Impact of cigarette tax increase on health and financing outcomes in four Indian states

Background: In India, about one million deaths occur every year due to smoking. Tobacco taxation is the most effective intervention in reducing smoking. In this paper, we examine the impact of a one-time large cigarette price increase, through an increase in excise tax, on health and financing outcomes in four Indian states. Methods: We used extended cost-effectiveness analysis to estimate, across income quintiles, the life-years gained, treatment cost averted, number of men avoiding catastrophic health expenditures and extreme poverty, additional tax revenue collected, and savings to the Ayushman Bharat Pradhan Mantri Jan Arogya Yojana (AB-PMJAY) with a cigarette price increase to Indian Rupees (INR) 10 plus 10% ad valorem in four Indian states. Results: With the price increase, about 1.5 million men would quit smoking across the four states, with the bottom income group having 7.4 times as many quitters as the top income group (485,725 vs 65,762). As a result of quitting, about 665,000 deaths would be averted. This would yield about 11.9 million life-years, with the bottom income group gaining 7.3 times more than the top income group. Of the INR 1,729 crore in treatment cost averted, the bottom income group would avert 7.4 times more than the top income group. About 454,000 men would avoid catastrophic health expenditures and 75,000 men would avoid falling into extreme poverty. The treatment cost and impoverishment averted would save about INR 672 crore in AB-PMJAY. The tax increase would in turn, generate an additional tax revenue of about INR 4,385 crore. In contrast to the distribution of health benefits, the extra revenue generated from the top income group would be about 3.1 times that from the bottom income group. Conclusions: Cigarette tax increase can provide significant health and economic gains and is a pro-poor policy for India.


Introduction
Tobacco use is the leading cause of preventable premature mortality globally 1 . In India, about one million deaths occur every year due to tobacco smoking 2 . Cessation substantially reduces the risk of dying; in particular cessation before age 40 reduces the risk of death associated with continued smoking by about 90% 3 . Yet, cessation remains uncommon in India. According to the Global Adult Tobacco Survey (GATS), in 2009-10, 5.7% of adults aged 15 years and above (10.3% males, 0.8% females) smoke cigarettes in India 4 ; this only decreased to 4.0% (7.3% males, 0.6% females) in 2016-17 5 . Tobacco use is also more prevalent among the poor, who in turn, fall into greater poverty, as poor families spend a larger proportion of their income on tobacco and are at a much higher risk of falling ill and dying from smoking-attributable diseases, thus imposing additional costs to the family 6 .
The most effective intervention to reduce tobacco use is a large increase in the federal excise tax on cigarettes that enables minimal downward substitutions to shorter, cheaper cigarettes 1,7-9 . Evidence from several low-and middle-income countries (LMICs) has shown that the impact of tobacco taxation would disproportionately favour populations on lower income 10-15 . To date, however, no study has examined the impact of tobacco tax increase across income groups at the subnational level. In this study, to support tobacco control policies through increased tobacco taxation in India, we estimated the impact of a cigarette price increase through an increase in excise tax from the 2018-19 tax structure (Underlying data 16 ) to 10 Indian Rupees (INR) plus 10% ad valorem, across income groups in four states in India.

Methods
We used the extended cost-effectiveness analysis (ECEA) model that was developed in the Disease Control Priorities Project 17 , and was previously used to estimate the impact of a 50% increase in the price of cigarettes on health, poverty, and financial outcomes in 13 middle-income countries, and the impact of cigarette price increase in Vietnam 12,14 . The model was used to estimate the impact across five income groups in four states in India of a cigarette price increase, resulting from a tax increase from the 2018-19 tax structure to INR 10 plus 10% ad valorem for all lengths of filtered cigarettes on smoking reduction, deaths averted due to major tobacco-attributable diseases (chronic obstructive respiratory disease (COPD), stroke, heart disease and cancer), life-years gained, treatment cost averted, number of men avoiding catastrophic health expenditures and extreme poverty, additional tax revenues raised from the tax increase, and cost savings to the Ayushman Bharat Pradhan Mantri Jan Arogya Yojana (AB-PMJAY).

