FCTC Ratication, Smoking Prevalence and GDP per Capita : Lesson Learn for Indonesia

Background The stagnated tobacco control progress in Indonesia needs to be accelerated through a more comprehensive implementation of Framework Convention of Tobacco Control (FCTC) measurement. Nevertheless, the tobacco industry argument concerning the negative economic impacts of tobacco control still hinders the government to ratify or even sign the FCTC, which has been ratied by more than 180 countries. This study aims to bring the empirical evidence on the tobacco industry argument concerning FCTC. This study applied two stage least square estimation strategy to unbalanced panel data at country level. On the rst stage we estimate the impact FCTC ratication on smoking activity, and on the second step, estimating the inuence of smoking activities on macroeconomic performance. Results The result of this study shows that FCTC ratication is negatively related to a country’s smoking prevalence, in which the ratifying party of FCTC has lower smoking prevalence. Moreover, country who raties FCTC longer is also associated with lower smoking prevalence. Whereas FCTC ratication is benecial in reducing smoking prevalence, the declining smoking prevalence is not related to the decline in GDP per capita. Conclusions The result of this study shows the decrease in smoking prevalence has nothing to do with the macroeconomic indicator. Hence, FCTC ratication, which is an important driver for tobacco control actions acceleration, should not be seen as a backre to the economy. Instead, FCTC ratication could be mutually benecial for the health and economic aspects as it provides comprehensive guidance and protocols by taking into account the well-being states of both aspects.


Background
The tobacco control situation in Indonesia is getting worrisome as during the past decade the tobacco control indicators show stagnated progress or even declining progress in some important indicators.
The non-optimal tobacco control effort in Indonesia resulting in stagnated progress is related to weak Framework Convention on Tobacco Control (FCTC) protocols enforcement in the country. Compared to other similar countries, Indonesia, being the only country that is neither a party nor a signatory of FCTC, is faced by though tobacco use situation as shown in Fig. 1 below.  Figure 1 above shows that the three countries had successfully reduced the overall smoking rates consistently and signi cantly since 2005. In contrast, within the same period, Indonesia experienced increasing smoking prevalence by nearly 5%. Brazil, amid being the world's largest tobacco exporter, has successfully driven its smoking prevalence rate to 13.9% below the world smoking prevalence rate of 20.5%. [2] Among the four countries, Brazil is the only country that has complete policies on all of the MPOWER measurements, with remarkably 83% of tobacco tax implemented. [3] While Pakistan and Bangladesh still have rooms for moderate and or minimal policies are some of the MPOWER measurements, both countries, under FCTC rati cation, can fully cover all the important policies measurement. [4,5] Whereas Indonesia, being neither the party nor signatory of FCTC, remains with very weak policies on tobacco advertising bans and cigarette affordability. [6,7] Hence, it comes as no surprise that Indonesia still struggles with tobacco control efforts.
Amid the alarming tobacco use condition, after 15 years, Indonesia still yet to ratify the Framework Convention on Tobacco Control (FCTC). Indonesia is currently the only country in Asia-Paci c to not ratify FCTC. FCTC, which was agreed upon by 192 WHO members country in 2003 and effective as per 2005, is currently rati ed by more than 180 countries as it is seen as an important protocol to enforce tobacco control regulation. Economic concern still is the main reason that hinders Indonesia from FCTC rati cation. It is believed that FCTC rati cation which allows comprehensive tobacco control policies enforcement in Indonesia will negatively impact the economy through the decline in tobacco production and tobacco industry economic activities. The argument against FCTC rati cation in Indonesia believed that the decline in tobacco production and tobacco industry economic activities will decrease the tobacco farmers' welfare, increases unemployment for tobacco industry workers, and eventually lead to major macroeconomic impact -poverty rate and unemployment rate increase and GDP rate decrease. Despite the strong statement provided by the opposition view, there is still no solid evidence to support this argument. This paper hence aims to provide evidence on how tobacco control actions under FCTC rati cation impacts the economy by comparing the economic conditions of the countries that have rati ed the FCTC between 2005-2010.
The topic of the economic impact of FCTC is very important for both academic discussion and the policymaking process. Lack of academic investigation on this topic leaves out the room for any negative speculation against the FCTC protocol and hence the academic research is very much needed to ll the space. Moreover, to investigate the economic impact of FCTC will help the policymakers to understand that while FCTC is highly important for tobacco control efforts, it does not necessarily relate to negative macroeconomic indicators as the tobacco industry does not signi cantly contribute to the Indonesian economy.

