Episode 11: Policy Feedback in Democracies and Non-Democracies (with Ross Buchanan)

Hello. Welcome to Hopsital Levin of the Policy Agendas Podcast, a podcast at the University of Texas

at Austin by the Policy Agendas Project. I’m E.J. Fagan. Today, I’m joined by the new project

manager of the Policy Agendas Project, Brook Shannon. Hello. Thanks for having me, Brooke. Congratulations

on your newfound power and position. Thanks. So far, it’s been a peaceful transfer of power. So, so far,

we’re gonna be talking about something along those lines today with Ross Buchanan.
Ross is a graduate of Gresser Graduate Research Fellow with the Policy Agendas Project here

at the University of Texas at Austin. He is a BD. He is on the market. Ross, welcome to the podcast.

Thanks for having me. This is I love the project. I mean, I we’ve all seen Ross work

on this project for years now. And and I’m very excited to talk about the part on the podcast. The working title

is Policy Responsiveness and Feedback With and Without Democracy. Syrus,

what what are you what are you getting at with this project? What’s the general overall research question?

So there are two real motivations for this project. The first was I wanted to really understand

the aspect of the policymaking process that most affects people’s lives. Policy outcomes

are the part of the process that most directly affects us all. And yet, ironically, it’s the part of the policy making

process that we know the least about. So I wanted to know of you to what extent does all this policy responsiveness

we found actually yield outcomes that we think that actually affect us? What is your definition

of responsiveness? So functionally it’s it’s the extent

to which the public gets what it wants. So, I mean, what the government is actually doing is

I mean, certainly a huge part of that. But I mean, what the government does doesn’t always succeed. So, I mean, the outcomes

are a very important part of that. And how do you measure that? So this is a big topic

in the big abstract topic. It’s been used. People done some budget work. People have done some work

on speed, on like specific survey questions. But you how do you do it? So I’m looking

specifically at air pollution because it does give us an objectively measurable outcome.

I think it is a representative issue for other like other issues

that have observable outcomes. But yeah, I look at air pollution as it has an objectively measurable outcome.

So what is the what what what types of air pollution are we talking about? So in this

paper I look at two specific air pollutants. There’s a NO2 nitrogen dioxide, which

is a really common. It’s a common component of smog. Like we’ve all seen it. We’ve

all smelled it. And another one is a PM2.5, which is tiny little particles that are emitted

by all sorts of processes, but they’re really, really harmful. A big problem in China. And

so they come from my coal power plants or from, you know, industrial processes. Basically, any combustion process

will release these pollutants to some extent. So, yeah, coal is a huge contributor. And

when countries want to limit these these pollutants, what do they do? So with PM2.5,

which is the more harmful of the two of you can. The filtration works. I mean, just a

also cleaner coal, if you’re just scrubbing coal beforehand, can help some extent and it is

harder to get rid of. I mean, you have to really reduce the amount of combustion processes overall

and how do you measure this? So we might we might measure this through

indirectly, but you’ve got you have a kind of a cool, cool method here. So I use satellite data.

So a local authorities run the world. China specifically have a lot of incentive to cheat

with with local with the ground sensors that they have control over. So I use satellite

drive data from the NASA and the European Space Agency

for these two air pollutants. So, you know, we know exactly, you know, with some measurement error, like exactly how much an O2,

PM2.5 or in the air, which makes you different from like a G.D.P. study or some. And that where you can read

the official statistics, an authoritarian country might be might be suspect. Right. And

there is always going to be some measurement error, but there will be no bias to that. I mean, there’s no incentive for

the European Space Agency to lie about, you know, Shanghai’s air pollution level.

All right. So I want to get to the individual studies here. So we have two papers that we’re talking about today. One of these papers we’ll put online

in a link in the description in as a working paper in the in

this episode. And I’m going to start with that one, which is policy responsiveness and feedback without democracy.

So this there’s this this paper uses the case of China. And your

argument is that if the Chinese government is responding to public opinion, even if

if it’s different than in a democracy. Kate Sookie’s explain how you you you you find that

the Chinese government is responsive. And so social has already established that the Chinese government

does to some extent do what the public wants to at least some of the time.

