There are many imporant differences between small and large water systems, and between rural and urban areas.

Common sense tells us that there are fundamental differences between small and large water systems, and between rural and urban areas. Understanding these differences should provide a foundation for addressing policy issues that will arise about the regulation of small water systems and ways to provide safe, reliable and affordable water service in rural communities.

Communities in rural areas tend to have limited financial resources. Photo by Paget-Clark.

New Definitions

Initially, a caution is in order. The U.S. Census Bureau is in the process of changing the definitions of "rural" and "urban" areas. These new definitions will provide, for the first time, uniform definitions throughout the country. They also will result in changing the label placed on some communities; that is, some historically "rural" communities will be classified as "urban" and vice versa.

The federal government has adopted new criteria to define the extent of a "metropolitan area." A metropolitan area can contain both urban and rural communities, but metropolitan areas tend to be much more heavily populated than non-metropolitan areas. Most of the people in the United States live in metropolitan areas, but they occupy just a small fraction of the total land area in the country. Simply, about 80 percent of the people live in a metropolitan area, but they occupy just 20 percent of the land area of the country.

The smaller the population density (the fewer people per square mile), the more expensive it is to build a water distribution network to connect the same number of people (assuming comparable geography, etc.). For example, the Environmental Protection Agency (EPA) reports that water systems that serve 100 people or fewer have an average of 33 customers per mile of water pipe, while systems that serve more than 10,000 people have an average of 71 customers per mile of pipe. Interestingly, this clustering of the population also has other important political and societal implications. For example, of the roughly 3,100 counties and similar political subdivisions in the country, only 854 are located in metropolitan areas, but they are occupied by 80 percent of the population (227 million people out of the 281 million people counted in the 2000 census).

We know most U.S. water systems are small. EPA reports that, as of the end of 1998, there were approximately 54,000 water systems in the country, but more than 32,000 of them served 500 or fewer people. An additional 14,000 water systems served between 500 and 3,300 people. In other words, there are more than 46,000 water systems (85% of the total) that serve fewer than 1,000 customers. In fact, the total population served by all of these small water systems is just 25 million people - only 10 percent of the population served by all water systems in the country. Many of these small water systems are not located in rural areas. Rather, they may serve a single residential subdivision in a suburban area, a mobile home park outside of an urban area or a small community that is just a few miles from an urban center. Other small water systems, though, serve fairly isolated rural communities. The distinction among these types of water systems is important for several reasons. For example, in some locations, it is economically feasible for small water systems to physically interconnect with a nearby water system or to have the operator at a nearby system take over the operations of the small system. In rural areas, though, it is far less likely that there will be another water system close enough for physical interconnection to be feasible (shared operations can be feasible over large distances if electronic communications are used to relay data about key pieces of equipment, such as pumps, chlorinators and storage tanks).

Getting Connected

One study that examined the potential for physical interconnection and satellite management in 17 states found that the results varied tremendously from one state to another, ranging from 8 percent of small systems to nearly 50 percent of small systems in a state having the potential to physically interconnect with a larger water system. There is no single statistic that would automatically tell you whether a state's small water systems have the potential to interconnect. The study concluded, though, that nearly all of the small water systems in the states that were analyzed had the potential to become part of a satellite management operation with a larger water system.

EPA has compiled some basic economic statistics to compare water systems of different sizes on a national level. These data indicate that small water systems (those serving fewer than 3,300 people) tend to have lower levels of income than larger water systems, at least comparable levels of poverty and much lower property valuations. All of these factors would indicate that, on a national level, customers of smaller water systems tend to have fewer financial resources than customers of larger water systems.

There are dramatic differences between the communities served by water systems in metropolitan areas as opposed to those served by water systems in non-metropolitan areas. Household income is between 15 percent and 30 percent lower and poverty rates are 30 percent to 40 percent higher for communities served by water systems in non-metropolitan areas than by those served in metropolitan areas. It appears that the real difference between water system demographics is not one based on the size of the water system, but one that is based on whether the system is in a metropolitan area. Income levels and poverty rates for water systems in metropolitan areas are fairly consistent, regardless of system size. But there is a dramatic difference - again regardless of system size - between systems in metropolitan and non-metropolitan areas. It is important to focus on the rural or urban character of a water system, and not simply its size. This is important not only because of differences in geography (distance between communities and population density), but also because of what appear to be significant differences in income and poverty levels.

