Sunday, October 07, 2007

Another look at Cigarette Taxes

Every now and then, things happen that produce what economists call a natural experiment. The last major one in the Phoenix area was when the gasoline pipeline from Tucson broke, and we lost 30% of our gasoline supply. It wasn't a lot of fun for gasoline consumers, but it provided a lot of interesting stuff for economists.

Last year, the voters provided us with a couple of others. One initiative raised the minimum wage and two others raised the cigarette tax. Back in August I commented on a news story about the effect of the cigarette tax increase. Two things made the tax increase interesting from an economic standpoint. First was the increase in tax stamp sales prior to the increase taking effect in January. This is a classic case of increasing demand today because you know that the price will increase in the (near) future.

The second interesting aspect of the tax increase is what it tells us about the price elasticity of demand for cigarettes. Governments like to tax things that are inelastic in either supply or demand. Inelasticity means that the quantity in the marketplace doesn't change much when the price increases due to the tax. Hence the tax base (quantity) is fairly stable and tax revenues increase by about the same amount as the tax rate. What economists also know is that elasticity is not constant along a demand curve. As prices increase, you tend to move from a more inelastic region of the demand curve to a more elastic region. Once you get into the elastic region, a price increase decreases quantities more rapidly which reduces total revenue and your tax base becomes unstable.

I wanted to revisit this for a couple of reasons. First, we've now got a couple of additional months of data. Second, congress and the administration are currently fighting over how much to increase the SCHIP program, and one proposal is to fund the higher spending with an additional 61 cent per pack cigarette tax.

For Arizona, if we've moved from the inelastic region of the demand curve to the elastic region, an increase in the federal cigarette tax would would have a potentially significant negative impact on state cigarette tax revenues. So where are we? (Like everyone, I'd like more time and more data, but we're in the middle of the debate today so it's time for a best guess.)

Here's a graph of some data.

It shows the implied monthly sales of tax stamps for cigarettes in Arizona. I got this by taking the monthly cigarette tax revenues recorded by the Arizona Department of Revenue in their monthly tax facts report and dividing it by the tax rate. The tax rate was $1.18 up until December, 2006 and $2.00 thereafter. I also plotted 3 month, 6 month and 12 month averages.

Given the data, the first effect of increasing tax rates is pretty obvious. Tax stamp sales increased significantly prior to the rate increase and fell substantially thereafter. This is in line with basic economic theory about demand.

The second effect should be a decrease in the monthly quantity of tax stamps due to an increase in the price of cigarettes. However, it is masked by the first effect so it's difficult to tell what the new month volume really is.

As best I can tell, if we hadn't raised taxes on cigarettes in January, we'd be selling about 21 million tax stamps per month today. Looking at the average for the last four months of data (April - July) it looks like a bit less than 17 million. That's a 20% decrease in quantity. Since the increase in the price of cigarettes was about 20% (from a little more than $4 to a little less than $5 or $0.82) that means the price elasticity of demand is now about 1.

Hence we're moving from the inelastic region of the demand curve to the elastic region. For every 1% increase in price, you get a 1% decrease in volume.

That also means that if congress raises cigarette taxes by $0.61 (a 12 % increase in prices in Arizona) volumes will fall by about the same or perhaps a greater percentage.

For the state of Arizona, a 12% decrease in volumes means average monthly cigarette tax revenues would decline from about $34 million per month to about $30 million. That's $48 million a year.

Note that Arizona is currently expecting a $600 million revenue shortfall for the fiscal year.



Anonymous Anonymous said...


The point of a pigovian tax, like a cigarette tax, is to reduce consumption of goods that cause negative externalities. The fact that cigarette consumption is falling, and consequently cigarette tax revenues, is a good thing.

October 16, 2007 at 11:48:00 AM MST  
Blogger Scott Gustafson said...

I didn't say it was good, bad or indifferent, just that the expected effect on Arizona tax revenues was negative.

Now you can look at what those revenues fund and decide if their benefit is greater than the benefit of what the federal tax would fund plus any possible benefits from reduced cigarette consumption.

From a purely economic standpoint, I'm persuaded that cigarette smokers provide a net benefit. They pay a lot of taxes, die earlier, and consume fewer public benefits in retirement than their healthier counterparts.

Note that I don't smoke and plan on living long enough to be a problem to my children.

October 16, 2007 at 8:46:00 PM MST  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home