I mean, what we didn’t know is did this actually matter? I mean, the responsiveness is an authoritarian regime.

The responsiveness is very limited. So we didn’t really know it does actually aggregate

in anything that actually affects people’s lives. And so I have a. Policy measure

in the form of the government reports over public opinion measure in the form of a stillings measure

of the Internet searches through their main search

Bing search engine, which is called Baidu Palmy. So there’s this

relationship between the government action and what the public wants. And then I have the air pollution outcome,

which shows kind of like what happens when the government does what it does and what happens there.

Pushing goes down. I mean, you have to control for things, most notably economic output. But when

you control for that, then when the public becomes more concerned with their pollution,

the air pollution level is lower than it would be otherwise. Yeah. So this is a podcast. And so I’m going to you. There’s an audio

here with me leafing through a piece of paper, but I can’t actually show. I think there’s a very innovative

way of presenting feedback results that Ross Cross came up with here. So you to get the Dan on the paper tape to

see it. But it’s it’s a full process you’re modeling, right. So there’s there is a change in in

in the in air pollution. This causes a response in in these search indexes

that then cheat, that translate into become a policy and that then translates

into lower air pollution, which then in turn lowers both government policy and air indexes.

So it’s this is thermal static model that you see inflation and you seen a lot of other places. But

I think in in in China, people are surprised about this. Right. It’s Souters. This is a

dynamic process and is very similar to the thermostat model. But there is one key difference here is

the thermostat model. There’s responsiveness and then the public is reacting to what the government does

in China. What we find is that the public doesn’t really know directly what the government does and that the political

process is opaque, but it still does observe the outcomes. So the public is reacting

to the outcome here, not the government action itself. So, yes, it is similar to the thermostat model, but there is that key difference

of the outcome of. The feedback is coming from the outcome, not the government policy. And I think that’s a good way to transition into

your other paper, which is looking at the same the same phenomenon. But the United States. So how does the process differ

in the United States? And so we know. So at the national

level, the the big difference is, is that the political process is reasonably transparent. So the government

there is responsiveness and the public does react to what the government does because the public can generally

see what the government’s doing in a very general way. At the local level, though, the government

is not so transparent. It’s more difficult to see what the government is doing and the governance

very fragmented at the local level. There’s a lot less media coverage. There are lots of reasons for this. But so

what I find is that the US, the city level, works a little bit more like China. I mean, there is responsiveness

and there are responsive outcomes, but the public has difficulty responding to what the government’s doing. Instead,

it is reacting to the outcomes it correctly observed. All right. So I think that covers the paper that

the two papers in the project overall pretty well. So, Brooke, we’re going to transition to our second half this podcast.

Let’s talk a little bit about this critically. Sure. So in authoritarian settings, as you say,

the public is reacting, but not necessarily to government actions. They see physical outcomes because

of the opaque nature of government action. They respond directly to those outcomes.

So what exactly do you think that that means for the

policy agenda, especially of authoritarian governments?

Do you have a hunch on different policy areas or like in general, what does it say about

the policy agenda? So I think a huge implication of this is that

the visibility of the policy area matters a lot for responsiveness. I’m only

at the moment I’m only looking at one area and it’s a very, very visible one. And certainly there are other

visible policy areas. I mean, people can see construction projects by their home. I mean, they have some idea

of how is of how how good local schools are. But there are lots of areas that are not so

visible. And so I would expect that when the political process is opaque,

that there’s going to be a lot less responsiveness in those in these less visible

areas. Right. I think air pollution is a really cool, cool

measure and a cool way to show policy responsiveness and government action

because it’s like the ultimate public good. Right? It’s like there is no

part there’s no part of air pollution. I guess that doesn’t somehow affect everyone. Right. You can’t hide from it.

Yeah, you can’t hide from it. Everybody’s got to breathe. And I think it’s cool that you also speak to this sort of

like classic agenda setting idea where like in 1971, this book on

on the U.S. also comes out and it talks about government action and like outcomes

and federalism and stuff like that. So I’m wondering, especially tapping into this agenda

setting idea, how does government action in an authoritarian and

a non-democratic area or even in a local area and the local government area that isn’t

known for being incredibly transparent or incredibly.

