Over at Crosscut, Knute "Skip" Berger weighs in on both sides. First, siding with the classic right-wing argument, he argues that tolls constitute oppressive social engineering---that, in effect, using tolls for any other purpose than to pay for the road on which they're levied constitutes an attempt to "penalize drivers and force them into transit or staying home." Second, from the left, he argues that tolls are a libertarian solution that penalizes people who can't afford to pay: Unlike the progressive income tax, tolls are regressive because everyone pays the same amount, regardless of their means. It's the Tim Eyman-Mike McGinn double punch: Tolls shouldn't be used to fund non-road improvements, and tolls are unfair to the poor.
So how should we fund amenities we want, like light rail on I-90, improvements to surface streets, and capital spending on transit corridors? The conventional options have been sales tax, gas taxes, user fees like the vehicle license fee, or money from cities' general funds, which are funded largely through property taxes, and sales taxes. Of those, sales taxes are the most regressive taxes of all (because poor people spend the largest portion of their incomes of any group); gas taxes are also regressive (everyone pays the same for gas regardless of income); flat user fees, as flat taxes, are regressive; and property taxes are arguably regressive as well (the little-old-lady argument: As property values go up, taxes go up, forcing little old ladies to sell their homes).
Given that most of the available taxes are regressive (and the state legislature isn't likely to pass an income tax), are tolls inherently worse than any other regressive tax or fee? Or, more to the point, should government services (like transit) be funded by fees on things other than those services (like tolls on bridges)?
I think it's perfectly fair to charge tolls, and to use those tolls for purposes other than the roads they're levied on, for two reasons. First, social engineering is a (perhaps the) fundamental role of government. For decades, that social engineering has been aimed at getting people to drive as much as possible (by making driving the cheapest, most convenient option), but that's no more inevitable than if governments were to invest huge amounts of gas-tax dollars in transit and bike infrastructure to encourage those alternatives instead, as they did in Copenhagen. If you support taxes on cigarettes, laws against drunk driving, building codes that keep your family safe, or food-safety requirements, you support social engineering.
Second, and perhaps more importantly, tolling is only "regressive" if people have no other choice than to drive across a tolled road. But they do have other choices---and tolling can help expand those choices. If I want to avoid the tolls at rush hour on 520, I have options: I can take one of several alternate routes, use transit (Metro and Sound Transit both offer frequent bus service across the bridge), combine trips or reschedule my trips for times when tolls are lower, telecommute, or ride my bike across I-90 (or, eventually, the new 520 bridge) instead. All these options won't work for everyone (alternate routes, for example, may be slower, especially at rush hour), but nearly everyone has at least one alternative other than driving alone at rush hour and paying the highest-level toll.
Moreover, collecting tolls will provide more options. And the more tolling you do, the more options you create.Assuming Eyman's I-1125 doesn't pass (and, admittedly, that's a big assumption), tolls on 520 and its parallel bridge, I-90, can be used to fund transit service across both bridges, giving people who can't afford to (or don't wish to) pay tolls yet another alternative. You can call that social engineering. Or you can call it good planning for a future where transit, biking, and other "alternative" modes of travel are even more attractive and accessible than driving.