The tax increase represents an important progress on implementation of the Framework Convention on Tobacco Control (SDG-T 3.a) [36], specifically implementation of tax measures to reduce the demand for tobacco (Article 6). The tax hike in the specific component from COP$ 2,350 to COP$7,000 increases the retail price of a 20-stick pack in 98.4% from COP$7,028.6 to COP$ 13,950.3. Based on data on exchange rates from the Central Bank of Colombia and the Worldbank (annual average of the monthly exchange rate for 2019 of COP$ 3,281.3 per US dollar–US$ and PPP conversion factor for Colombia in 2019 of COP$ 1,317.1 per international $–INT$), the price change equals to an increase from US$ 2.14/INT$ 5.33 to US$ 4.25/INT$10.59.

Reduction in demand for cigarettes translates into lower exposure to tobacco, one of the main risk factors of Non-Communicable Diseases (NCDs). At the extensive margin, the tax increase is estimated to reduce the number of smokers from 4.51 to 3.45 million smokers. At the intensive margin, smokers who keep smoking after the tax increase also reduced their consumption, lowering their exposition to the risk factor from 2.01 to 1.78 cigarettes per day (median). Out of both margins, only the reduction in the number of smokers leads to gains in health; in other words, reduction in smoking intensity from non-quitters does not lead to gains in health as the health benefits of smoking reduction, specially if it is not substantial, are not significant [37]. The combination of the effect on smoking at both the extensive and intensive margins is shown in Figure 2. Both effects combined reduce the number of annual cigarettes smoked in Colombia from 332.3 to 215.5 MM of 20-stick packs.

The decrease in the number of smokers reduces the health risk on NCDs, decreasing the premature mortality caused by smoking from 2.3 to 1.8 MM deaths, representing a progress of 445.0 M averted deaths (SDG-T3.4): 233.2 M from heart disease, 99.6 M from stroke, 82.3 M from COPD and 30.0 M from lung cancer. As of July 15, 2021 the number of deaths due to Covid-19 in Colombia was 113,839; also, in 2019, Colombia had 244,355 non-fetal deaths; finally, the armed conflict between 1958 and 2012 in Colombia caused at least 220,000 deaths [38]. Therefore, the tobacco tax hike averts the equivalent of 3.91 times the total deaths from Covid-19, 1.82 times the number of annual non-fetal deaths in Colombia, and 2.02 times the deaths from the armed conflict. The averted deaths translate into a total gain of 7.9 million years of life. In addition to these gains, the tax also averts 15.9 M deaths from exposition to second-hand smoke (SHS) in the household (at home).

The averted morbidity and mortality from both smoking and second-hand smoking has direct effects on health systems as it translates in lower healthcare utilization and thus, to lower costs. The estimates suggest that the tax increase reduces healthcare costs by COP$2,157.4 MMM: COP$798,740 MM from heart disease, COP$692,868.8 MM from stroke, COP$343,721.3 MM from COPD and COP$322,095.1 MM from cancer. Healthcare for NCDs is expensive and, when a significant part of those costs are covered by Out-Of-Pocket (OOP) payments, they can lead to poverty. The characteristics of the health system in Colombia imply that most of the costs of healthcare caused by smoking are absorbed by the national risk pool, leaving only a small portion of those costs to be payed by directly by the person/household (out-of pocket expenditure). Regarding the latter financial burden, it is estimated that the tax increase leads to a reduction of COP$194.2 MMM in OOP expenditure in healthcare (SDG-I3.8.2), representing protection from financial risk (SDG-T3.8).

Households who are not poor can be pushed below the poverty line by high healthcare costs of NCDs treatment caused by smoking; thus, averted deaths from smoking translate into averted new cases of households falling into poverty. The estimates suggest that the tax hike averts 28.96 thousand people from falling into poverty. Along the same line, catastrophic expenditure on healthcare is defined as household’s out-of-pocket expenditure on healthcare higher than 10% of the household’s total income. The results suggest that the tobacco tax hike averts 337.6 thousand people from having catastrophic expenditure.

Increases in tobacco taxes are progressive both in the short-run and in the long-run; in the short-run because individuals with lower income are more sensitive to changes in prices, and therefore the gains of changing from paying a positive amount of tax to paying zero tax because of quitting concentrate in those populations; in the long-run, the gains in health both in morbidity and mortality exceed by far, at the aggregate level, the additional costs imposed on smokers who did not quit [20, 39]. Figure 3 shows the Concentration Curve of smokers before and after the tax increase. Since current smokers are the focus of the study, before the tax increase the distribution perfectly matches the 45 ^◦ line of equity. The dashed line shows that the tax increase benefits all segments of the population by reducing exposition to smoking all over the income distribution. However, the effect is not homogeneous; it can be seen that such benefit is higher in low-income individuals, showing that the distributional effect of the tax increase is progressive. This happens because the price elasticity is lower in low-income individuals, meaning that their quitting response to the tax increase is higher as compared to the middle or high income individuals, who also benefit from the policy but in a lower magnitude. In general, tobacco taxes act as a fiscal policy that contributes to greater equality (SDG-T10.4). When results are aggregated at the subnational level, the same mechanism leads to concentration of the gains of the tobacco tax increase in the poorest regions in the country. This subnational effect is particularly relevant for Colombia because subnational authorities are partially responsible for health insurance coverage of the population with the lowest income (subsidized regime), and regions with the highest needs are the ones with the lowest resources to invest in local public health actions to counteract and deal with the consequences of the tobacco epidemic. In addition, subnational authorities are in charge of tobacco tax administration.

