# Counting On Reliable Statistics

Folk wisdom tells us there are "lies, damn lies and statistics." Simply put, statistics are numbers arranged to provide useful information, but using the wrong numbers or misinterpreting their meaning can cause problems. Like other tools, the fault lies not with the tool itself, but with the person who uses it.

Some of my most frustrating moments in the solid waste field have been trying to get usable statistics. Sometimes the problem was lack of information, and sometimes the wrong information was gathered either through ignorance, or because it yielded favorable results.

Once I wanted to know how many automated containers required repair so I could estimate how many extra containers to order. I contacted cities that used automated collection and asked how many containers re-quired monthly and yearly repairs expressed as percentages of the to-tal containers owned. Some respondents had no records. One told me that he guessed "lots" of his containers were repaired but could not be exact. One large city had no hard data but estimated .5 percent per year as a reasonable guess. None I contacted had a reliable figure.

At the time, Los Angeles was buying 1.4 million containers at a probable price of \$70 each. If I applied the estimated one-half of one percent to my operation, 7,000, a total of \$490,000, would have to be purchased as extras. Moreover, each one percent deviance in that figure meant 14,000 containers and \$980,000. This estimate does not include the cost of storage facilities, personnel, administrative costs and transport vehicles. These figures made me thankful that I wasn't a taxpayer in any of those cities.

In today's troubled economic times, many are beginning to re-examine the costs of doing business. Hard figures on comparable or alternative operations are useful tools. Unless you compare similar operations on a level playing field, comparisons often can be misleading.

Estimating costs demonstrates how statistics are gathered and used to predetermine the outcome, and makes comparison difficult. Everyone defines costs differently, and it is difficult to find a standard list of item costs that everyone can agree on. To some, the cost of land may mean the price paid. To others, it includes lost revenue from the land not being used. Still others add in depreciation of structures, while loss of interest on mortgage money may also be included. Obviously, an agreement on the definition of terms is essential in statistical analysis. To compare apples to apples, you must define an apple with great precision.

Numbers are concrete elements; they always have the same value. Numbers comfort people who need consistency. For those of us who have to interact with, and rely upon, people, weather and other less reliable factors, statistics are only indicators, not absolutes. Knowing this, we use statistical probability. Let's say 75 percent of a city's residents are happy with the refuse collection system in place. That number won't tell us whether a new system will work better, or even how well it will be accepted, but it is one factor to let us know the public probably will need a good reason to change.

In this age of electronic information gathering and dissemination, it is important to remember that decisions still have to be made by people. Machines cannot differentiate between valid information and speculation. More information doesn't always mean better information. People still have to decide what data to keep, what form to keep it in and more importantly, what it all means.

Industry representatives need to discuss issues of uniform information gathering. I would like to see information on the number of homes served vs. the number of homes collected. One tells you what your customer base is; the other tells you how many trucks and personnel you need. They are not interchangeable.

In recycling, set-out rates are used to determine equipment and personnel needs, and participation rates are used to measure program success. Some think these rates are the same thing. A standard measure for the rate of efficiency would allow for comparison between operational techniques and jurisdictions. Current indicators include the number of hours needed to collect a ton of refuse, or the number of containers collected per hour or the number of homes passed or all three.

Why do we need uniformity? If I am considering changing my method of operation to become more efficient, I want to make that decision based on factual, apple-to-apple comparisons, not apples-to-oranges. Many millions of dollars, as well as jobs and reputations, are based on statistics that have no real relationship to each other. Comparative statistics that are actually incomparable support decisions that affect the lives of millions of people every year, and decide the fate of multi-million dollar programs as well.

Send questions about your operations to Knapp at 3336 Vista Rocosa, Escondido, Calif., 92029; fax 619-740-9177; or call 619-741-5349.