Elsevier

Journal of Economic Psychology

Volume 58, February 2017, Pages 77-88
Journal of Economic Psychology

Goal attainability and performance: Evidence from Boston marathon qualifying standards

https://doi.org/10.1016/j.joep.2017.01.001Get rights and content

Highlights

  • Clear, challenging, and attainable goals prove important to results.

  • Marathon runners improve as Boston marathon qualification becomes more attainable.

  • Increasing the difficulty of the goal worsens performance.

Abstract

In this paper we test if performance improves once goals become more attainable. Goal-setting literature suggests that workers respond to challenging but achievable goals with increased performance. Empirical evidence supports the notion of goals increasing performance; however the evidence on how attainability of goals affects performance is mixed. This paper tests whether efforts increase, improving performance as the goals become more attainable. We are employing a unique set of publicly available marathon data from 1970 to 2015 to directly analyze the effect of goal attainability on performance. With the probable goal of qualifying for the Boston Marathon, we test if runners increase their effort, and consequently improve their performance if they enter a new age group and as a result have a more attainable goal. We find that runners who enter a new age group perform better than the runners whose qualifying time did not change. This effect is seen with runners in younger age groups, but not found in the results of runners in more advanced years.

Introduction

Meeting objectives requires measuring productivity in a quantifiable way. Whether creating products, taking a test, or losing weight on an individual level, concrete goals are often set in advance to establish results to work toward. This creates an objective that can be achieved, rather than a vague statement of purpose like “increased profitability” or “losing some weight.” Striving for these goals affects every facet of life, but is setting a specific goal all that is needed? While goals are intended to increase productivity and improve results, setting a goal that is beyond reach could have little or no effect. In this paper we address the importance of the attainability of a goal, and its contribution to the likelihood of achieving the desired results by observing the race day performance of marathon runners. We exploit the availability of marathon data from 1970 to 2015 to analyze the effect of a goal, the Boston marathon qualifying (henceforth known as BQ) time standards, on performance. The unique and key feature of the data is our ability to track individual marathon runner performance over years along with changes to BQ time standards, allowing us to isolate the impact of goal attainability on performance.

Goal-setting and its impact on effort and performance has its origins in psychology. Locke (1968) was among the first to create a theory which states the setting of a challenging yet achievable goal increases intrinsic motivation and as a result improves performance. Latham, Mitchel, and Dossett (1978) examined Locke’s theory in workplace production and confirmed the link between goal-setting and workplace performance. When describing the characteristics of successfully set goals, Locke and Latham (1985) suggested that in order to boost performance, goals must be specific, challenging, and also attainable and/or realistic. They hypothesized that unrealistic and unreasonable goals will result in failure, which will lead to a drop in motivation and lower overall performance.

Several empirical studies have supported the idea that goals need to be specific and challenging, and that easy, non-specific goals do not provide enough motivation (Goerg and Kube, 2012, Mento et al., 1987, Smith et al., 1996, Wu et al., 2008). Locke and Latham, 1990, Locke and Latham, 2002 review the extensive amount of laboratory and field studies and in a large amount of the reviewed articles, they find either strong or mostly strong support for the idea that setting challenging goals improves performance more than setting less challenging goals.

Recently, studies have examined the specific impact of goal attainability on performance. More specifically, these studies focus on the idea that if goals are too challenging then a person realizes achieving them is not possible and quickly gives up. So, for goals to elicit the largest gains in effort or performance, they must not be set to be too easy or too difficult, but at a performance level that one may realistically obtain. Heath et al., 1999, Wu et al., 2008 formalized this idea by using prospect theory to explain this result. Once a reference point (goal) is set, people view the outcome as binary, you either fail or succeed in achieving your goal. If one has no chance at succeeding, then effort decreases as it will be wasted on a task certain to fail. If one has a task that is very easy, then effort decreases once the goal is met, even though one may have been capable of greater performance.

Economic theory can be used to further explore the nuances of goal-setting that do not involve monetary rewards and their impact on effort and performance outcomes. Smithers (2015) notes that the motivational tool of goal-setting has been overlooked by economists. Goals are set in many different economic realms in an attempt to achieve desired outcomes, from the principal-agent problem in managing a work place to self-control issues in personal finance. In order to place the contribution of our study, we must discuss the strands of economic literature on goal-setting.

First, we note the difference between binding and non-binding goals. Binding goals refer to situations in which monetary rewards are related to success or failure and are most often seen in the context of the principle-agent problem where monetary rewards are used to elicit effort from workers (Laffont & Martimort, 2002). Alternatively, non-binding goals do not involve monetary rewards. Recent attention has been given to situations in which principals set non-binding goals for agents. Gómez-Miñambres (2012) develops a theory incorporating intrinsic motivation of agents and notes that unattainable goals can be detrimental to performance. Corgnet, Gómez-Miñambres, and Hernán-Gonzalez (2015) use a laboratory setting to analyze the impact of non-binding goals on performance. The authors found support for the notion that goals must be challenging yet attainable.

Another strand of economic literature focuses on the theory of non-binding goals that examine self-control issues (Koch and Nafziger, 2011, Hsiaw, 2013) which are applicable to weight-loss and exercise plans. There is a wealth of weight-loss experiments that are inconclusive as to the impact of attainability on weight-loss success (Grave et al., 2005, Ames et al., 2005, Byrne et al., 2004, Fabricatore et al., 2008, Gorin et al., 2007, Linde et al., 2004, De Vet et al., 2012). While our study is not about self-control issues, the BQ time does serve as a pseudo ‘self-set’ goal, in that an outside agency sets the BQ time, but a runner must decide for themselves to set BQ time as their goal. There have been numerous experiments in the field of athletic performance producing mixed results on the impact of goal attainability on performance (Bar-Eli et al., 1993, Bar-Eli et al., 1997, Garland et al., 1988, Weinberg et al., 1991). The major downside to most athletic performance studies, similar to weight loss studies, is their reliance on experiments which often contain a relatively small number of observations which may not produce robust results. Our study uses real-world runners outside of an experimental setting, which is important as there is a lack of studies on goal-setting outside of a laboratory setting.

