econometrica 1989;57(6):1361-1401. [ Links ]

24. Perron P. Erratum [The great crash, the oil price shock and the unit root hypothesis]. Econometrica 1993;61(1):248-249. [ Links ]

25. Engle RF, Granger CW. Co-integration and error correction: representation, estimation, and testing. Econometrica 1987;55(2):251-276. [ Links ]

26. Johansen S. Estimation and hypothesis testing of cointegration vectors in Gaussian vector autoregressive models. Econometrica 1991;59(6):1551-1580. [ Links ]

27. Indec. Estimaciones y proyecciones de población 2010-2040: total del país. Buenos Aires: Instituto Nacional de Estadística y Censos; 2013 [accessed January 15, 2015]. Available in: ]

28. Guindon GE, Paraje GR, Chaloupka FJ. The impact of prices and taxes on the use of tobacco products in Latin America and the Caribbean. Am J Public Health 2015;105(3):e9-e19. [ Links ]

29. World Health Organization. WHO technical manual on tobacco tax administration. Geneva: World Health Organization, 2010. [ Links ]

30. Blecher E. Targeting the affordability of cigarettes: a new benchmark for taxation policy in low-income and-middle-income countries. Tob Control 2010;19(4):325-330. [ Links ]

31. Eriksen M, Mackay J, Schluger N, Islami F, Drope J. The tobacco atlas. Atlanta: American Cancer Society, 2015. [ Links ]

32. Congreso Nacional. Índices de Precios al Consumidor difundidos por el Congreso Nacional, 2014 [accessed July 3, 2015]. Available in: ]

33. MacKinnon JG. Critical Values for Cointegration Tests. In: Engle RF, Granger CWJ, ed. Long run Economic Relationships: Readings in Cointegration. Oxford: Oxford University Press, 1991. [ Links ]

34. MacKinnon JG, Haug AA, Michelis L. Numerical distribution functions of likelihood ratio tests for cointegration. Journal of applied econometrics 1999;14(5):563-577. 10.1002/(SICI)1099-1255(199909/10)14:5<563::AID-JAE530>3.0.CO;2-R [ Links ]

35. Ministerio de Trabajo, Empleo y Seguridad Social de la Nación. Boletin de remuneraciones de los trabajadores registrados [database on the Internet] 2014 [accessed January 15, 2015]. Available in: ]

Received: March 16, 2016; Accepted: September 15, 2016

*Corresponding author: MSc Germán Rodríguez-Iglesias. Fundación InterAmericana del Corazón, Argentina. Postal: Arévalo 2364 “1°A”, Buenos Aires, C1425FZF, Argentina. E-mail:

Declaration of conflict of interests. The authors declare that they have no conflict of interests.

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License