Study population
We selected four states in India, namely Karnataka, Assam, Uttar Pradesh, and Maharashtra, based on recommendations from the Indian Ministry of Health and Family Welfare. These four states are also characterized by diverse socioeconomic demographic characteristics, tobacco use, and health insurance coverage. We focussed on male smokers aged 15 years and older, as they constitute more than 90% of all cigarette smokers in India 5 . To estimate the number of smokers by age and income groups in each state, we applied the age-specific smoking prevalence for males in each state from the second round of GATS survey in India in 2016-17 to the male population in each state 4 . The male population in each age group for each state was estimated demographic data from the 2011 Census of India, applied to the Indian male population and projected by the United Nations Population Division for 2018 18,19 . As the GATS survey did not report prevalence based on income, we used education level as a proxy for income group, such that those with no formal education were classified as the poorest 20% of the population and those with college education or higher were classified as the richest 20% of the population.

Cigarette price and price increase
We calculated the base price of the different lengths of filtered cigarettes sold in India (less than 65 mm, 65-70 mm, 70-75 mm and longer than 75mm) using data from Market price A.C. Nielson for 2014 20 , and inflated the price using the inflation rate in India for 2018 obtained from the World Bank Development Indicators 21 . We focussed only on filtered cigarettes as they constitute more than 92% of the market shares in India. Applying the Goods and Services Tax (GST), GST Compensation Cess and the National Calamity Contingency Duty (NCCD) in accordance with the tax structure for cigarettes in India in 2018-19 to the base price 22 , we obtained the 2018-19 market price of a pack of 20 cigarettes. Using the proposed tax of INR 10 (which includes GST, GST Compensation Cess, and NCCD) plus 10% ad valorem, we calculated the percentage increase in market price. We assumed that the tax increase will be passed on to consumer prices, as recent analyses in India showed that in nearly all cases, most of the tax hikes were passed onto smokers, but tax decreases did not reduce consumer prices 23 . The market price was assumed to be uniform in all of the four states studied.

Price elasticity
For this study, we considered a conservative scenario and used the average price elasticity for cigarette demand in both high and low-and middle-income countries of -0.40 12,24,25 . The price elasticity in India is likely to be higher, as John (2008) reported a price elasticity of -0.34 26 . We used this elasticity in our sensitivity analysis (see Sensitivity Analysis section below). Research suggests that young people and smokers who are on low income are more price sensitive than older people and smokers who are on high income 24,27 . Hence, we used two times the price elasticity (-0.80) for young smokers aged 15-24 years, as reported by the U.S. Department of Health and Human Services and International Research Agency for Cancer 24,28 . For smokers in the lowest income quintile and those in the highest income quintile, we used the price elasticity of -0.64 and -0.12, as used by the Global Tobacco Economics Consortium (GTEC) 12 . We also assumed that the price elasticity of quitting is half of the price elasticity of demand.
Effects of cigarette price increase on quitting, life-years gained, disease costs, income poverty, tax revenue, and cost savings to AB-PMJAY We used the ECEA model to estimate the impact of the cigarette price increase on quitting, number of deaths attributable to four major tobacco-attributable diseases (chronic obstructive respiratory disease (COPD), stroke, heart disease and cancer) averted, life-years gained, treatment cost averted (all treatment cost, most of which are paid out of pocket in India) due to the four tobacco-attributable diseases, number of men avoiding catastrophic health expenditures and extreme poverty, as defined by the World Bank as income of under $1.90 per day in purchasing power parity, and additional tax revenues collected 12,14 . To estimate the cost savings to AB-PMJAY─ the National Health Protection Scheme in India which provides a health insurance cover of INR 5 lakh to families living below the poverty line─ we added the cost savings in AB-PMJAY due to men prevented from falling below the poverty line to the treatment cost for those living below the poverty line before the tax increase. To obtain the former cost, we applied a 50% risk of dying from tobacco-attributable diseases, the proportion of smoking-related deaths due to COPD, stroke, heart disease, and cancer obtained from the Indian Million Death Study 2 , and the treatment cost of each disease to the number of men who would otherwise have continued to smoke and hence, fall below the poverty line. The data inputs and sources of data are available as Underlying data 16 . For indicators where state-level estimates could not be obtained, national estimates were used. All treatment costs obtained from the literature were converted to INR and inflated to the costs in 2018 using the exchange rate and consumer price index obtained from the World Bank Development Indicators 21 .

Sensitivity analysis
We conducted sensitivity analyses to examine the impact of a 25% and 100% price increase with the price elasticity of demand for cigarettes of -0.40, and that of INR 10 plus 10% ad valorem with the cigarette price elasticity of -0.34 in India 26 . For the lowest income group, we used a price elasticity of -0.635, as done by GTEC 12 .
All analyses were performed using Stata version 15.1 29 .

