Archive for December, 2018

Energy policy and referenda

December 1, 2018

I feel the need to rant about referenda today.


So what the hell is Taiwan’s energy policy supposed to be now?


Last Saturday, voters passed the referendum #16, commonly labeled as “go nuclear to go green” (my clumsy translation), which deleted a clause in one of the laws setting the phase-out date for nuclear power. (At least I think that’s what it did. I’m actually not sure, which, eventually, is the point.) They also voted overwhelmingly against the Shen-ao coal-fired power plant project (which the government had already cancelled). And Taichung voters elected KMT candidate Lu Hsiu-yen in a landslide, partially because she campaigned on the poor air quality caused by the huge coal power plant. She further promised to stop sending electricity generated in Taichung to northern Taiwan.

But while those are the most recent results, we also have to think about previous lessons from public opinion. Five years ago, the government wanted to have a referendum on whether to open the fourth nuclear power plant, but it was so unpopular that KMT legislators weren’t even willing to vote to propose the referendum. Also, after the KMT lost the 2016 presidential election, Eric Chu singled out increases in electricity (and propane gas) prices as one of the major reasons that the public rebelled against the Ma government. Finally, let’s remember how much outrage there was last summer when an accident at one power plant caused one day of blackouts over much of the island.

To summarize, the voters don’t want clearly coal. They definitely don’t want nuclear, or maybe they do. They don’t want any power plants in their neighborhood, and they definitely don’t want electricity generated in their neighborhood to be sent elsewhere. They want low prices, and they absolutely demand a stable supply of electricity.

It should be easy to satisfy all those demands simultaneously. I’m glad we used referenda to clear up this entire matter.


I have three big objections to this attempt to use referenda to decide energy policy. First, voters are not forced to consider trade-offs. None of the proposals suggested that cutting coal power might be possible if an increase in electricity prices spurred less electricity consumption. Voters in central Taiwan were not asked if they supported refusing to send electricity generated in central Taiwan northward even if it resulted in companies in the Hsinchu Science Party (read: Taiwan Semiconductors) being forced to cut production. Trade-offs are exactly what governments do. The Tsai government restarted nuclear reactors that had previously been offline over the protesting screams of its anti-nuclear wing because it was much more afraid of blackouts. Taiwan could cut pollution by using higher prices to suppress demand, but that would be unpopular. It could also cut pollution by retrofitting some of its older coal plants, but that is extremely expensive and it would take a few years. None of the options are ideal. You can’t have everything you want; you have to make trade-offs. Referenda almost never present the question this way.

Second, energy policy takes years to implement. Five years ago, the Ma government bowed to public pressure and shuttered the fourth nuclear power plant. Unfortunately, this left Taiwan’s energy reserves precariously low. The Tsai government has invested heavily in wind power, which is starting to come online now, and it is planning a natural gas facility in northern Taiwan. However, in the meantime, the choice was essentially getting more electricity out of the existing coal plants or the existing nuclear plants. In fact, the government had to do both. It’s a reasonable stopgap measure, given the long-term strategy. Unfortunately, Taiwan’s referendum law has a short-term orientation. Under the December 2017 revision, the thresholds for proposal and passage are ludicrously low. A successful referendum doesn’t necessarily reflect a deeply held consensus in society. It can just as easily reflect a short-term blip in public opinion. A few years ago, nuclear was extremely unpopular. Now, that has faded somewhat, so maybe this year coal is the villain. What if we do this next year and find that public opinion has shifted again? Are we supposed to fundamentally shift energy policy every two years just because 29.8% of eligible voters said yes to some unintelligible question on the ballot?

And that brings me to the third and most basic problem: information. Referenda place extremely high demands on voters to become educated, and there is very little evidence that voters are up to the task. I’ve been reading Democracy for Realists by Chris Achen and Larry Bartels, and they make this point forcefully. Voters simply do not have the time, capacity, or desire to become fully informed on any given question. We all have better things to do. At any rate, division of labor is a hallmark of modern society. Why should we think that society is better off if everyone neglects their other responsibilities (that is, the things that they are good at) and spends months learning about energy policy? That’s crazy. Instead, people find shortcuts. There are usually plenty of people who are happy to advise them how to vote, but that isn’t necessarily good advice. The people with the strongest incentive to give advice are the people who will directly benefit from the outcome. Not surprisingly, referenda tend to favor the wealthy. The promise of referenda is that voters can bypass the disgusting politicians and go directly to the people. Unfortunately, the people have to rely on an even more disgusting set of people (who are actually also politicians in a different guise) for advice.

Achen and Bartels cite a couple of stories vividly illustrate these problems. In one, many counties in Illinois adopted a requirement that any increase in taxes to fund fire departments had to be approved by referendum. In other counties, the local administrators and councils made this decision. Predictably, voters refused to pay higher taxes, and the quality of fire departments in the referendum counties declined noticeably. Training was neglected, equipment became outdated, staffing was thinner, and response times were longer. Wait, maybe that’s what voters wanted. Maybe they were willing to accept worse fire protection for lower taxes. It seems unlikely; most people also want better services, especially when those services involve life and death. However, they did not save money. They paid lower taxes, but they paid higher fire insurance rates. The county administrators understood this, but voters did not. Poorly informed voters made self-harming choices, and this problem, unlike national energy policy, was fairly easy to understand. Voters are ALWAYS underinformed.

But what if some voters could be fully informed? Would other voters defer to them? In Canada, the province of British Columbia tried to find out. There was a movement to reform the electoral system by putting in some form of proportional representation. The provincial government took a large group of citizens and basically gave them a college class for a few months. Various experts came in and taught this group all the pros and cons of the various proposals. Eventually, the group formed an overwhelming consensus for a specific proposal which was put on the ballot. The voters rejected it by a decisive margin. The voters apparently weren’t impressed by all the study that the select group had done. Average voters made their decisions based on their own limited knowledge rather than assuming they, like the people in the select group, would see things differently if they were more fully informed.

Remember at the beginning when I stated that I wasn’t exactly sure what referendum #16 did? Of course I don’t! I’m underinformed. I’ll bet you are too. I want experts who have spent their careers thinking about the details and tradeoffs involved to sit down with politicians who have spent their careers thinking about how to balance the aggregated demands of society and figure the damn thing out. If I try to set energy policy, I’m probably going to overlook something very basic and end up with expensive, dirty, and unreliable electricity. I might even end up burning my own house down.

Referenda are a terrible way to make public choices.