The alcohol consumption J-curve

People and governments love to simplify problems to a single issue - Speed kills, Helmets save lives, Stop the boats. It helps them appear to be ‘doing something’. But real life is not so simple.

Take alcohol. While heavy drinking has long been acknowledged as being socially disruptive, more recently, the fight against alcoholism has been partly driven by arguments around health impacts.  Yet their are both positve and negative health effects from alcohol, and the positive effects are usually overlooked.  The unintended consequences of policy are also rarely considered.

The alco-pops tax was one measure aimed at curbing binge drinking, but it was a fizzer. Sales of other alcoholic beverages increased significantly, offsetting much of the claimed benefits of the tax.

Additionally, no one considered that more expensive alcohol might encourage binge drinking at the expense of casual drinking. If your preferred alcohol is more expensive, there is less incentive to drink in a casual setting where you don’t end up drunk. Why spend the extra money on alcohol unless your intention is to get drunk?

It’s a thought that has crossed my mind when considering the drinking patterns around the world. Those countries with the most expensive alcohol, usually due to alcohol taxes, usually have the most extreme binge drinking culture (that's been my personal observation, and I have no hard evidence to back up the claim).

But alas, these considerations are a little too real for the average policy maker to consider.

The Australian government’s health message about alcohol follows the single issue simplification pattern precisely (their emphasis).

Due to the different ways that alcohol can affect people, there is no amount of alcohol that can be said to be safe for everyone. People choosing to drink must realise that there will always be some risk to their health and social well-being.

But alas, the evidence seems to strongly contradict this simplified message (although the alcohol consumption guidelines are a little more generous).

The overwhelming conclusion from large scale longitudinal studies is that moderate drinking improves longevity. The graph below illustrates.

Men who never drink are just as likely to live as long as men who average 4 drinks per day, with those who drink about one drink per day (or 7 per week) likely to live longest.

The results are partly attributed to the social interactions that are associated with alcohol consumption.

One important reason is that alcohol lubricates so many social interactions, and social interactions are vital for maintaining mental and physical health. (here)

Somewhat surprisingly there are no other plausible explanations at hand that I know of. The debate appears stuck on the ‘is this relationship real’ stage, without moving on to considering why it might be real.

So here is a suggestion.

Often our body has systems that work on a use-it-or-lose-it basis. We use muscles, they grow. We don’t, they atrophy. Our bodies have a built in system (ethanol metabolism) to break down alcohol. Perhaps the very act of digesting of excess alcohol keeps the system healthy for longer.

As my good friend Wikipedia says

If the body had no mechanism for catabolizing the alcohols, they would build up in the body and become toxic.

In any case, the health impacts of alcohol consumption are another example of how common understanding and resulting policy is often detached from the more rigorous academic research. It also highlights repeated failure of policy makers to consider the unintended consequences of well meaning policy.

HT: Eric Crampton at Offsetting Behaviour

Cannon's Law

Cannon's Law says that if it will work, then the government won't do it. If regulation would really bite, the regulated parties will work the political system to kill it. (here)

With so many simple effective regulatory reforms that would greatly enhance economic efficiency being constantly overlooked, this spoof law resonates with my experiences. It is particularly relevant to the mining tax, the carbon tax, and other reforms currently being considered (even to the Greek debt situation).

The point is that economists often oversimplify policy matters. They consider government decisions is isolation of the interaction between affected parties and politicians. Remember, most policy changes involve winners and losers, and the relative political clout of each group can determine the actual outcomes.

My version of Cannon’s Law would be a little more subtle and say –

The more effective the regulation in achieving outcomes for the public good, the less likely the regulation will be appropriately implemented due to the increased likelihood of regulatory capture.

For those not familiar with regulatory capture, the following definition might be useful.

For public choice theorists, regulatory capture occurs because groups or individuals with a high-stakes interest in the outcome of policy or regulatory decisions can be expected to focus their resources and energies in attempting to gain the policy outcomes they prefer, while members of the public, each with only a tiny individual stake in the outcome, will ignore it altogether.[1] 

Regulatory capture refers to when this imbalance of focused resources devoted to a particular policy outcome is successful at "capturing" influence with the staff or commission members of the regulatory agency, so that the preferred policy outcomes of the special interest are implemented.