Results
Smoking prevalence in Karnataka, Assam, Uttar Pradesh and Assam Among the four states studied, Assam has the highest male cigarette smoking prevalence of 9.7% (Table 1), and Uttar Pradesh has the highest absolute number of cigarette smokers of about 6.5 million men. Maharashtra has the lowest smoking prevalence overall as well as for almost all age groups. In Assam and Maharashtra, cigarette smoking prevalence is higher among the higher income groups, whereas in Uttar Pradesh, the prevalence is higher among the lower income groups. In Karnataka, the prevalence was comparable across all income groups.
Before the price increase, an estimated total of about 10.7 million males older than 15 years smoked cigarettes across all of the four states studied (Table 2). Men in the bottom income group (poorest 20% of the population) accounted for 21%, while men in the top income group (richest 20% of the population) accounted for 15% of the total number of male smokers. ). An estimated total of 665 thousand deaths due to COPD, stroke, heart disease, and cancer would be averted among current smokers due to quitting. The number of averted deaths in the bottom income group would be 7.4 times that in the top income group (210,289 vs 28,610). The deaths averted due to quitting would yield an estimated 11.9 million life-years, with the bottom income group gaining 7.3 times more life-years than those in the top income group (3,751,930 vs 513,319).

Sensitivity analysis
Our sensitivity analysis yielded similar results. Across all three scenarios (25% price increase, 100% price increase, and price increase to INR 10 plus 10% ad valorem with a price elasticity of -0.34), the ratio of the additional life-years gained, treatment cost averted, and number of men avoiding catastrophic health expenditure between the bottom and the top income groups is similar to that in the baseline scenario (Figure 1a-c).
Compared to the baseline scenario, with a 25% cigarette price increase, the ratio of the additional tax revenue collected from the top income group to the bottom income group increases in all four states, while with a 100% price increase, the bottom income group would accrue tax savings as a result of quitting and reduced consumption (Figure 1d). The cost savings to AB-PMJAY, in Assam and Uttar Pradesh, from a 25% price increase would be less than half that from the baseline scenario (INR 26 crore 55 crore in Assam, and INR 290 crore vs 618 crore in Uttar Pradesh), while that from a 100% price increase would be almost double (INR 103 crore in Assam and INR 1,160 crore in Uttar Pradesh) (Figure 1e). The findings using the Indian price elasticity of -0.34 are similar to that of the baseline scenario.

Discussion
This study confirms that significant tax increases on filtered tobacco would be associated with pro-poor health outcomes as well as significant reductions in poverty and levels of catastrophic health expenditure in the Indian population. Increasing the market price of filtered tobacco is estimated to have a higher impact on the smoking rates of lower income Indian households, and as a result, poorer households bare a lower share of the overall net increased market cost. This study reaffirms that tobacco taxation is an important public policy approach that is associated with a pro-poor impact in India, as also found in other LMICs 10-15 .
The findings from this subnational analysis of four states are broadly consistent with the trends seen in the national results 12 . However, the magnitude of effects differs across states. For example, the ratio between the outcomes between the first and fifth income groups is largest in Uttar Pradesh due to a wide gap in the rates of cigarette smoking between the first and fifth income groups (11.4 vs 6.3%). In this state, the poor would benefit the most from large tax increases and this benefit is significantly higher than the national estimate. The subnational results provide an important economic platform to engage with statelevel decision makers and other stakeholders to demonstrate the differential impact that higher cigarette taxes can achieve within their jurisdictions. These results should be used to stimulate action from the local level and foster a groundswell of political buy-in to advance the national agenda to adopt higher tobacco taxes. Leveraging sub-national data to develop an economic case for stronger tobacco taxation policies is a strategy that is currently being pursued in several other settings including Mexico and Colombia 30 .
The number of men avoiding impoverishment is minimal in Karnataka and Maharashtra, which is likely because of a higher income level overall in these two states. Similarly, cost-savings to AB-PMJAY is also small in these states as fewer people are eligible to AB-PMJAY. While the potential to alleviate poverty was minimal, the higher taxes would be associated with an avoidance of a significant amount of treatment costs and reductions in catastrophic health expenditures in both states and the highest share of these reductions would be among the poorest groups. By applying the ECEA method, we have demonstrated the impact that higher cigarette taxes will have on the two key indicators that are used to assess progress towards achieving financial risk protection and universal health coverage, namely avoidance of illness-related poverty and catastrophic health expenditures 31 . This work highlights the criticality of this fiscal policy measure for achieving universal health coverage and the Sustainable Development Goals (SDGs).
There is now robust evidence generated from several LMICs that support the role of tobacco taxation policies in achieving significant health outcomes, particularly among the poor, as well as an overall reduction in poverty 10-15 . By curbing rates of tobacco-related non-communicable diseases in LMICs, tobacco taxation is also a means to advance economic development in these settings. Yet, the momentum for implementing and sustaining this fiscal policy approach has yet to be achieved in India and other settings. Advancing more aggressive taxation of tobacco across India and other LMICs is urgently needed to align with the recommendations of the FCTC. Recent reports suggest that for most countries, the increases that have been made to date are still far too small to significantly alter smoking rates 32 .
This study has limitations. First, we used education as the proxy for income. This may underestimate the true socioeconomic status of households. Second, a subset of subnational inputs was not available and so national inputs were used. In some cases this could under or overestimate the effects, though we anticipate the impact would be minimal as national data were substituted for less than 20% of overall inputs. Third, AB-PMJAY has expanded to beyond covering only those living below the poverty line 33 . Hence, savings to AB-PMJAY due to averted treatment costs is likely much higher than estimated. Fourth, we have not accounted for the potential substitution to other tobacco products.