Data collected by the Census Bureau shows that the median income level in non-metropolitan areas is (and has been) consistently below incomes in metropolitan areas. The median household income in non-metropolitan areas has been consistently about 75 percent of the income in metropolitan areas. National income data also show that not only is the median income different in metropolitan and non-metropolitan areas, but these differences exist throughout the income distribution curve. Thirty percent of non-metropolitan households had incomes that were less than $20,000, while only about 22 percent of metropolitan households had incomes at the same level. Moving up the curve, approximately 10 percent of non-metropolitan households had incomes above $80,000, but almost 25 percent of metropolitan households had incomes of $80,000 or more. There are very real differences in incomes between metropolitan and non-metropolitan areas - not just at the median income, but over the whole range of incomes. Non-metropolitan areas have a much higher percentage of low-income households than metropolitan areas. Moreover, non-metropolitan areas do not have as many high-income households to provide support for community services.

About 80 percent of Americans live in a metropolitan area, but these areas occupy only 20 percent of the land area.

Discretionary Income

The Bureau of Labor Statistics conducts surveys of consumer expenditures each year. The Consumer Expenditure Survey reports results for various categories of expenditures and, importantly, presents the results separately for urban and rural households. It should be noted that the most recent results of the survey are for 1999 and, of course, do not use the Census Bureau's new definitions of urban and rural areas. Data from the survey show some interesting differences between urban and rural households. For example, urban households have incomes that are 40 percent higher than rural households ($45,597 vs. $32,414), yet urban households spend only 20 percent more than rural households ($37,905 vs. $30,831). The effect is that rural households spend all of their after-tax income to meet their needs (expenditures of $30,831 vs. after-tax income of $30,708), while urban households are able to save a significant portion of their after-tax income (expenditures of $37,905 vs. after-tax income of $41,741). Simply, rural households are living from paycheck-to-paycheck, and even then are barely meeting their basic needs, while urban households are able to meet their needs and still have money left over for savings and investment.

The expenditure patterns themselves also disclose some important differences between urban and rural households. Urban households spend substantially more on shelter (mortgage or rent) than rural households, but interestingly their expenditures on utility services are the same. In fact, despite the dramatic differences in incomes and overall expenditures, rural households spend more for electricity and home heating than urban households ($1,373 vs. $1,225). This means, of course, that rural households are spending a much higher proportion of their income on electricity and home heating than are urban households (4.2% vs. 2.7%).

While the expenditures on water are higher in urban areas, the survey counts every household with a private well (or that otherwise does not pay a water bill directly, such as apartment dwellers in many locations) as having an expenditure of zero. Therefore, the Consumer Expenditure Survey should not be used to draw conclusions about differences in water costs between urban and rural households.

The bottom line is that the typical rural household is spending all of its after-tax income. Any increase in water costs (or any other expenditure category) is going to force a reduction of expenditures in other categories. While some other categories may be viewed as "luxuries" (for example, entertainment, charitable contributions and tobacco), the fact is that the typical rural household will be forced to make a choice between competing needs, while the typical urban household will not be forced to make the same choice (because its expenditures are nearly $4,000 less than its after-tax income).

Very little research has been done into the differences in the cost of water in rural areas as opposed to urban areas. Most water rate surveys (for example, the biennial survey conducted by Raftelis Financial Consulting or AWWA's Water:\Stats project) focus on larger water systems. More comprehensive data collection efforts, such as EPA's Community Water System Survey (last conducted in 1995), present results by system size rather than by system location.

Water Costs

EPA's characterization of small water systems, based primarily on the Community Water System Survey, does not distinguish between rural and urban systems (or even attempt to distinguish among systems by state or region). That effort also does not separate revenues by the type of customer served (residential, commercial, etc.), making it difficult to compare small water systems with larger ones. For example, that report shows that the larger the water system, the greater the revenues per service connection. But by failing to break down that figure between residential and non-residential customers, it is not possible to analyze whether the difference is the result of differences in costs, rate structure philosophies, or simply the presence of more commercial and industrial customers in larger systems.

More Research

The 2000 census should provide researchers with more current information about consumer expenditures for water in every municipality in the country. However, a full release of housing-related information is not scheduled until the summer of 2002. Further, even the summer 2002 release may not incorporate the new definitions of "rural" and "urban." The Census Bureau indicates that a release later in 2002 may be the first that will incorporate those new definitions. In addition, if the census 2000 products are similar to those produced from the 1990 census, the only information on the cost of utility services, including water, will be in the Public Use Microdata Sample files (surveys of 1% and 5% of all households in the country that received the census "long form"). These files are not scheduled for release until the 2002-2003 time frame. Given the currently available data, therefore, it is not possible to draw any reliable conclusions about differences in water prices between urban and rural areas, and this appears to be an area that should be investigated further.