To our knowledge, there are only two studies on non-binding self-set goals in real-world situations. Harding and Hsiaw (2014) develop a theoretical model of consumer demand for an energy conservation program that involves non-binding, self-set goals. They present empirical evidence supporting the notion that goals must be realistic in order to motivate performance. In their study, consumers who chose realistic goals for energy conservation consistently saved more, ‘achieving savings of nearly 11% more than those choosing very low or unrealistically high goals’. In a paper which also uses marathon runner data, Markle, Wu, White, and Sackett (2015) find runners with a set goal improve performance by 6 min as compared to those runners that do not have a goal.1

Our paper fills an important gap in the literature on non-binding goals by incorporating the available information about previous performance to help determine and control for the potential for improvement toward a goal. Additionally, our study differs from previous literature in that the goal is not actually self-set, but determined by an outside agency. In this context, the BQ time serves as a symbol of ‘superior’ performance, it has been estimated that only 10% of all marathon runners will run a BQ time in their careers.2 Furthermore, the results of this study may be applicable to standardized testing for college and graduate school admission. For example, students often take the SAT repeatedly in an effort to improve their score and gain admission into ‘elite’ colleges, very similar to how runners repeatedly run marathons in an effort to BQ. As an example, 75% of Harvard’s most recent admissions class scored higher than 2100 on their SAT, in the coming year, high school students with the goal of entering Harvard may use this number as a benchmark for performance.3 A student’s score on their first taking of the SAT would set the level of attainability if 2100 is their goal. Thus their score improvement in subsequent takings of the SAT may depend on how close they were to 2100 on their first attempt.

The results of our analysis indicate that attainability of a goal plays a key role in performance once previous results are controlled for. As the BQ time becomes more attainable, performance improves and in cases where the BQ time becomes more challenging, performance declines. In the section that follows, we discuss in great detail the process for BQ time changes and how this variation is important to identifying our main result.

Section snippets

Context, data, and descriptive statistics

The Boston Marathon, which began in 1897, is the oldest annual marathon in existence and is arguably the most prestigious marathon in the world.4 Due to its overwhelming popularity and capacity constraints, qualifying standards were introduced for runners in 1970. In order to qualify for the Boston Marathon one must run a marathon on a certified course within a specific time limit as determined by the Boston

Methodology

We estimate the effect of a change in goal attainability on two separate models measuring performance. First, we estimate the effect of a change in age group on a runner’s performance during the marathon. The model uses the following general form:Race_timeijt=α0+α1BQ_changeit+α2Ageit+α3Genderi+α4Misseditj-1+α5Experienceij+α6Experience_Sqij+α7Races_in_yearit+α8Running_PRij+α9BQprevi+α10Monthij+ijtwhere Race_timeijt is athlete i’s finish time in year t at marathon j.

Results and discussion

Table 4 provides the results of a random effects least-squares regression. Our dependent variable in each of the specifications is the marathon finish time in minutes for each runner in an individual race. Coefficients on month are excluded for space considerations. Coefficients are reported with robust standard errors.15

In each of the regressions

Conclusions

The goal of this paper is to examine the relationship between realistic goals and performance. Using the data from numerous different marathons and analyzing the differences in performance of runners over the course of many years, we find support for the hypothesis that performance improves as a goal becomes more attainable, if past performance is properly controlled for. We see that when qualifying standards get easier, it has a positive effect on performance and that in cases in which the

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References (32)

  • M. Bar-Eli et al.

    Effect of goal difficulty on performance of aerobic, anaerobic and power tasks in laboratory and field settings

    Journal of Sport Behavior

    (1993)
  • M. Bar-Eli et al.

    Effect of goal difficulty, goal specificity and duration of practice time intervals on muscular endurance performance

    Journal of Sports Sciences

    (1997)
  • B. Corgnet et al.

    Goal setting and monetary incentives: When large stakes are not enough

    Management Science

    (2015)
  • E. De Vet et al.

    Ain’t no mountain high enough? Setting high weight loss goals predict effort and short-term weight loss

    Journal of Health Psychology

    (2012)
  • Grave, D. R., Calugi, S., Molinari, E., Petroni, M. L., Bondi, M., Compare, A., Marchesini, G., & QUOVADIS Study Group....
  • A.N. Fabricatore et al.

    Weight loss expectations and goals in a population sample of overweight and obese US adults

    Obesity

    (2008)
  • Cited by (12)

    • Energy saving in a simulated environment: An online experiment of the interplay between nudges and financial incentives

      2021, Journal of Behavioral and Experimental Economics
      Citation Excerpt :

      Achieving goals also increases intrinsic motivation by leading to a sense of self-achievement (Gómez-miñambres, 2012). The effectiveness of goal setting has been widely documented in a variety of domains, such as performance (Burdina et al., 2017; Lent & Souverijn, 2020), healthy lifestyle (King et al., 2013; Samek, 2019), and energy conservation (Harding & Hsiaw, 2014; Loock et al., 2013). Hence, we hypothesize the following:

    • THE PSYCHOLOGY OF RUNNING

      2023, The Psychology of Running
    View all citing articles on Scopus
    View full text