Consistent in terms of acting in a responsive way, how does that differ in

in these different types of governments? So I

think this is. This is a major question I’m hoping to kind of get to.

My work continues. I suspect that there is a difference

in the consistency of the government’s focus. I mean, the time period I’m looking at

Russia and China is pretty limited. But one thing I want to do

over time is I want to I want to expand that. The time window and to see

how consistent is this responsiveness over time. I mean, there is always anecdotal evidence that there’s that there

are huge punctuations and Chinese responsiveness, maybe more than in the United States, although that story may be determined

by that. So I think in the early 2000s, air pollution in China was terrible, but it was not on the public

agenda. And then it kind of sort of just kind of lay with the huge bang of shows

up in the late, very early 2010s is changing. Air pollution in China

normally distributed it. It increases that have relatively consistent rate

for a fairly well the last 20 years or so, 20, 30 years. But

it just Iraq, it’s under the public agenda around like 2011, 2012. That sounds like a punctuated

equilibrium story. Yes, it does. Could you use bricking, Eric? I think he wrote this a few months ago.

Could you use change in air pollution as the same way in the same way we used budget data

to measure peaty across systems? Potentially, yes. There’s

a data availability is kind of an issue here. I mean, the Chinese data, the satellite data itself for air pollution

just doesn’t go back that far. I mean, I guess it technically starts the 90s, but only starts getting good like in

the late 20 knots. So we’re kind of limited right now. But going forward, yes. That

there’s a potential use for this data. It’s interesting cause, you know, when you know, we’re so limited

in budget data and we’re limited to just a few specific types of systems, it is really cool to

see just the Petey’s story using the same data in, you know, across,

you know, given rise of authoritarianism, different, you know, different levels of development

and that we could learn a lot that we haven’t learned just from these. Kinkel

is. Is the Olympic story in with air pollution? And maybe you

can tell our listeners the story, which I’m referring to, which we’re all referring to. The

decrease in air pollution around the Olympics. Is that a peaty story? Of.

No. So you’re referring to a 2008. I mean, the air pollution

in Beijing, which is notoriously bad. I mean, suddenly got much better for a very short period of time.

I. There was a very quick fix, basically,

and there’s a lot of evidence that, in fact, air pollution kind of got worse after that. I mean, so there

is. Over the course of over a year, it makes very little difference because like what we can see

is that when there’s a lot of pressure for factories have shut down in the very short

term to make the skies pretty for the Olympics on that, then they ramp up production afterwards

to kind of make up for the lost quota. So these things only happen in China. And I should mention that I’m

using annualise data to kind of get rid of these kind of weird blips. Like there’s there’s a term

impact blue referring to that, the APEC conference and that, you know, the authorities in Beijing will pressure

all of the factories around there to to shut down and have a conference. So they have a nice blue sky for the photo.

And I dont for the purposes of my project, it’s not really responsiveness. I mean, that’s just the elites who want a good photo

op for a very short period of time. Got it. Hi, Brooke. So transitioning to the

U.S. paper, I’m a state and local person, so this paper’s super exciting for personal reasons.

And I was really taken by your findings that at the state level,

feedback is really driven by government action versus at the local level, which is driven by outcomes. And

I was especially interested in your I guess in your model

and your control variables that you use. So you control for economic output and a bunch of fixed

effects for geography and region and things like this. And I’m wondering if

you found any difference in states and locations or regions that have a history of dealing with this

type of air pollution. Thinking of California’s wildfires or Ohio and Pennsylvania

like acid rain in the 70s. Is there a path dependent story with air pollution?

Yes. I mean, the source of power is a very big one. I mean, coal is

terrible for the air, but investments suddenly where you’re getting your power take time.

And I mean, some areas are just they’re lucky enough to be really conducive for a hydropower or

there a lot of investments in nuclear power. Decades ago. And then the geography

matters a lot. If you have, you know, Los Angeles mean one of the reasons the air pollution so bad is this kind of a

on the kind of traps and all their pollution from cars and

geography does change, but it takes a very long time. So, yeah, I mean, there is a very strong there is path dependance

and kind of like where you’re getting your power and these are long term investments kind of payoff over time. And then

there’s geography, which are kind of stuck with. But is that path dependance change the politics, doesn’t

it? Question To better understand that we’re going to need a longer we’re going to need more time

in the data? I would suspect so. But I mean, the public is reacting

more to levels of change than they are to absolute levels. So if things are bad just because

of geography, we’re doing power. The public does kind of factor that in and kind of how they’re and what

they’re expecting. But I mean, things are getting worse regardless of the baseline and the public becomes more concerned.