Conclusion
Tobacco taxation remains the single most effective public policy approach for curbing smoking rates. In this study, we confirm that the poorest will benefit the most from an increase in cigarette taxes, with more life-years gained, more premature deaths and treatment costs averted, and more cases of poverty and catastrophic health expenditures avoided compared to the richest income groups. By applying the ECEA method at the subnational level, we provide first-ever estimates of the differential impact of cigarette taxes within the Indian population. The estimates generated for each of the four states provide a powerful economic input that will equip state-level decision-makers to make a strong case for higher taxes within national debates. From this work, state-level decision makers can also make the important link between cigarette taxes and the SDGs (e.g. SDG 1: End poverty in all its forms everywhere), given the substantial impact that higher taxes will also have on alleviating poverty within the population.  This is an interesting paper that considers the impact of a sizable increase in the excise tax in four Indian states. By international standards, smoking prevalence in India (and in these states) is not high. The authors show that an increase in the excise tax to 10 INR (presumably per stick, although this was not clear) plus an additional 10% tax would result in positive public health and fiscal results in all four states.

Data availability
The rationale for the choice of these four states is not clear, but one gets the impression from the Discussion that this research was done with the aim to advocate for higher taxes in these four states. That is a worthy cause, although, given the size of India, a tax at the national level would have a multiplied effect, compared to a tax increase in just four states.
The significant contribution of this study is that it indicates that an excise tax increase is strongly progressive/pro-poor, in that the poor benefit much more than the rich.
While the results are generally believable at the intuitive level, the authors do not provide sufficient information to allow others to replicate the study. The DOI link to the data did not work when I tested it.
The finding that the excise tax increase will result in many thousands of men escaping extreme poverty, requires further explanation and justification. The authors indicate (quite correctly) that the poor's demand for tobacco is more price elastic than the non-poor. Thus a greater percentage of the poor will quit when faced with a higher price of cigarettes (or tobacco more generally). But what about those smokers who do not quit smoking when they are faced with a higher price? Some may cut back on their consumption (there is evidence of that in the literature), but even for them, their total expenditure on cigarettes/tobacco may increase. For some poor smokers who are unable/unwilling to reduce their consumption, a sizable increase in the price of cigarettes will have a significant effect on their total expenditure. In some cases this will push some smokers into extreme poverty (for those who are on the verge of extreme poverty before the tax increase), or push them further into extreme poverty. The paper says nothing about this

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Reviewer Report 07 July 2020 https://doi.org/10.21956/gatesopenres.14311.r28868 © 2020 John R. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Rijo M. John
Centre for Public Policy Research (CPPR), Kochi, Kerala, India