With gender being one of the defining characteristics of individuals in the artificial society, the microsimulation model allows to address, at least partially, some of the drivers of disadvantages in terms of poverty and health for women levered by a reduction in tobacco consumption, and it is an attempt to bring an intersectional approach to explore interactions and outcomes within households that are influenced by tobacco consumption [40]. The next lines describe the most general aspects of that analysis. Regarding the distribution of risk factor by sex, the estimated number of smokers by sex is 2.8 million men and 1.66 million women; the tobacco tax reduces this exposition to 2.17 million male smokers and 1.27 female smokers. One of the mechanisms included in the model is second-hand smoking, which is relevant to identify the consequences of the power relation between smoker members of the household and other members (mostly women and children) [41]. Since tobacco taxes reduce smoking rates, they contribute to disease prevention due to decreased exposure in the household [42]. We calculate that effect based on previous evidence on prevalence of second-hand smoking at home for Colombia [43], and on epidemiological parameters for morbidity and mortality from second-hand smoke [44, 45]. The results suggest that the second-hand smoke averted mortality is one of the effects of the policy affecting gender equity; despite the higher number and prevalence of male smokers, most of the averted deaths from second-hand smoke (8.71 M out of 15.9 M) correspond to women. In addition, there are other mechanisms operating in the model. First, the reduction of the crowding-out effect of tobacco expenditure caused by the tobacco tax increase studied in several LMIC, including Chile, with similar cultural and socioeconomic conditions [46]. Second, the averted loss in the pool of income at the household level. Both mechanisms suggest that the tobacco tax policy promotes gender equity.

Human capital plays a central role on economic growth [47] as well as on development [48], with education and health [49] being “the two most important sources of human capital: knowledge capital and health capital” [50]. In addition to the loss in health capital represented by morbidity and mortality, smoking and NCDs lead to losses in knowledge capital via the years of education lost with premature death. For that reason, by averting deaths from smoking, tobacco taxes also contribute in avoiding losses in knowledge capital. The tax increase averts the loss of 10.1 years of education per smoker (≈ the number of years needed to complete middle school); adding up this effect for every quitter leads to a total of 3.37 MM of lost years of education averted by the policy, contributing to SDG-T4.4. This gain is a conservative estimate because, since the model is static, individuals do not increase their educational attainment over time; however, it is expected that younger cohorts increase their years of education over time and that structural changes in the educational system also lead to changes in population’s level of education.

Tobacco taxation is a source of tax revenues; therefore the tax increase opens fiscal space and strengthens domestic resource mobilization (SDG-T 17.1). The increase in the excise rate more than compensates for the previously mentioned drop in the number of cigarettes in the Colombian market, leading to an increase in tax revenue from COP$ 797.1 MMM to COP$ 1,359.3 MMM. There is an additional positive, although unfortunate, effect on tax revenues. Because tobacco taxes are low in Colombia, some cigarettes are introduced legally to the country, taxes are paid, and afterwards they are smuggled into other countries in the region. This market tactic has been reported in Ukraine [51] and Belgium [52].

When such mechanism is taken into account, the initial tax revenue from tobacco taxes is estimated at COP$1,089.7 MMM, and the tax hike increases it to COP$ 2,158.3 MMM. This is explained by the cigarette surplus entering Colombia: before the tax increase, it is estimated that 425.2 MM 20-stick packs of cigarettes enter legally into the country, and after the tax such quantity is expected to decrease to 308.3 MM packs. However, it is important to notice that the excess tax revenue is artificial in the sense that it is expected to significantly be reduced once Colombia catches up with the region in the level of tobacco taxes and effective policies to counter illicit cigarette trade [53] are accurately implemented [54], reducing corruption and organized crime, and strengthening rule of law (SDG16).

A final point of public finance is worth mentioning. The tax increase transfers resources from the tobacco industry to the government, and such transfer is socially desirable as the tobacco industry imposes health and environmental externalities for which the government, and the society in general, has to pay. Despite the tax increase, the results show that the profits from the tobacco industry are expected to increase from COP$ 236.8 MMM to COP$ 344.7 MMM. This unintended increase is caused by the tobacco industry’s typical response to tobacco taxes, overshifting the tax to the consumer to maintain and increase their profits [55]. The implication of this result is that even after an important increase in tobacco taxes as the one simulated in the model, there is still space for future further increases in tobacco taxes.