I wrote a paper on cigarette smoking (with Bao Jia Tan) while at BEDA in Beijing in 2010 – here is a pre-print. While in China this time I have tried to argue the point with Chinese students that there was a strong case for increasing tobacco taxes in China and that this tax should be a specific rather than an ad valorem charge. I got sufficiently interested in clarifying my new arguments that I wrote them up (after discussing with Echo, Kening and Xixi) . Thoughts are preliminary.
The Case for a Large Specific Tax on Cigarette Smoking in China
Cigarettes are a legal product in China but a product that can kill (as it does everywhere) when consumed as intended. It is responsible for the deaths of 1.2 million Chinese smokers each year and 100,000 more from the effects of secondary tobacco smoke. There are forecasts that, in China, deaths could treble by 2030 unless more drastic policy actions to limit smoking are taken. There has been nearly a 5-fold increase in lung cancer deaths since the 1970s although part of this is due to air pollution problems as well as smoking.
There are two major arguments for imposing hefty taxes on cigarettes that are fairly specific to China.
- The existence of substantial “secondary tobacco smoking costs” that smokers impose on non-smokers. Often these costs arise from men smoking and the smoke impacting on their wives and children. In China it is estimated 100,000 deaths occur each year in this way. This is a straightforward negative spillover externality that motivates corrective taxes but also motivates rules regulating smoking in inappropriate situations and moral suasion arguments against smoking in private cars and the home when non-smokers are present..
- There are also asymmetric information externalities that arise if smokers do not realize the damage that smoking can cause to their own health. Unfortunately it seems to be a fact that many people – including doctors – in China do not understand the connections between smoking and both heart disease and lung cancer. Even if these diseases do not occur smoking long enough creates the high likelihood of contracting emphysema or a loss of elasticity in your lungs. Emphysema can mean a miserable old age with breathing difficulties and a difficult death. In China emphysema kills more people than heart disease. The best way of addressing this type of externality is through public information (propaganda) campaigns although taxes are also useful.
It is sometimes argued that extra health costs are an externality but this is not so if people pay their own costs and does not seem a relevant issue in China. Even if the government did pay health costs through a social insurance scheme, the cost of such health costs will often be less than the taxes paid – even for low taxes. Smoking makes the government money because the taxes it gets exceed health costs of smoking particularly because smoking kills you quickly. It is a sad fact of life that if you contract lung cancer you do not live long.
So in a country like China heavy taxes on cigarettes make sense primarily because of secondary tobacco smoke and asymmetric information issues. There are paternalistic motivations as well. It is sensible to advocate policies that avoid unnecessary death and disease because people, particularly youth, ignore well-known information regarding the health consequences of smoking.
Currently China has about the cheapest cigarettes anywhere. Inexpensive brands such as the local brand “Baisha” cigarettes, sold in Hunan Province, sell for 5.5 RMB which is about 90 cents. As argued in Clarke and Tan (2010), cigarette prices in China have increased more slowly than incomes have grown over the past 20 years so they have become more affordable. If the Chinese Government is serious about reducing the incidence of cigarette smoking then a substantial increase in the price of cigarettes needs to occur through increased taxes on cigarettes. Provided the demand for cigarettes is not too inelastic this should reduce the demand for cigarettes and the incidence of smoking.
The evidence on price elasticities of demand for cigarettes in China is that they are very low. This is puzzling since a standard Chinese objection to steeper cigarette taxes is that they would substantially reduce revenues from cigarettes. At first sight, with the very low price elasticities, government revenue from taxes could, in fact, be greatly increased by levying higher taxes.
This is misleading however. The very inelastic demands in China are due to the fact that taxes here are mainly levied ad valorem – as a fixed percentage of the pre-tax sales price. This is analyzed in White et al (2013) who reason that part of the reason observed elasticities in China are so low is brand-switching from more expensive to cheaper brands. That this is encouraged by ad valorem taxes can be made transparent by means of an example.