Regulatory capture does not have to involve intentional ‘ship jumping’ by agencies. The simple fact that you are spending a lot of time talking to the industry you regulate makes you identify with that industry.

For example, the agency who is charged with regulating the amount of water available to be used in each river system would talk with farmers quite a bit. They would begin to think that they are in the ‘water business’ rather than the ‘protect the public interest’ business, thus subtly shifting their subconscious focus.

To be clear, there are two key political realities that economists usually overlook.

1. If a new regulation will be very effective, it is unlikely to see the light of day; and
2. Even if a new regulation is adopted, it is likely to be watered down by vested interests ‘capturing’ the regulatory body.

In the end, perhaps our political leaders are not as powerful as we think. Our democracy appears to have a secondary feedback loop between politicians and interest groups that chugs along behind the main cycle of politicians reporting to the people at the ballot box. The power rests with the people and the alliances we form in business and social practices.

Perhaps the power is distributed unevenly due to the differences in wealth between groups.  And maybe there are simple ways around that, like capping and declaring political donations.  Or maybe this wealth difference is not so important - even big business needs to keep customers happy.

Maybe these thoughts are no surprise to you. But it is nice to lay it all out and ponder our own positions is this social game.

Warwick McKibbin tells it straight

RBA director Warwick McKibbin has a reputation for speaking his mind on key economic matters. It appears he is at it again, and it is worth considering his informed views on some global and local matters.

Here are my favourite points from the linked article.

Referring to the most recent global economic crisis as a mere ''blip'', he said the coming crisis could undo the mining boom and bring on inflation of the kind not seen since the 1970s.

The response globally to the financial crisis was mostly to kick the can down the road.  At some point this must stop, and the longer it goes on, the worse the resolution must be.

Joking that he could not talk about Australian interest rates, which were in any event ''always appropriate'', the Reserve Bank board member warned that the inflation would spread worldwide.

I would say that Australia has been severely buffered from global inflation by our exchange rate.  Who knows how long this can last.  My suspicion is that if interest rates go up to fight inflation, our local economy will flounder and we will end up having to drop interest rates severely and get our share of inflation anyway.  

Australia needed a sovereign wealth fund to store mining income while it lasted, ideally stored in a separate account for each taxpayer so the government could not raid it.

Of course I agree about using the tax revenues raised from the mining investment boom to save for our future.  The idea about giving each taxpayer an account seems particularly interesting.  I haven't put much thought into it but at first glance like the idea.

The $50 billion national broadband network epitomised the sort of waste Australia could not afford. ''I would say to any politician who thinks that spending is worthwhile, take your salary as shares in NBNCo. If you think it's a good investment, you'll be ahead,'' he said.

While I think the idea is great, the positive externalities generated by the NBN should factor into the equation, yet these can't be captured by revenue from broadband access.  But in general I like the idea. 

Apologies - overzealous comment spam filter

A number of people have made thought-provoking comments on this blog in the past months, only to have their comments disappear moments after submitting.  They have been accumulating in the blogger spam filter.  I have let through the legitimate comments now.

My apologies.  I will keep an eye on this from now on.

What pay rise?

I work with a bunch of economists. Now is the time of year we have performance reviews and negotiate pay rises and promotions. But no one has yet discussed the fact that high effective marginal tax rates (EMTR) greatly reduce the real take-home benefits of a pay rise.

(EMTR is an estimate of the change in take home income after tax, and after accounting for reduced welfare payments)

An EMTR higher than 50% is very common in Australia for low and middle income earners, and any push for greater middle class welfare will simply increase this perverse tax incentive.

For example, you get a pay rise approximating CPI of 4%.  Your average tax rate is 25% (meaning you get 75% of your gross pay in the hand) and your EMTR is 50% (meaning that you only get 50c out of every extra dollar of your gross salary). In this case you actually have a pay reduction in real terms. 

You take home pay only increased 2.7% (but the government net tax portion of your income has increased by 8%).

In the above example for your take home pay to keep pace with CPI you need an 6% pay rise.  If you are a lower income earner whose cost of living increases typically exceed CPI, you will need an even higher pay rise just to break even.

