Movie production incentives don't pay off

I've written before about the supposed economic impact of movie production (see here and here). The argument is that movie production increases employment and contributes to growth, both in its own right, and in its ability to increase subsequent tourism. I question both of those conclusions, or at least question the idea that the benefits of movie production incentives exceed the costs of those incentives.

And now I have some new research support for my conclusions. In a new article published in the Journal of Economic Geography (sorry I don't see an ungated version online), Mark Owens (Penn State University Erie) and Adam Renhoff (Middle Tennessee State University) look at movie production incentives granted by U.S. states. They constructed a dataset of all movie productions in the U.S. over the period from 1999 to 2013 (some 16,725 movies), including filming locations and their characteristics, as well as the characteristics of the movies. They also collected data on four types of movie production incentives (emphasis mine):

A refundable tax credit allows the production studio to receive cash back when the value of the tax credit exceeds their state tax liability. A transferable tax credit allows the movie studio to sell their outstanding tax credits to a third party if the value of their tax credits exceeds their state tax liability. Standard non-refundable, nontransferable tax credits, which are not very common, offer movie producers significantly less financial flexibility. Cash grants or rebates are cash transfers (treated in this research as a percentage of the qualified production spending) from the state to the movie production studio. They are not tied to the company’s tax liability.

They also collected data on the minimum spend required to qualify for incentives. Using a reasonably sophisticated discrete choice modelling approach, they find that:

 ...movie production incentives vary in their ability to attract films, and vary with respect to firm size. The magnitude of the effect of incentives is the largest for mid-sized studios (the so-called ‘mini-major’ studios) where each incentive has a positive and significant impact on location choice. None of the production incentives significantly impact the location choice for independent studios, the smallest firms with smaller budgets, possibly because independent movies are less likely to meet minimum requirements and/or because they are less likely to be profitable. Consistent with intuition, we find that major producers respond more favorably to refundable and transferable tax credits than to standard (non-transferable, non-refundable) tax credits. Refundable tax credits are more effective than transferable tax credits in attracting major studios.

In other words, movie production incentives are effective in attracting movie productions. No surprises there - the studios are simply responding to incentives - if a particular location offers greater benefits than another location (including in the form of tax credits or other incentives), then the studio is more likely to want to produce their movies in that location.

However, Owens and Renhoff then conduct a back-of-the-envelope calculation weighing up the costs and benefits of movie production incentives for U.S. states. They have to make a number of assumptions here, because they don't have complete data on movie production spending, etc., but their results can be taken as indicative:

Based on our estimates, the movie production incentive programs are all revenue-negative, implying that a dollar awarded in tax credits leads to an increase in tax revenues, that is, less than a dollar. In this sense, the efforts to attract film production to a state do not ‘pay for themselves’ with higher resulting state tax revenues.

In other words, costs are greater than benefits. Movie production incentives do not pay back the taxpayer. However, if a state government's is simply to increase employment, Owens and Renhoff do offer some hope:

The tax incentive programs appear to increase employment at a lower cost to the state than the cost of directly increasing the number of state employees.

Of course, that doesn't mean that there aren't more cost-effective ways of increasing employment. Asking if a policy increases employment at a lower cost than if the government employed people to dig holes in the ground and then fill them in, is setting a very low bar.

Unfortunately, while this research provides justification to get New Zealand out of the Tiebout competition that is providing large movie production incentives, I fear that it will fall on deaf ears.

Read more:

• Lord of the Rings, tourism arrivals and harvesting
• More on hobbits and tourism

The minimum wage and suicide

One of the more robust findings in the happiness economics literature is the negative impact of unemployment on happiness (as just one example, see this paper). So, to the extent that unhappiness (including depression) is associated with suicide or attempted suicide, we might expect higher unemployment to be associated with more suicide.

There is ongoing debate, and there is no settled consensus, but I believe that the weight of evidence seems to suggest that higher minimum wages do reduce employment (and therefore increase unemployment) - see my most recent post on this topic here. So, to the extent that minimum wages increase unemployment, and unemployment increases unhappiness, then we might expect higher minimum wages to be associated with more suicide.

However, that isn't end of the story. Higher minimum wages are associated with higher incomes (in the simplest sense, among those that receive the minimum wage), and higher incomes might buffer against depression (and suicide). So, there are potential effects in both directions for the relationship between a higher minimum wage and suicide.

A recent article in the Journal of Epidemiology and Community Health (ungated here, and see this New York Times article as well) by John Kaufman (Emory University) and co-authors looked into this relationship using monthly state-level data from the U.S. over the period from 1990-2015. They found that:

When further controlling for state-specific time-varying economic variables, we estimated a 3.5% reduction in the suicide rate for every additional dollar in state minimum wage...