So do you think that change is more evident in public levels of attention or

or even in salience like using Google or search engines in general to find this stuff

in terms of local like locations with all these idiosyncrasies of

L.A. being in a basin or Denver being in a basement, where you see that like brown tire on the mountains

when the smog is bad versus at the state level to where they expect more like a national

type of politics. Sir, can you can you repeat your question,

please? Yeah. Do you expect like this sort of this sort of

like path dependance to come out and idiosyncrasy of a location versus the state level,

which is more like ideological politics that’s reflective of national level discourse?

So I think so your question is kind of getting it like what is the difference between the local and the state

level, right. And so a big issue, a I think a key difference

is what can you make of the things you observe around you and

at the state level of you? You do have these kind of very noisy signals of how

well the state government is doing. Like any. There is some correlation with your local air pollution versus the state level

air pollution. But that correlation is extremely noisy. So I think that that

I think that that’s a really key difference as far as how can you interpret what’s around

you and how can you evaluate the government based on that. So that makes me think of

of a study of a Soroka and vision paper on Canada provincial politics, where people

like the the federalism system just kind of makes it difficult for people to know who to be mad

at. So I’m wondering, like in terms of responsiveness. How

does responsiveness change at the different levels of government? Is it all the same tactics?

Or what does it look like? So you’re saying like, what does the public actually do when it becomes concerned?

Yeah. What does the public do? And then also, what does the government do? Are there any differences in the responsiveness like

in China? For example, they tell the factory owners to stop producing the government. Authorities do.

Right. So what does it look like in levels of federalism? So in China,

the public the public does at least implicitly threaten the government. I mean, the government does

keep track of like how concerned is the public because, like, they want to avoid civil unrest. I mean, they. So they put a

lot of effort into monitoring, just like how angry is the public at various issues?

And then they react preemptively. And sometimes the public does protest or threatened to protest and

government officials will actually minimally to kind of avoid that in the US. I mean, there

are electoral incentives, obviously, but there mean so I think especially the local level, there is some degree

of what kind of see in China. I mean, no one wants a huge scandal on their

hands. I mean, no one wants people shouting at them in their office. And when people do do that kind of thing,

I mean, Brooky, you’re you’re you’re the expert on local politics. I mean, how how effective is is a

city like Austin? Realistically, a Democrat is probably not going to be electorally

threatened. I mean, how much of an incentive is it just to not have people yelling at you to get

you to get right? I imagine it’s pretty high, especially considering most local governments are nonpartizan.

So it doesn’t matter really if there’s a Republican, if there’s a lot of smog and people’s kids are,

you know, showing physical effects, I’m sure that the incentive is pretty high for any local official

anything at our local bureaucrats in China at all? Oh, most definitely. Yeah.

The huge incentive built into their political system is avoiding

avoiding unrest. I mean, that the regime needs to survive and it wants to survive. So

centers are officials have very strong incentives to avoid even just the displays of civil

unrest. So there’s a. So they’re being chased by a crocodile. Right. And so they’re running really fast because this crocodile behind

them. Does that make them more responsive or more effective at responding to public

opinion than the United States government or democratic government? Or are there still

things blocking them in an authoritarian government from being? Are they just being responsive enough to stay out of the reach

of the alligator? Crockett Whatever, whatever the metaphor is. Or are they, you know,

really get it? So this is a really good question, one I am hoping to get to when they have

a longer time series. So this is kind of conjecture at this point, but

a major difference maybe time horizon is that

there’s a very strong incentive for Chinese officials to make things better.

Now, for however long, their term is usually there in a region for to say, you know, three or four years

or so and then transferred somewhere else. So that Chris incentive to centered can make things better for the moment.

