"Stick" Works Better than "Carrot"

Is there a study (or a strong argument) showing that "sticks" are more effective at changing behavior than "carrots?" This study could be for any sort of behavior change, not just for commuting behavior. Here is some background for my query: A "parking cashout" is where an employer pays employees not to park at the work place. One Bay Area employer (who we will call "Employer X") has a severe parking shortage, so pays employees $4 per day to not park. Cashout is a "carrot," a benefit used to motivate commuting behavior change. Employer X's $4 per day cashout program has reduced SOV (single occupancy vehicle) commute mode share from 74% to 70%. Before cashout, Employer X had 26% green commutes (walk, bus, train, telecommute, carpool, etc). To implement cashout, Employer had to pay the existing 26% green commuters $4 per day (out of fairness) before motivating 4% new green commuters. Hence, the cost per day per new green commute at Employer X is $30 per day. Unfortunately, this is a very ineffective way to reduce SOV commuting. (Employer X could fund lease payments so that each employee could have two Priuses for the same amount per day.) Some historical, discontinued cashout programs (from 1995 and before, not in Bay Area) were more effective, but cashout has rarely been tried in modern, auto-centric, suburban office settings. The recent findings of Employer X are not encouraging for Bay Area suburban office cashout. "Parking charges" are where employees have to pay to park at the work place. Charges are a "stick," not a carrot, changing behavior by penalizing SOV commuting. High parking charges (combined with good transit and bad auto traffic) in downtown San Francisco dramatically reduce SOV commuting. SOV commute mode share is less than 50% in San Francisco, but SOV commute mode share at the Bay Area's 17 large suburban-style office parks (more than 594,000 employees) with free parking is greater than 80%. Irritant theory We have an unproven "pop psychology" theory as to why cashout is not very effective. High-paid office workers are willing to ignore small benefit programs such as $4 per day cashout. This carrot is not a sufficiently large motivator to cause commuting behavior to change. Employees will not think about the cashout on a regular basis. We believe that parking charges will "irritate" SOV commuters. These SOV commuters will think about the parking charges every day they commute. Eventually this irritant gnaws at them long enough to cause many to change behavior. Changing commuting mode choice is a significant decision because of relatively high barriers to changing away from the convenience of driving alone. This difficult decision is not a "snap decision" and may require pondering over many weeks. We believe that the same $ value of irritant/stick has a much higher impact than the same $ value of cashout/carrot. A 1989 academic paper ("Parking Subsidies and Commuter Mode Choice: Assessing the Evidence," by Richard Willson, Donald Shoup, and Martin Wachs) reinforces the idea that carrots are less effective than sticks: "A program of transit and vanpool subsidies as well as preferential parking for carpoolers had little effect until [Twentieth Century Corporation in Los Angeles] raised the price of employee parking from no charge to $30 per month for solo drivers. Solo driving decreased from 90 to 65 percent after pricing."

- Steve Raney,
Cities21, Palo Alto, CA