High effective marginal tax rates might be a contributing factor in the rise of middle class welfare.  High EMTRs mean that employers payroll costs must grow at a rate much faster than CPI just for employees to break even.  If these type of pay rises are not supported by the real growth of the economy governments may increase welfare to maintain standards of living.  This further increases EMTRs in a reinforcing cycle.

An additional point is that the significant impacts of effective marginal tax rates on changes in take home pay is generally ignored when comparing changes in gross household incomes to the cost of living, or the cost of housing.

In sum, one major economic problem with high EMTRs is that your employer faces a 4% increase in the cost of employing you for a 2.7% increase in your net pay.

Smoking decreases health costs to society

The academic literature generally concludes that smoking reduces health costs to society. This is in stark contrast to commonly held beliefs that there are substantial health care costs borne by society from 'vices' such as smoking, alcohol consumption and fatty foods (which are the target of future regulations).

In fact I will argue (slightly tongue in cheek) that as a society we would be better off if more people would take health risks, and it would be a simple solution to the aged care burden many fear will occur when the baby boomers retire.

The following academic results are typical (my emphasis).

Health care costs for smokers at a given age are as much as 40 percent higher than those for nonsmokers, but in a population in which no one smoked the costs would be 7 percent higher among men and 4 percent higher among women than the costs in the current mixed population of smokers and nonsmokers. If all smokers quit, health care costs would be lower at first, but after 15 years they would become higher than at present. In the long term, complete smoking cessation would produce a net increase in health care costs, but it could still be seen as economically favorable under reasonable assumptions of discount rate and evaluation period.

And from here

Until age 56, annual health expenditure was highest for obese people. At older ages, smokers incurred higher costs. Because of differences in life expectancy, however, lifetime health expenditure was highest among healthy-living people and lowest for smokers. Obese individuals held an intermediate position.

As I have said repeatedly

My core argument in this field has been that increasing preventative health care, while having the benefits of a healthier and long life, often come at increased total lifetime health costs, rather than decreased costs as is often proposed. Remember, we all die some day, and any potential cause of death postponed will allow another to take its place, which of course has its own health costs. Alternatively, a more healthy existence may make us more productive for longer and lead to us contributing more in taxes over our lifetime than the potential increase in health costs which were paid through the tax system for our preventative care.

I argue that most unhealthy vices provide a net benefit to society in terms - they reduce health costs by more than the reduction in tax contributions to health care which may occur due to illness.

The reason is simple. Most of the serious health problems associated with drinking, smoking and obesity take a long time to present. A smoker whose habit had no impact on their lifetime employment, but dies as a result of lung cancer upon retirement at age 65, has still contributed all their lifetime productive efforts to society, including plenty of transfers to others via taxes on tobacco itself, but avoided ongoing health costs from ageing, and costs of the pension.

It sounds cruel, but it is true. The rest of us are better off if people die soon after they retire (unfortunately they are not). The costs of these health vices are therefore borne directly by the people who partake in them, to the benefit of those who choose not to. Perhaps an alcohol and tobacco subsidy is in order?

The only situation where relatively healthy people are worse off from the poor habits of others is if the illness resulting from some unhealthy habit or behaviour occurs early in life and is a barrier to employment and social contribution in general.  In this case the 'unhealthy choice' would result in a massive reduction in their own well-being AND incur costs on others.

The academic literature seems to suggest that this situation is relatively uncommon compared to the alternative, where apparently unhealthy habits do not radically decrease people's productive contributions during their working like.

We can see then that the aged care burden we face is a result of people living healthier and longer lives, especially in the period after retirement. This is mostly the results of better nutrition, lack of war, and importantly, far greater medical knowledge and technology. Unhealthy consumption habits, like smoking, actually have a net effect of reducing the health care burden to society.

As a final note, the amazing gap between academic understanding, public perception, and political ramblings, suggest that taxes on tobacco and alcohol are more about raising revenue than reducing society wide health care costs. The counterintuitive and technical nature these academic conclusions make them easy to keep isolated from policy discussions, allowing politicians to keep any debate at the most superficial level.