But then I mean, not now take a longer term investments to kind of improve things

in the U.S. so that the incentive may exist. It’s but that’s. This is a question

I’m kind of hoping to get to with ongoing research. I mean, I’ve seen the peaty leisure, you know, kind of use the

informational advantage versus institutional efficiency argument. And

for the most part, in the larger the information advantage always wins. Right. That that over the long term, you get a much more

less efficient pattern of change in in Chinese government. So I’m wondering if

if you get these kind of like really inefficient, but big changes under China. So

like, you know, the Olympics is the example that comes to mind. But the other the impact was sky.

All those examples are like incredibly inefficient. Right. We’re just now shutter a factory for

a week. That that doesn’t seem like the, you know, the best public policy in terms of, you know, for factory

policy. But like short term solves your problem. Whereas the United States, you know, you got you

know, you that they might take a little bit maybe less efficiency in

acting, but they act better over the long run. Here is a very good question. A big

a major difference is, is that the Chinese government isn’t constrained by laws in the same

way that most American political authorities are. So the Chinese government can

more easily kind of acting big, dramatic ways, like it can just force people to relocate, to build dams,

for instance, whereas without having to go through a lengthy legal process like in the US.

However, I do think that this I do think that when American authorities are motivated,

I mean, they can make things happen. I think one of I strongly suspect one, the reasons

why Chinese officials have acted so dramatically in recent years is just because their pollution is so

bad in China. If the air pollution were comparably bad in the U.S., it has been in the past,

although I don’t have good data that far back, I think that the U.S. government would find

ways to act very dramatically. So the difference in the difference. Between what

the political systems are actually doing may not be quite as dramatic as it may appear.

The I saw some interesting paper at the part of the genius project

over the summer and they were comparing that Hu Jintao and Xi Jinping rate of regimes

and their their kind of flavors of authoritarianism. You might probably assume that the Hu Jintao

regime is less authoritarian and that Xi Jinping regime is certainly less repressive. But they actually

showed a less efficient model under Hu Jintao that he was essentially trying to protect the status quo, not not upset the

apple cart and keep going. And I think if most of your data is

is as Egypt, maybe you have a little bit of both. Both regimes. Right. Is there a change from regime

to regime? Yes. So given the limited data. One, be careful

about drawing any kind of definitive conclusions about these regimes. But at the end of whose

reign you have, that’s when air pollution is just then becoming a really huge

issue. So the. So in recent years under Xi Jinping, more a

lot more has been done about air pollution. But this is kind of as it has been as it was.

They have a change over just as the air pollution is kind of rocketing onto the public agenda.

So be very careful about attributing it to the the nature of

their theadministration’s. But can you draw comparisons then between the two

leaders or between like the system in general, then to

say this is how leaders would act in an authoritarian or non-democratic

setting? Instead of like this one president or

this one leader acted in a certain way and someone reacted very differently to to

public say to salience of an issue or a public response of public opinion, ‘scuse me.

When that then strengthen your argument for this is just how it works in an authoritarian setting.

I guess once you have more data. Yes, it is a possibility.

I mean, so I think a major so this is conjecture. So I want to be very clear about that. I can’t

be sure that all the time. But so this is reasonable conjecture

that I think I can Witter establishes that. I think a major issue is

our major difference is which level of government is acting. So we’ve seen a lot more

effort at the national level in recent years and in authoritarian China as the issues become bigger.

And but there still was an issue going further back. But I suspect that most of the

issue was kind of being initiated more at the local level, kind of less there’s less like actual strategy, but there still was a lot

of government action. It’s just more localized. Are you familiar with the M4 anthropic principle

in astronomy? Yes. OK. So there’s the anthropic anthropic principle for everybody who doesn’t

know is the notion that

life can be very rare. But if a civilization that lives would only observe would only

be able observe a civilization that has life in it. So there’s this is essentially that

the fallacy that that says that just because we’re alive, that doesn’t mean that there’s billions of other civilizations out there.

We could be only only civilization out there and it wouldn’t be statistically weird.

I think that’s correct. I wonder if there’s anthropic principle for authoritarian regimes. So if an authoritarian

regime did not respond to two public concerns,

would that regime be around still? No. Old regimes have to to

some extent satisfy public what the public wants.

I mean, they may not do it very efficiently. They may not do it very well. They may not do it in all policy areas.