Methods
The study employs unbalanced panel data using two-stage least square (2SLS) estimation analysis. The technique that is used in 2SLS regressions requires two stages: in the rst stage, we estimate the impact Page 4/13 of FCTC rati cation on smoking prevalence activity. In the second stage, the instrument-estimated value from stage one is then computed as a predictor when estimating the effect of smoking activity in the economy. The sample in the panel dataset consists of 87 countries over the period between 2000 and 2016. Due to our main focus on this research is to compare the enforcement of FCTC rati cation at the country level, the primary sample is countries that rati ed FCTC between 2005 and 2010.
The model to be estimated is given by (1) In which a Y is the dependent variable of smoking prevalence, X are the independent factors that explains the smoking behaviour, T is a variable of time counts of rati cation or the length or the duration of a country's FCTC rati cation in year, and D is a dummy variable representing country that is the ratifying party of FCTC. Subscripts i and t explains correspondences of country i at time t, whereas ε is an error term.  (2)), macroeconomic proxy as dependent variable used in this study is per capita GDP. The effect of FCTC rati cation is denoted as Ŷ, smoking prevalence hat (the difference between actual smoking prevalence and the rst regression residual value). M denoted other macroeconomic variables representing correlation between macroeconomic indicators and per capita GDP. Subscripts i and t explains correspondences of country i at time t, whereas ε is an error term.

Source: Authors Calculation
The FCTC coe cients, both on the FCTC rati cation duration and the FCTC rati cation participation appear to signi cantly affect smoking prevalence. The effect of dummy FCTC is signi cantly negative at the signi cance level of 1%. This means that smoking prevalence in a country will decline if the country is a ratifying party of FCTC. Moreover, the longer a country participated in FCTC rati cation, the lower the smoking prevalence will be. Whereas the percentage of productive ages population shows a positive signi cant relationship with the smoking prevalence. This means that the higher the share of productive ages population in a country, the higher is the smoking prevalence in the country. Whereas the percentage of value added taxes is not statistically signi cant to the smoking prevalence. This means that the smoking behaviour in a country is not driven by the tax rate. Standard errors are shown in parentheses. ***, ** and * denote p-value < 0.01, < 0.05 and < 0.10 respectively.

Source: Authors Calculation
Holding other variable constant, the effect of smoking prevalence hat is not statistically signi cant. This result showing an insigni cant relationship between smoking behaviour and per capita GDP. Intriguingly, the smoking prevalence hat shows an inverse relationship with per capita GDP. This indicated that the higher is the smoking prevalence, the lower per capita GDP will be. Unfortunately, the relationship between the two variables is not statistically signi cant.
However, other control variables such as mean year of schooling, exchange rate, consumer price index, value-added agriculture, democratization index, and corruption index are statistically signi cant to GDP per capita. Among the other controlling variables, only the openness is not statistically signi cant to the per capita GDP.
Based on the regression estimation result, when a country is a ratifying party of FCTC, the country will experience a declining smoking prevalence. However, the changes in smoking prevalence is not statistically signi cant to the macroeconomic indicators. In general, it can be concluded that being a party of FCTC rati cation will not directly impact the macroeconomic activities of a country.