*I am not a smoker, but am an occasional drinker, and generally want to live a long time, so I selfishly choose to stay as healthy as I can.

Myth: Tight rental market boosts home prices

A common housing market myth is that low vacancy rates lead to rent increases, which lead to price increases (or at the very least, put a limit on any loss in home values). For example -

...this market imbalance will at some points cause an acceleration in rentals growth and a tightening in rental vacancies, so setting the stage for a recovery in prices through 2012.

Unfortunately, if history is anything to go by, this argument fails in real world conditions.

The two graphs below make the point clearly. In the early 1990s, vacancy rates soared and prices remained flat. But in the early 2000s, rental vacancies matched these highs during the strongest period of price growth observed in 25 years. How can these two opposing relationships been reconciled?

(Images from here and here)

I have a hypothesis.  During boom times overbuilding results in a slight glut in homes entering the rental market (eg 2000-2005). As the construction boom subsides, these homes are slowly absorbed by rental demand. When the market begins to fall (bringing much of the economy with it) potential sellers become reluctant landlords, boosting rental supply (eg 1990-1995). Additionally, nervous householders reign in spending on housing, resulting in an increased occupancy rate and lower rental demand.

There are many ways the occupancy rate increases, which don’t necessarily imply a shortage of homes. Downsizing leads to more efficient use of existing homes -

For example, the parents of a family whose adult children have moved out with friends or partners might find that the upkeep of a large house conflicts with their ‘grey nomad’ retirement plans. They can sell their 5-bedroom house and move into a new 2-bedroom unit, pocketing the price difference for their retirement.

In this scenario the construction of a 2-bedroom apartment resulted in a 5-bedroom home being available to meet the housing needs of population growth.

Other ways include university students moving home with their parents, and grandparents moving in with their children’s families.

If my hypothesis holds, then the ‘rental market cycle’ has two periods for each economic cycle, and tight markets are a signal of a price boom only if the previous trough was prior to a price fall. Therefore our next 'rental market cycle' will be one accompanied by falling prices, or flat at best.  The evidence in Brisbane seems to suggest that this pattern is beginning to occur (although prices have already fallen 10%).

(I also have a suspicion that auction results show a similar cycle - increasing in booms and busts, with low clearance rates at turning points.)

Interesting TEDx video on risk taking and helmets

...and a light read on safety measures that don't work (but probably make us feel better).

Helmet laws hit the headlines - again

The public debate about mandatory helmet wearing laws in Australia has raged since Sue Abbott won an appeal to the District Court last August defending her failure to wear a helmet. Since then media coverage on the matter has been generally poor, often confusing the effectiveness of helmets in reducing head injuries following a fall, with the net social benefits of the law itself.

The debate

The debate is about mandatory helmet laws (MHL). The pro-choice side advocates repealing the law so that helmet wearing is voluntary (not compulsory non-helmet wearing as some mistakenly believe).

The argument is about whether the law itself provides net social benefits – not about whether an individual rider involved in a fall is more or less likely to injure their head by wearing a helmet. Evidence points to the fact that yes, a falling rider with a helmet will, on average, suffer less severe head injuries than a bareheaded rider.

But is this a justification for a law?

Not at all. You see wearing a helmet while walking and driving will also prevent head injuries in the case of an accident. But one side of the debate seems happy to leave these other activities alone, even though it fits logically with their argument. 

Health Costs

Indeed, supporters of MHLs often cite taxpayer-funded public health care as a justification. Yet this makes no sense whatsoever for the MHL debate, the tobacco taxes, or any other preventative health care issue.

As I have said before 

…that increasing preventative health care, while having the benefits of a healthier and long life, often come at increased total lifetime health costs, rather than decreased costs as is often proposed. Remember, we all die some day, and any potential cause of death postponed will allow another to take its place, which of course has its own health costs.

Alternatively, a more healthy existence may make us more productive for longer and lead to us contributing more in taxes over our lifetime than the potential increase in health costs which were paid through the tax system for our preventative care.

Governments, and subsequently economists, worry about these things because many health care costs are borne by others though tax revenue, yet the net economic effect is anything but straightforward.

The arguments 

The only argument remaining in favour of MHLs is that we are saving people from themselves. It is a pretty weak argument for making law in my view.