But every regime needs to kind of keep enough people happy enough

or at least not so angry to stay in power. I agree that I’m so. But we probably

can’t say that they all do, though, right? Because the ones that don’t. Get revolt.

There’s a revolution in. And they get they get deposed in favor of regime that’s either democratic or in a certain regime that

made that does respond to public opinion. I mean, maybe maybe there’s intothat totalitarian regimes

that don’t. But, you know, I suspect rather. Do you suspect. If we could do this in

North Korea or whatever some other real totalitarian country, do you think there would actually be

responsiveness? Yes, in fact, yes, I do. I don’t think. I don’t think it responds for

such a regime usually responds very well. But you have to keep at least a significant

portion of the population happy. I mean, even the most authoritarian,

a centralized leader at least has to keep the security forces happy. Right. So it may not be

very what we would normally think of as good responsiveness, but there would be responsiveness of some kind.

And if you were to go back, this would be slide and the dependent variable. But if you go back and look at major, major

protests in authoritarian countries, do you think you would find those follow a

authoritarian regime that is not responding? So let’s say before the Arab Spring,

if you can go to the Arab Spring, what do you think you would find? You would find a failure to respond there,

I’m sure. So I think that responsiveness to some extent is politics as normal, even in highly

authoritarian regimes. I mean, protests can happen for all sorts of reasons. I mean, they can happen because people

are very frustrated because maybe there is responsiveness, but it’s just not enough. But it can also

be triggered because of responsiveness. I mean, people kind of seem like, oh, we’re getting a little bit of what we want. We want more.

So it’s my city isn’t attract is an attached to any one strategy.

I mean, the public uses a lot of strategies and it adapts as the government adapts. So we protest is one

of those. But I don’t want to. I don’t wanna give people the impression this is like the only thing. Even if we’re cheering machines

that’s kind of pushing the government. It seems more difficult for governments at

the local level and also in other sort of non-democratic

or what we consider, I guess, traditionally Democratic, such as the American states to

almost prove to the public to sort of satiate their claims.

Right. That that this is a problem that needs immediate action. It seems like the

state level can sort of get off scot free anyway. Right. Like if if they go if the people

are responding to a government action, then the government can. The governor can

put it in the annual state of the state speech. Right. Like they can say, this is our

new mandate and we’re creating this new agency to deal with air pollution. But it seems like

they might not actually have to do anything. It’s interesting that like outcomes are more important in a in a

non-democratic slash local American setting because people need

to tangibly see or breathe the change. Whereas at the state level and in my mind,

also sort of at the national U.S. political rhetoric level. Right. Like they would have to put

something in place and not necessarily have an outcome. I think that’s a really interesting part

of your story. Yes. That’s a good question, because the you

raise a really interesting issue here of government action doesn’t always work. Right. Even very

well intentioned policies. Sometimes they just fail or they can even be counterproductive. So what

happens that does so we know that the public is reacting to what the government is

doing at the national level, not so much the outcomes. But what about when the outcomes

are terrible? And so in this issue area, for the time period

I’m looking at. I mean, I don’t really I can’t really get at this issue yet, but there would be a learning process. I suspect

that if if, let’s say the governor trying to address the economy, but everything is doing is

just making it worse, which is what we saw in Europe, really Great Depression, then

eventually the public is going to kind of adjust to that. I suspect that

that they would eventually start reacting to the outcomes if they they weren’t if they begin

to not trust what the government the government action that they’re seeing. Well, this has been great.

We’re gonna wrap up real quick with our normal wrap of questions. Ross, what should other people read if they’re interested in the subject

other than your own work? So a really informative book, I think, as Susan Butler’s The Summers

State. I mean, a lot of ways I see my work is kind of the downside to this podcast. It’s like the inverse

of that almost because I’m looking at sort of the non non submerged state like the if it if

it’s emerged people are responding to the outcomes, if it’s not submerged, people are responding to government self. So I

really recommend that book. And or you can also go read the listen to the podcast we record with her on

the cover expanded version of that same work,

which is which title I’m forgetting off top my head right now. Ross, thank you so much for joining us. Ross is on the market this

year. So please hire him. Brooke is not on the market this year. Thank you for joining me. And this has been

your Policy Agendas podcast.