Those results were based on the suicide rate among people without a college education (who might be expected to be more affected by the minimum wage). They found a statistically insignificant effect on suicide among college-educated people. So, it appears that the positive effect of the minimum wage on incomes may outweigh the negative effect of unemployment on suicide.

Kaufman et al. also report that:

We found a statistically significant interaction between minimum wage and unemployment... indicating that the impact of minimum wage varies across the unemployment rate... When unemployment is high (>6.5%), progressively higher minimum wages are associated with lower suicide rates, while at low unemployment (3.8%–6.5%) the effect of minimum wage is attenuated, with little effect observed at very low unemployment (<3.8%). We observed the highest suicide rates when state minimum wage was no higher than the federal and unemployment was high. Curiously, the lowest suicide rates were observed when both minimum wage and unemployment were high (eg, unemployment >7% and minimum wage ≥US$1.75 above the federal).

These results are interesting, but perhaps problematic. Since it appears that unemployment and minimum wage rates are negatively related, it is difficult to disentangle the two effects in this way. So, I would put much more weight on the headline result than on this additional analysis. However, I think more work needs to be done on this research question, as the mechanisms underlying the negative relationship (which has also been found in some other studies) remains under-explained. That might usefully be achieved by looking at individual-level data, rather than aggregate state-level data.

[HT: Marginal Revolution]

This study tells us nothing about whether studying economics make students less ethical

Does studying economics make students less ethical? The research question is interesting and potentially important (and I've written about similar questions before - see here and here). Answering it, though, is pretty challenging. Simply comparing some measure of ethics between students that have studied economics and students have not studied economics is no good, because perhaps less ethical students choose to study economics. That would create a problem of selection bias. Similarly, comparing some students who chose to do economics early in their degree programme, and students who chose to leave economics for later, faces the same problem.

So, I was quite disappointed when I read this book chapter by Christian Mastilak (Xavier University) and co-authors (ungated earlier version here). Promisingly, they conduct a lab experiment with business school students, some of whom have studied microeconomics already, and others that haven't, and test how ethical their choices are across a few experiment tasks. They find that:

...participants with exposure to agency theory assumptions through either an experimental manipulation invoking a competitive, wealth-maximizing frame consistent with common agency theory or prior microeconomics coursework acted more unethically than participants who had neither exposure to agency theory.

By "agency theory" here, Mastilak et al. are really referring to economic theory more generally (and in fact that's how they refer to it in the ungated version of the paper). They started their experiment by having participants play a prisoners' dilemma game. The game had two different framings, and each participant saw only one framing: (1) as a competition between two competing firms, with an emphasis on each firm's individual payoff; or (2) as a collaboration between two NGOs, with an emphasis on the joint payoff to society as a whole.

They then went on to have their participants engage in other tasks, a few of which were used to evaluate ethical behaviour (or otherwise). They find no differences between participants on two of the tasks, but they do find statistically significant differences on one task, that involves participants overstating a budget request, where they would receive a payout that enriches them.

Mastilak et al. compare the budget request between participants who were exposed to the competitive or cooperative framing of the prisoners' dilemma, and find that participants exposed to the competitive framing made larger budget requests (thereby acting less ethically). However, comparing participants who had completed prior economics with those that had not, there were no statistically significant differences. Now, see if you can follow these bits of the paper:

Panel B reports a two-way ANOVA. The effect of the experimental condition (frame) is significant (p = 0.011).

Initially, it appears there is no effect of prior microeconomics coursework as the effect of prior microeconomics coursework is not significant (p = 0.193)... coursework. Tests of simple main effects are reported in Table 1, Panel C. For participants who had taken microeconomics, the agency frame condition had no effect (p = 0.201). For participants who had not taken microeconomics, the agency frame condition had a significant effect (p = 0.026).

We interpret our tests of simple effects as indicating that either agency frame manipulation or prior economics coursework is sufficient to increase unethical behavior...

Wait - read that last bit again. They interpret their results as indicating that "either agency frame manipulation or prior economics coursework" increases unethical behaviour. And yet, one paragraph earlier, they clearly say that "the effect of prior microeconomics coursework is not significant", and that the effect of framing on participants who had taken microeconomics was also not statistically significant.

And even their results of the agency frame having a significant effect for those who had not taken microeconomics is shaky, because their whole sample size of participants who had not taken microeconomics was 11 students. And because microeconomics is compulsory for business students, those 11 students are students who had left economics until later in their degree programme, so can be assumed to be meaningfully different from the other students (self-selection bias).

