Economic impact

What factors help us understand trends in waste generation and disposal? Old standard metrics of raw tonnage and per-capita rates of disposal and recycling have been useful but only scratch the surface of understanding the dynamics of waste, especially in relation to other relevant economic data. Clearly, economic activity affects the discard stream, but what kinds of economic activity matter most?

This article takes an econometric approach in analyzing one state’s disposal history and shares observations on how various factors stack up statistically to waste trends. Sources of the data used for this analysis include the North Carolina Department of Environment and Natural Resources (DENR) Division of Waste Management (DWM) Solid Waste Management Annual Reports, the U.S. Bureau of Labor Statistics, the U.S. Federal Reserve Economic Data and Log Into North Carolina (LINC). An econometric analysis reveals that disposal is driven primarily by industrial production, in particular construction and tourism, as well as by income. Three variables that help to explain disposal rates include tourism payroll, percent urban population and the number of local landfills.
 

Unmet goals

In 1991 the North Carolina General Assembly amended the state’s comprehensive solid waste management act and established goals on a per-capita basis. This change was made to officially recognize that disposal rates are driven in part by population growth. The number of tons disposed per person in fiscal year 1991-1992 was 1.08, and the state’s goal was to reduce that amount by 25 percent by 1993 and by 40 percent by 2001 (or to 0.65 tons per person).

Neither of these goals was met. In fact, by fiscal year 1993- 1994, only a 6 percent reduction in the per-capita disposal rate had been realized, and, by fiscal year 2000-2001, the per-capita disposal rate had actually increased 10.3 percent from fiscal year 1991-1992. In recent years, however, per-capita disposal declined from 1.33 tons per person in 2006 to 0.94 tons per person in 2013, a 29 percent decrease over seven years.
 

A closer look

After failing to meet the 1993 goal, solid waste professionals across North Carolina were perplexed about how to reduce the amount of waste disposal. In response to the dismal outlook, the North Carolina DENR Division of Environmental Assistance and Customer Service (DEACS) presented some observations to suggest that the state’s progress toward its goals was actually better than it appeared at first glance. The primary observation was that it is necessary to account for factors outside of the control of waste disposal and recycling policies and programs when analyzing their effectiveness. The underlying reason for this is because it is inappropriate to compare “apples” with “oranges.” In other words, an inherent problem in comparing data from 1991 with 1993 or 2001 is that the underlying economic conditions are different. Therefore, just as the North Carolina General Assembly amended waste disposal goals to account for changes in population, we also needed to control for changes in the economy when analyzing progress made toward achieving solid waste goals.

A discussion of the long history of academic research investigating the links between economic and social factors and the quantity of waste disposed is beyond the scope of this article. However, one finding that is robust to both location and time is that for each 1 percent increase in income, waste generation increases by less than 1 percent.

Some previous studies also have investigated the effect of household size, age composition, education levels and other economic factors on household waste generation. Johnstone and Labonne provide a concise literature review of the economics of solid waste generation (Johnstone, Nick and Julien Labonne; 2004; “Generation of Household Solid Waste in OECD Countries: An Empirical Analysis Using Macroeconomic Data,” Land Economics 80[4]: 529-538). This analysis also finds a strong correlation between waste generation and per-capita income.

The correlation coefficient indicates the degree of linear dependence between two variables and can be used as a starting point for identifying drivers of disposal. The variables most highly correlated with the percentage change in per-capita disposal tons are related to the construction industry. In other words, the percentage change in construction employment, earnings and income closely follow the same path as the percent change in per-capita disposal tons. This relationship makes intuitive sense. We can expect more construction and demolition waste when there is more construction activity. Thus, construction activity increases per-capita disposal holding all other economic variables constant.

Another industry that seems to play a large role in describing the amount of per-capita disposal is tourism. It is easy to imagine how tourism may increase the amount of municipal solid waste generated. Consumers’ purchasing and consumption habits are likely different between vacation and home, and individuals also may be unfamiliar with local practices and discard materials they might otherwise choose to recycle. Tourism may be one explanation for such a seemingly high per-capita disposal rate in Dare County, a popular beach destination. However, it should be noted that Dare County does not seem to be an outlier with respect to tourism variables. One economic variable alone cannot explain all the trends in per-capita disposal around the state. In fact, other counties around the state, such as Wake (home to Raleigh, the state capitol) and Mecklenburg (home to Charlotte, the state’s largest city) spend the most on tourism payroll and generate the most state and local taxes from tourism.
 