General comments:
This study examines the implication of an arbitrary cigarette tax increase on select outcomes such as life years gained, tax revenue generated, amount of catastrophic health expenditures averted and savings on a national health care scheme. It focuses its analysis on 4 select Indian states-Karnataka, Assam, Uttar Pradesh, and Maharashtra-chosen rather arbitrarily. This is one of the few recent studies to have examined the impact of tax increase at the subnational level and the kind of outcome variables the study focuses are quite relevant from a public health policy point of view. The study and its conclusions are quite useful for tobacco control policy in India, and taxation in particular. An important policy relevant conclusion from the study is that cigarette tax increases are disproportionately beneficial for India's poor. This counters the age-old tobacco industry argument that tobacco taxes are regressive. This study establishes it is not.
Although the study provides insightful analysis of the impact of a proposed tax increase, it comes with certain limitation most of which may be addressed rather quickly. Below are my specific comments on the same, not necessarily in the order of its importance.

Specific comments:
Methods para 1 says "Tax increase from the 2018-19 tax structure to INR 10 plus 10% ad valorem for all lengths of filtered cigarettes". It is not clear why this particular combination of the tax was selected as a proposal and whether this will be in addition to the existing taxes or inclusive of the existing tax. What is also not clear is where this is proposed to be applied given that current cigarette tax structure has varied components including excise tax, national calamity contingent duty (NCCD), base GST rate (28%) and compensation cess. However, later on, from the appendix table provided, it becomes clear that the proposed tax is to be applied as a compensation cess. These details may not be obvious to a less informed reader.

1.
Study population: "we used education level is a proxy for income group". While it may be fine to use education as a proxy for income, GATS allows estimation of wealth quintiles as it 2.
provides a variety of information on the possession of different assets. Principal component analysis is usually applied to estimate such wealth quintiles from this type of surveys. The use of education as a proxy for income may be the reason for some strange prevalence figures in Table 1. For example, cigarettes, especially the filter cigarettes, are usually consumed by the relatively well-off people in India and the poor tend to smoke more of bidis. In Table 1, however, it is seen that the bottom two quintiles have cigarette smoking prevalence many times higher than the upper quintiles in UP. Although this proxy is mentioned as a limitation, I think, this is something that can be overcome using more appropriate techniques for generating income groups.
Cigarette price and price increase: "we obtained the 2018-19 market price of a pack of 20 cigarettes". It is not clear why FY 2018-19 was chosen as the benchmark year for the study simulation. Union budget for FY 2019-20 re-introduced excise taxes on tobacco products, albeit nominally. This changes the tax structure and it would've been better had the analysis made after taking this into account. In addition, the union budget for FY 2020-21 more than doubled NCCDs applied on cigarettes. Because the benchmark year is kept as 2018-19 and there have been two rounds of tax increase including one structural change in the subsequent annual budgets, the actual estimated impact of this proposal may not be accurate as the actual prices are already higher than what is assumed in the simulation. It would be better for the simulation benchmark to reflect the actual existing taxes (both rate & structure) on cigarettes in order to estimate the true impact.

3.
Price elasticity: Income group wise price elasticity estimates are available in India from a study in 2015 by Selvaraj et al. (2015 1 ). The authors might want to try these out or use these at least for sensitivity analysis.

4.
The subsection "Impact of tax increase to INR 10 + 10% ad valorem" need to make it clear upfront whether the proposed increase is on top of existing taxes or inclusive of existing tax. Upon examining the supplied appendix table 1, I was able to figure out that it is indeed inclusive of the existing compensation cess. But, the simulation benchmark already says the average existing ad valorem compensation cess is 16.7% while the specific cess is 3.27. The proposal is actually to increase this 3.27 to Rs.10 and decrease the 16.7% to 10%. This needs to be spelled out as such. The authors should also provide the rational for this proposal which appears to be arbitrary in nature. Why would the proposal seek a reduction in existing ad valorem rate while proposing a hike in specific tax? Why Rs. 10 specific and not 15 or 5 or 20? Why there's an ad valorem rate at all and increase the complexity of tax structure and why not try to make a proposal that will simplify the structure instead? A discussion must address these questions.