Discussion
The result above implied two imperative discussions that are the main interest of the study. Firstly, the countries that participate in FCTC rati cation will have higher probability to succeed in reducing their smoking prevalence. This is consistent with the previous studies that found negative association between FCTC implementation and the countries smoking prevalence. [8][9][10][11] Moreover, this study also found the negative association between the years of FCTC rati cation and the changes in smoking prevalence. The negative relationships between FCTC variables and the smoking prevalence inferred the importance of FCTC rati cation in supporting a country's tobacco control effort.
The second key nding of the study is that the changes in smoking prevalence -which is also determined by FCTC rati cation -does not necessarily related to the macroeconomic indicator (in this case GDP per capita). GDP per capita is a very broad macroeconomic indicator determined by numbers of factors. This study nds that cigarette consumption -portrayed by the smoking prevalence is not one of the determinants of the countries' GDP. This nding is counterintuitive with the premise that the declining tobacco industry contribution will negatively impact the economy. The data on tobacco industry economic contribution (i.e. through tobacco taxes, labour absorption, and tobacco farming, and even tobacco exports), primarily in major tobacco producing countries, is seemingly positive. However, the magnitude of the tobacco economic contribution is oftentimes overstated and resulted in the misleading argument that tobacco control will be an economic back re. [12] Moreover, the tobacco industry economic contribution still does not worth the health cost borne by the country due to increasing smoking related NCDs, loss of productivity, and premature deaths. [13,14] The failure to calculate for the negative externalities created by smoking frequently overlooked by the tobacco industry proponent. A study in Canada implied that the total tobacco-related costs from direct, indirect, and various induced types of impact exceeded the tobacco industry contribution through tobacco taxes by not less than $291.8 Million in 1977. [15] In addition, the study in China as the largest tobacco producing country also found that compared to the other cash crops, tobacco has the lowest economic rate of return while at the same time the government tax revenue from tobacco industry continues to decline. [16] Intuitively, it is safe to say that FCTC rati cation (leading to declining cigarette consumption) is not necessarily related to the changes in GDP. As opposed to the tobacco industry argument, the nding of this study suggests that declining smoking prevalence as the result of a more comprehensive tobacco control effort under FCTC rati cation might reduce the country's healthcare cost. Previous studies argued that smoking cessation is related to reduction in healthcare cost, primarily in short-term healthcare costs. [17,18] Moreover, reducing smoking prevalence could be done at relatively low costs, resulting in higher long-term bene ts due to increasing health aspects. [19,20] In addition, one of the way this could be done with mutually bene cial outcomes for the economy and health aspect is by raising tobacco taxes. Increasing tobacco tax rate will simultaneously increase government tax revenue and decrease tobaccorelated medical costs. [21] This adds to yet another evidence on the importance of demand-side tobacco control for both economic and health aspects.
As important as demand-side control, supply-side tobacco control will also lead to the desired economic and health bene ts. Controlling for tobacco supply, which is also part of FCTC agreement, will be highly important in protecting the impacted groups welfare. As one of the main concerns that still hinders Indonesia on becoming FCTC ratifying party is the tobacco farmers and tobacco industry workers welfares. Nevertheless, the current Indonesian tobacco business practice which highly relies on imported tobacco products, potentially brings a substantial loss to Indonesian tobacco farmers as well as the economy. [22] In contrast, article 17 of FCTC mentioned the support for viable alternative economic activities for the impacted groups.

Conclusions
Upon discussing the importance of controlling for tobacco demand and supply, FCTC rati cation, which in this paper found to be one of the factors that in uences the decline in smoking prevalence, will be the important key to accelerate the tobacco demand and supply control in a country. As FCTC provides comprehensive measures on tobacco control actions, FCTC rati cation, which this paper found to have no direct impact on the macroeconomic indicator, will assist the country to accelerate the decrease in smoking prevalence. Contrary to what the industry believes, not only is FCTC bene cial to the health aspects by the declining smoking prevalence, some key tobacco control efforts in FCTC rati cation are also advantageous for the economy.

Declarations
Ethical Approval and Consent to participate Ethical approval is not applicable.