Little correlation

It also is interesting to note that there is very little correlation between disposal and recycling in the state. When disposal increases 1 percent per capita, recycling only increases 0.2 percent. This may suggest that as income increases around the state, we cannot rely on the current state of the recycling industry to curb disposal rates. To reduce the amount of disposal, it will be necessary to understand and change people’s behavior.

The analysis so far has revealed that statewide, per-capita disposal is driven by industrial production and income from the construction and tourism industries in particular. To predict and control disposal, it will be useful to establish a more detailed relationship between disposal and various economic indicators. Using county-level census data from 2000 and 2010 obtained from LINC combined with per-capita disposal rates, it is possible to establish whether a simple linear relationship exists between per-capita disposal and economic indicators. A linear relationship is only one possible specification for the relationship, and analyses conducted under different specifications may yield different results.

The purpose of this article is to simply shed light on drivers of per-capita waste disposal and it is beyond this scope to identify the correct relational specification and develop a flawless model.
 

Variable effects

Except for a few variables, the estimated coefficients are of the expected sign, but some variables are better at explaining differences in disposal rates than others. The income / wealth variables, production variables, consumer characteristics and landfill variables are all correlated with annual county per-capita disposal as economic principles would suggest. Unfortunately, the housing variables overall are not very helpful in explaining variation in per-capita disposal for this sample of annual county observations. While all of the production variables are positively correlated with disposal rates, a deeper investigation into the relationships between these variables reveals that they are all positively correlated with the income / wealth variables. In other words, many of them lose their explanatory power when combined with other production or income / wealth variables.

The landfill variables yield some interesting results. When considered by themselves, all of the landfill variables except the number of tire landfills are positively correlated with per-capita disposal but explain only about 5 percent or less of the variation in per-capita disposal. However, when considered along with other income / wealth, production or consumer characteristic variables, the sign of the estimated coefficient flips and becomes negative. Such a relationship may perhaps be explained by a “not in my own backyard” mentality, where consumers who live near landfills are more cognizant of what they throw away.

A final model for per-capita disposal is specified as a linear function of the natural log of millions of dollars in tourism payroll, the percentage of urban population, the total number of landfills in the county and an indicator variable to distinguish year 2000 observations from year 2010. After eliminating some outliers based upon standard statistical tests, this model explains more than 62 percent of the variation in per-capita disposal in the county-year observations. The estimated coefficients suggest that: (1) a 1 percent increase in income or production is correlated with an increase of 0.09 tons per capita; (2) a 1 percent increase in the percent urban population leads to another 0.41 tons of disposal per person; (3) each additional landfill in the county reduces per-capita disposal by 0.02 tons per person; and, finally, (4) holding all other variables constant, counties in 2010 disposed of 0.15 tons per person less than in 2000, a decline of 12.7 percent.

The analysis described above has served two purposes. First, it further emphasizes the point made by North Carolina’s Division of Environmental Assistance and Customer Service that change in per-capita disposal needs to be considered relative to what else is happening in the state. Second, it provides a starting point for thinking about how working to decrease per-capita disposal might be best managed in North Carolina.

Considering that the construction and tourism industries seem to be such drivers of disposal, it may make sense to study the costs and benefits of policies or programs that encourage the use of recycled content in new construction or increasing recycling education budgets in areas with lots of tourism or large urban areas. Such policies or programs that appear empirically beneficial to society should then be evaluated for implementation.

Chris Giguere is a doctoral candidate in economics at North Carolina State University. He is working on his dissertation, “Environmental and Resource Policy in North Carolina: Vehicle Emissions, Disposal and Recycling,” and has worked with the North Carolina Department of Environment and Natural Resources Division of Environmental Assistance and Customer Service. He can be contacted at giguerecs@gmail.com.