A factor-analytic logit approach to truck driver turnover

Journal of Business Logistics, 1995 by Richard, Michael D, LeMay, Stephen A, Taylor, G Stephen

Less-than-truckload carriers in the United States must devise strategies for retaining truck drivers in an environment that is increasingly competitive on all fronts. Customers demand low rates and high service levels. White males, the traditional truck driver labor pool, account for a diminishing portion of new entrants into the labor force.(1) Over the next decade, 50% of the drivers for the unionized, LTL carriers will probably retire.(2) The U.S. labor force is growing at the slowest rate since the 1930s, at approximately 1% per year.(3) This means that most companies will not be able to afford to "buy" drivers from other fields. It also means that drivers have many potential employers from which to choose.

In order to retain drivers, then, trucking managers must first understand why drivers leave one company to go to another or to another career. Trucking managers often attribute driver turnover to low pay or insufficient time home.(4) While these variables certainly influence drivers' loyalty to the company or to the profession, they are not necessarily the primary causes of driver turnover. Attributing turnover to pay and time home does one very important thing, however. It frees managers from having to make difficult changes in the company. The nature of truckload operations, as traditionally practiced, prevents most drivers from getting home on a regular basis. Moreover, current freight rates do not allow significant wage increases. As these matters are beyond the manager's control, driver turnover is seen as an insurmountable problem that carriers simply must learn to tolerate. However, problems with pay and time home do not account for the fact that most drivers leave one over-the-road (OTR) trucking job for another, usually one that offers the same level of pay and the same amount of time home.

So why do drivers choose to leave? This paper illustrates the use of a combination of quantitative techniques to identify the reasons why truck drivers leave a particular company. Similar techniques have been used to model choices of mode, route, destination, and carrier using nominally scaled, binary dependent variables. The most common of these models is the LOGIT model.(5) These models use sets of explanatory variables to predict the probability that an individual will choose a particular outcome.(6) The explanatory variables constitute a utility function that influences the probability of choosing a particular outcome over other available outcomes. The purpose of this paper is to present and empirically test a model of driver turnover behavior that may assist managers in controlling turnover. This study deals with a variety of variables that management can influence, regardless of general economic conditions or the financial condition of the firm.

A second purpose is to demonstrate an approach to choice modeling that alleviates some of the problems associated with multicollinearity. The approach combines Factor Analysis with LOGIT to predict probability of choice. The binary choice for the driver is either to leave or remain with the current employer. Results of this Factor-Analytic LOGIT approach are compared to those obtained by a LOGIT-only model. Thus, this study attempts to find practical solutions for managers, while also addressing technical problems sometimes associated with the conduct of academic research.

DRIVER TURNOVER IN THE MOTOR CARRIER INDUSTRY

Driver turnover in the truckload industry, defined here as the voluntary movement of drivers from one carrier to another, has proven problematic for managers. Indeed, with annual turnover rates often exceeding 200%,(7) the loss of drivers greatly increases a carrier's cost of doing business. To fully appreciate the magnitude of this rate of turnover, the median 1992 turnover rate for all U.S. companies was 8.4%.(8) While the turnover rate for all trucking firms has been calculated at 38%,(9) that number was heavily influenced by lower levels of turnover in large, less-than-truckload firms.

The constant loss of drivers significantly increases a carrier's cost of doing business through higher recruiting costs, excess paperwork, drug tests, and physical examinations caused by new drivers constantly being added. Drivers with questionable qualifications are likely to be hired simply to keep a tractor on the highway. This can lead to increased accidents, abandoned trucks, lowered morale among other drivers,(10) as well as conflicts with shippers and receivers. Such conflicts can prove very costly, for as Beier(11) has noted, cost-savings come only from stable, long-term relations among shipper, receiver, and carrier. Drivers can either enhance or tarnish this relationship.

As already noted, the conventional wisdom among trucking company executives is that higher pay and more time home would stem the flow of drivers through the firm and through the industry.(12) If these reasons are in fact the cause of such excessive turnover, then there is little management can do to address this problem. Unfortunately, because of the nature of irregular route motor carriage, network management to assure more time home is rarely feasible. The highly competitive nature of the irregular route motor carrier industry also makes significant pay increases unlikely, at least for most firms. Moreover, even if increasing pay were economically feasible, this in all likelihood would pressure other carriers to follow suit. The result would be more expensive turnover, but probably not lower turnover. Two large firms recently raised mileage rates five cents per mile. Management reported short-term improvement in turnover, followed by a gradual slide back to the original turnover levels.(13)