Exploring the seasonality of cigarette‐smoking behaviour

Seasonality has become a factor in the once‐stable tobacco industry … with so many indoor smoking bans right across the country; smokers have more chance to smoke in the warm‐weather months.—John Barnett, CEO of Rothmans Inc

Seasonality has been shown to influence cigarette sales during certain times of the year.¹,²,³ Although seasonality is a relevant issue to tobacco control, little research has explored factors that contribute to seasonality.¹,³,⁴ Some of the proposed reasons for seasonal effects include tax increases, weather conditions and timing of quitting efforts (eg, New Year's resolutions).¹,²,³ Thus, we further examined factors believed to contribute to the effect of seasonality on cigarette consumption.

Monthly cigarette sales were obtained from the New Jersey Department of Revenue for fiscal years 1999–2006, and data on monthly weather patterns were obtained from the office of the New Jersey State Climatologist for the same period. We conducted a stepwise multilinear regression to examine the effect of average monthly temperature, number of days per month with temperatures below freezing, number of days in the month and tax rates on monthly cigarette sales.

Consistent with previous research,¹ we found that in New Jersey, during the time period examined, February had the lowest average monthly cigarette sales (240 million),⁵ whereas June had the highest average monthly cigarette sales (329 million).⁵ It is important to note that New Jersey raised the state cigarette excise tax three consecutive times; each increase coincided with the beginning of fiscal years 2003, 2004 and 2005. Thus, high rates in June and lower rates in July (292 million)⁵ may be attributed to wholesale distributors “hoarding” cigarettes in June in anticipation of higher prices in July when new tax increases take effect.

All correlations between monthly cigarette sales and the predictor variables were significant; the strongest correlation was for excise tax rate (r = −0.826, p<0.00). In the regression, cigarette excise tax rate remained the strongest predictor of cigarette sales, followed by the average daily temperature and the number of days in the month, which were also significant (table 1).

Table 1 Multiple linear regression for monthly cigarette tax stamps or sales, 1999–2006.

Predictor variables	Outcome = monthly cigarette sales
	Unstandardised β	Standardised β
Intercept	2574105.8
Average daily temperature	123746.2	0.261**
Excise tax	−8721370.0	−0.825**
Number of days in the month	1280219.4	0.142**
		Adjusted R2 = 0.779

**p<0.00

These results further support the notion that the recent increases in indoor air restrictions, which force smokers outdoors to smoke,⁶ may contribute to the seasonal changes in cigarette‐smoking behaviour. Our findings point to the importance of controlling for the number of days in a month when examining seasonality. Previous research suggested that February, a winter month in the northern hemisphere, had the lowest cigarette consumption, but February also has the distinction of being the month with the fewest days. Thus, February's status as the month with the lowest cigarette sales¹ may be because of the weather conditions and also because it has fewer days for distributors to purchase tobacco. Although these data are limited to New Jersey for the period 1999–2006, these results are consistent with previous research¹,²,³,⁴ and add to the evidence that cigarette‐smoking behaviour has a seasonality component.