The pro-choice advocates usually cite a variety of factors to demonstrate that any benefits an individual may receive by wearing a helmet can be significantly offset by their own risk compensation and the changes to the behaviour of other road users.

For example

1. Drivers will pass helmeted cyclists closer than bare headed cyclists (with cyclists with long blonde hair getting the most room).
2. Helmets make cyclists feel safer, and they adjust by taking more risks (risk compensation) 
3. Helmet laws decrease the number of cyclists on the road, making car drivers less familiar with cyclist behaviour and making each remaining cyclist less safe.
4. Helmets increase diffuse axonal injuries of the brain and neck due to their increased diameter (and increased the likelihood of impacts due to the larger volume). As Sue Abbott argued in her court case – a helmet can increase angular acceleration which an oblique impulse imparts to the head, increasing the risk of damage to the brain, especially diffuse axonal injury 
5. Helmets can be a hazard in many circumstances (with many child deaths recorded as a result of helmet wearing) 
6. Any deterrent to cycling is likely to increase time spent on sedentary activities, further contributing to the obesity epidemic. 
7. The law allows governments to appear to be acting in the interests of cyclist safety, while neglecting other measures to improve cyclist safety, such as bike lanes or driver education.

Added together, mandatory helmet laws are not a clear winner on any social benefit measure. 

Missing the point

Most media commentary has missed the point of the debate. The pro-choice side does not argue that helmets are worthless for any individual rider if they are to hit their head. They simply claim that helmets are not as effective at reducing injuries as they are made out to be and the many flow-on social effects that further reduce cyclist safety are not considered.

Even academics have a hard time finding strong evidence that helmet laws have reduced head injuries significantly. The Voukelatos and Rissel paper I referenced in a previous post, showed evidence that the benefits of helmet laws in reducing the ratio of head to arm injuries for hospitalized cyclists was insignificant compared to other road safety improvements in the late 1980s and early 1990s. It was later retracted after criticism over data inaccuracies (corrected data in the graph below), with the critics now publishing their own study using similar statistics to examine the effect of the law in NSW. They find that there is a statistically significant impact of the law in reducing the ratio of head to arm and leg injuries.

Unfortunately, their model also found that the helmet law led to fewer hospitalisations of pedestrians with arm injuries.

For cyclists who do fall in a manner leading to significant injuries, a helmet may reduce head injuries. That it is so difficult to see the effect of helmet laws in the data suggests that any benefits of helmet wearing must be very small, even at an individual level. MHL supporters usually feel that helmets prevent almost all head injuries. But this is not the case. They at best provide a marginal improvement in head safety.

I hope this brings a bit of perspective to the issue for readers who stumble across helmet headlines.

Go back to where you came from - Australia's talking

Last night SBS aired their new three part series Go Back To Where You Came From, where six Aussies take part in a 'reverse refugee' experience.

I think most viewers would agree that it was particularly interesting to watch the participant's reactions to meeting refugees and visiting detention centres.  Participants are from quite different backgrounds, and they have a variety of opinions on refugee policy.

The show apparently has twitter all a buzz, and is generating quite a deal of media commentary.  Much of the reaction focuses on the apparent ignorance of one particular participant to the real situation of refugees - especially in light of their strong opinions on the matter.

My wife suggested that there was a clear pattern in the participant's attitudes - those with broad travel experience seem to have more tolerant views.  It was telling that a couple of participants had never left Australia before the show.

What I found missing from the show, which would have been a nice complement to the emotional dimension, is reference to the actual statistics on refugees, their country of origin, the proportion coming by boat, and the changes in refugee numbers over time.

This is important because the public debate usually overlooks a couple of key points.

1. Boat people are a minority of asylum seekers and a tiny fraction of total immigration (graph below from here)

2. The number of asylum seeker arriving in Australia correlates strongly with global numbers, suggesting that it is not so much the policy of the destination country that influences the number of arrivals, but the situation in the country of origin (see the graph below).

For more detailed analysis of the factors involved in refugee outcomes, read this detailed article. I look forward to the follow up episodes tonight and tomorrow, and recommend the program to anyone even slightly interested in the topic.