5.
Given that the model uses 2018-19 benchmark as mentioned above, it is not capturing the current excise component in existing tax structure. Moreover, to my knowledge, NCCD is applied along with the excise. The GST and ad valorem compensation cess are applied on this excise inclusive price. However, the model simulation and the estimation of taxes as given in the appendix table 1 doesn't seem to do it this way. On what basis the proposal assumes that the proposed increase in tax to be applied on compensation cess and not on, say, excise or NCCD? Depending on where this is applied, the implications and estimated final prices and tax burden will differ. It should be also remembered that compensation cess is set to be expired 5 years from the time of GST introduction which means only remaining 2 6.
years for its expiry. If the proposal is to apply the tax increases on compensation cess how can this be sustained when the compensation cess expires? Is proposing the increase on excise a better idea? I suggest authors discuss these issues somewhere on the paper.
The state-wise discussion of the results talk about the tax revenue generated from the proposed tax increase. It inadvertently give the impression that these tax revenues are accrued to the respective states implementing this tax increase. However, the compensation cess in its entirety goes to the union govt. which in turn uses this to compensate all state governments for revenue shortfalls, if any, they face up to 5 years post GST. Also, under the GST, this increase in compensation cess can't be made unilaterally by any single state. Any proposal to increase compensation cess will be done through the GST council and shall be applicable to all states. To my knowledge, there is no mechanism for individual states to change the current tax structure on products under the GST. The authors need to clarify/discuss what their proposal will entail in practice.

7.
The mentioned limitation talks about not accounting for substitution. Although there is no strong evidence of substitution between cigarettes and bidis the same cannot be said about substitution between filtered and un-filtered cigarettes. If tax is increased only on filtered cigarettes, one must realistically expect a certain amount of substitution towards unfiltered cigarettes at least by the relatively lower income group of people. It is not clear why the increase proposal cannot be made such that the increase happen simultaneously on all tiers of cigarettes. This will largely take care of such possible substitution and capture the overall impact more accurately.

8.
It is also not clear how the model simulation captures the effect of income growth as there is no discussion on the same. The paper only talks about price elasticities. However, annual per capita income growth is significant in LMICs like India. While price increase has negative impact on consumption, income growth can compensate for part of these impact. A combination of price and income are important in determining true affordability of these products. It would be ideal to capture these in the model.

10.
Typo: Appendix Table 2. Data inputs and data sources (Study data input with sources for analysis)): On row 51 it says the amounts are in VND. It should be INR.

11.
© 2020 Rao M. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

M. Govinda Rao
Takshashila Institution, Bengaluru, Karnataka, India Comments on the Study, "Impact of cigarette tax increase on health and financing outcomes in four Indian states": This is an important study analysing welfare implications of increases in tax on cigarettes in four Indian States namely, Karnataka, Assam, Uttar Pradesh, and Maharashtra. The authors use an extended cost-effectiveness analysis (ECEA) model developed in an earlier Disease Control Priorities Project to estimate the impact of cigarette price increases caused by tax increase in 2018-19 across five income groups in four states in India. The welfare implications are analysed in terms of smoking reduction, deaths averted due to major tobacco-attributable diseases (chronic obstructive respiratory disease (COPD), stroke, heart disease and cancer), life-years gained, treatment cost averted, number of men avoiding catastrophic health expenditures and extreme poverty, additional tax revenues raised from the tax increase, and cost savings to the Ayushman Bharat Pradhan Mantri Jan Arogya Yojana (AB-PMJAY).
The study estimates that, in 2018-19, an increase in the price of cigarettes by 53.25% due to increase in the tax for a pack of 20 would have led to about 1.5 million men quitting smoking across the four states, with the bottom income group having 7.4 times as many quitters as the top income group. It would also have averted 665 thousand deaths due to COPD, stroke, heart disease, and cancer among current smokers due to quitting. The deaths averted due to quitting would had added an estimated 11.9 million life-years, with the bottom income group gaining 7.3 times more life-years than those in the top income group. The cost saving on hospitalisation for avoiding treatment of major tobacco-attributable diseases is estimated at Rs. 1,729 crore ($Int 955 million and the saving is estimated at 7.4 times higher for the bottom income group. This would have helped almost 75 thousand people to falling into extreme poverty. The saving on treatment cost and impoverishment is estimated at Rs. 672 crore ($Int 371 million) in AB-PMJAY and additional tax revenue generated is estimated at Rs. 4,385 crore ($Int 2,422 million).
Thus, the study confirms that significant tax increases on filtered tobacco could improve pro-poor health outcomes and appreciable reductions in poverty. This study reaffirms that tobacco taxation is an important pro-poor public policy. This is a competent piece of work with significant public policy implications. I recommend the article for indexing.

Is the study design appropriate and is the work technically sound? Yes
Are sufficient details of methods and analysis provided to allow replication by others?