China has very cheap cigarettes but a wide range of cigarette prices. Cigarette prices range from the 5.5 RMB for Baisha cigarettes to very expensive brands such as the brand “Harmonization” which sells for 100 RMB. There is a continuum of cigarette prices across this range.
Doubling the current 35% ad valorem tax on a 5.5 RMB pack will increase its cost to around 6.8 RMB while a packet of “Harmonization” would increase in price to 135RMB. People may respond to a tax increase simply by switching from very expensive brands to much less expensive brands. The propensity to do this is strong because with an ad valorem charge the cheap brands do not change much in price as a consequence of the tax. (This example might not be entirely accurate since it might be that China charges slightly higher ad valorem tax rates on what it calls “product brands” though the basic story does not change).
This type of tax increase might not lead to much less smoking but because it will simply cause switches to smoking less expensive brands. Ad valorem tax increases rank poorly in terms of their effectiveness in changing smoker behavior particularly at the inexpensive end of the cigarette market.
Moreover, tax revenue to government might not increase much since the increased tax will fall on less expensive brands which yield relatively less revenue and the brand switch is towards such cheaper brands. The major advantage of an ad valorem tax is that it is progressive and therefore ranks highly in terms of equity. It mainly falls on wealthy rather than poor smokers. Poor smokers will continue to die however from the effects of smoking.
An alternative to the ad valorem tax is a specific tax. This is a tax per cigarette or per packet of cigarettes. For example, this might be an extra tax of 5 RMB per packet of 20 cigarettes. Now a 5.5 RMB pack would cost 10.5 RMB and a packet of Harmonization would increase in price to 105 RMB. Now there is much lower tendency toward substitution towards cheaper brands since the cheaper brands become relatively much more expensive. In addition, since tobacco prices increase at the low price end of the market, the policy will be much more effective in reducing smoking. Finding cheaper substitutes for currently low-priced cigarettes might be difficult although there are possibilities for purchasing bulk tobacco and for smokers growing their own tobacco plants.
Poor people in China are known to have much more elastic demands and that is likely to be true also of young people who are considering taking up smoking.
It is also likely to be the case that government revenue increases with such a specific tax rate increase. Although revenue is lost at the low-priced end of the market not much revenue was raised at this end originally. At the high priced end wealthy consumers will continue to pay the higher price since their demands are much less elastic.
Finally, a specific tax has the strong theoretical advantage of better reflecting the external damages of cigarettes that are proportional to the quantity of cigarettes consumed not their value.
The case for increasing specific taxes looks strong except for one adverse effect namely that this tax is regressive rather than progressive in its impact. It will primarily impact on low-income smokers. Those low-income smokers who quit smoking gain the advantage of better health outcomes but difficulties arise for poor smokers who continue to smoke. Moreover. it is difficult to think of compensations that might be used to ensure such people are not severely affected by an increased specific tax. Perhaps the government might use revenues from such taxes to compensate those disadvantaged with subsidized quit campaigns or subsidized nicotine replacement products.
In rural areas of China, where people are generally poorer and less well educated, the use of increased specific taxes as well as increased spending on public information are important means of reducing smoking in China. The main difficulty, as mentioned, is in devising compensations for poor people.
Demographic issues might also have an impact. A frequent suggestion is that villages in rural China have many old and many very young members with intermediate-aged workers living in cities. This too might impact on rural smoking behavior in China. Old smokers will be often those who have smoked for many years and may face difficulties quitting. A sensible argument in such settings would be to target those young males who are at an age where initiating smoking is a possibility.
Most people in villages have access to TV sets so it would have thought that public programs warning of the health risks of smoking might be used to influence smoking behavior in these areas. In fact very little is being done in this regard. A plausible implication is that not much is being done generally in terms of public policy despite official claims to the contrary. Cigarette taxes are low, little is being done to publicly indicate the dangers of cigarette smoking and bans and restrictions on smoking in public places and restaurants are widely ignored because they are not enforced.