Who peer reviewed this book chapter? The headline result is basically not supported by the analysis. The effect of the prisoners' dilemma framing on subsequent unethical choices is interesting (although I have some issues with the particular way they conducted the framing), but it tells us nothing about whether studying economics affects ethical behaviour.

Experimental studies can often help us to better understand relationships in a controlled environment. However, studies like this one do not help at all.

Read more:

• Does education in economics make politicians corrupt?
• The impact of business and economics education on moral competence

Framing, loss aversion, transaction utility, and reusable coffee cups

This article in The Conversation yesterday, by Sukhbir Sandhu, Robert Crocker, and Sumit Lodhia (all University of South Australia) caught my attention, because it nicely illustrates some of the concepts from behavioural economics that I discussed with my ECONS102 class last week:

Many cafe owners offer discounts ranging from 10c - A$1 to customers who bring in their own reusable cups.

But our findings reveal these discounts are ineffective in changing consumer behaviour.

A cafe owner we interviewed described how, despite providing a 20c discount for reusable cups, she didn’t think saving money motivated her customers:

The regulars were people who’d happily drop in a dollar tip into the jar kept on the counter. They were therefore not that concerned about 20c discount.

We know from previous behavioural psychology literature consumers are more likely to be what’s called “loss averse” as opposed to “gain seekers”. In other words, people hate paying extra for takeaway coffee cups more than they like getting a discount for bringing their reusable cups.

So, if you own a cafe, focus on making consumers pay extra for choosing takeaway coffee cups rather than offering discounts for reusable cup use. It’s more likely to motivate customers.

Let's say that a cafe owner wants to encourage customers to use reusable cups. They might do this because of concern for the environmental effects of disposable coffee cups, or the cafe owner might simply recognise that disposable cups cost them money, and so offering to fill a customer's own cup must be slightly more profitable for the cafe owner, because then they don't incur the cost of providing a cup.

Putting aside any cost differences, cafe owners could discourage their customers from disposable cups by making coffee sold in disposable cups more expensive. We know that when something is more costly, rational consumers will buy less of it. This also makes coffee in reusable cups relatively cheaper, and so would encourage some consumers to switch. Let's consider two different framings of the price difference: (1) consumers who use a reusable cup receive a 20 cent discount; or (2) consumers who use a disposable cup have to pay an extra 20 cents. 

How many consumers would switch? If consumers were purely rational, it wouldn't matter how the price difference was framed. A 20-cent discount for using a reusable cup and paying 20 cents extra for a disposable cup are exactly the same (provided the prices are the same in each case). Both options would lead to the same number of customers switching to a reusable cup.

Now here's where some behavioural economics comes in. Consumers (like every decision-maker) are not purely rational, they are quasi-rational - they are affected by cognitive biases and use heuristics when making decisions. Framing makes a difference to quasi-rational consumers, but it's not clear which framing should make consumers use fewer disposable cups.

One of the cognitive biases that quasi-rational decision-makers are affected by is loss aversion - decision-makers dislike losses much more than they like equivalent gains. In this case, the loss in utility (or satisfaction, or happiness) for the consumer from paying 20 cents extra for a disposable cup, is 'worth' much more than the gain in utility (or satisfaction, or happiness) for the consumer who receives a 20-cent discount for using a reusable cup. So, we would expect the 'loss framing' (20 cents extra) to have a much bigger effect on consumer behaviour than the 'gain framing' (20-cent discount). 

However, another cognitive bias that affects consumers is transaction utility, which I have blogged about before. Transaction utility recognises that consumers not only receive utility from the good or service that they purchase, but also from the act of purchasing. If a consumer feels that they are 'getting a good deal', this makes them happier (higher utility), and makes them more likely to purchase. So, based on transaction utility, we would expect the 'gain framing' (20-cent discount) to have a bigger effect on consumer behaviour than the 'loss framing' (20 cents extra).

Given that Sandhu et al. found that the negative framing had a bigger effect overall, it appears that the loss aversion effect is larger than the transaction utility effect. It would be good to see more research on this though, that disentangles those two effects more.

Overall, the takeaway message from this research is that if you, as a seller, want to steer consumers away from something using a price difference, present it as involving a loss to them (they have to pay extra). On the other hand, if you want to steer consumers towards something using a price difference, present the alternative as involving a loss to them. At least until this has been investigated a bit more, it appears that paying extra is a more powerful motivator for changing consumer behaviour than a discount.