The poverty curve

In Bill Gates' annual newsletter, 2014 edition he addresses 3 myths about global poverty.

1. Poor countries are doomed to stay poor
2. Foreign aid is a big waste
3. Saving lives leads to overpopulation

All 3 sections are interesting; I have mixed thoughts about number 3, though noticed after I had read it that is was authored by Melinda Gates.

In the first section Gates has a graph on the change in individual wealth of the world over the last 50 years. Note that the x-axis is logarithmic.

This comes about through productivity and trade. Wealth is created. Some of the big drivers of productivity include rule of law, private property, and low tax which suggests that had we seen optimum conditions in the last few decades we may have seen even better improvements. Still, this is encouraging. Gates mentions several other drivers such as health and education. Not mentioned (as a cause of productivity) is energy which allows significantly more output per person. The climate change paragraph can be ignored, developing nations need cheap and reliable energy, not low carbon energy.

Public health and unwarranted conclusions

I read a fair number of articles. This is a paraphrase of one but I have seen similar several times over.

In conclusion our results show that exposure X is associated with significantly increased chance of outcome Y. Public health recommendations/ government agencies should reduce/ ban exposure X.

(This is assuming outcome Y is bad, a converse argument could be made if exposure X is thought to be good.)

This is frustrating for several reasons.

1. The word significantly usually means statistically significant. That is the authors are confident that the association they have found is a real one, not a chance one. For various reasons I think many findings that are claimed to be real are actually chance, so I may not be convinced the statistics justify the conclusion. But assuming the statistics do justify it, the significance relates to the degree of confidence in the result, not the size of the result. The phrasing "significantly increased chance" sounds like the size of the association is strong. It may be minor. A risk ratio of 1.003 (1.002–1.004, p <0.001) is statistically very significant but not functionally significant. Even a risk ratio of 3 (i.e. you are 3 times more likely to develop outcome Y) may be irrelevant if the outcome is extremely rare. Does it really matter if you increase your risk from 1 in a million to 3 in a million?

2. Association is not causation. Yes it may be a real effect, and it may be a relevant one, but it still may just be an association. We need to establish causation. Addressing a problem if it is causative may not resolve it. Addressing an association that is not causative definitely will not resolve it. And it could potentially worsen it. We need studies that show definite causation. Then we need studies that show intervention to reduce exposure X actually reduces outcome Y.

3. Nothing in the research relates to public policy. The study does not show that the policy was enacted and was effective. Even convincing knowledge that reducing outcome Y by preventing public exposure to X does not imply anything should be done by the state about X. Should the state ban hang-gliding because it is associated with increased mortality? Should it make every vice illegal because of detrimental effects on self? And there are further question about enforcing a ban. What about the monetary cost considerations? What about liberty? Will the unintended consequences be worse than the problem? Perhaps all that is warranted is education. For example the government can mandate labelling without banning a substance.

I am not arguing against any public policy. It just seems that socialism is so embedded in some people's psyche that new information to them logically implies government intervention.

Candy Irony

Candy Dynamics sell the Toxic Waste brand of candy bars which they import from Pakistan. Their Nuclear Sludge variety has tested high for lead prompting a recall.

A recent test performed by the California Department of Public Health has indicated that a lot (#8288A) of the cherry flavor of the above-listed product contains elevated levels of lead (0.24 parts per million; the U.S. FDA tolerance is 0.1 ppm) that potentially could cause health problems, particularly for infants, small children, and pregnant women.

Out of an abundance of caution, the company has determined to recall all lots and all flavors of the product distributed from the product's inception in 2007 through January 2011.

The products are identified as: Toxic Waste® Nuclear Sludge® Cherry Chew Bar (UPC 0 89894 81430 6), Toxic Waste® Nuclear Sludge® Sour Apple Chew Bar (UPC 0 10684 81410 7), and Toxic Waste® Nuclear Sludge® Blue Raspberry Chew Bar (UPC 0 89894 81420 7). Each chew bar has a net wt. of 0.7 oz (20 g).

No other "Toxic Waste®" brand product is affected by this recall.

No illnesses have been reported to date in connection with this problem.

This is unlikely to be of significance, maximum allowances are likely to be well below any level that causes harm. Still, somewhat amusing.

Gapminder: Resource to compare countries

I have recently become aware of an interesting Swedish site called Gapminder. They have aggregated a large amount of statistics and allow you to plot them in a variety of ways. It covers health, income, disease and mortality, and several other parameters.

What I found of interest was their income and life expectancy graphs on Gapminder World. The basic frame is graph with a variety of options.

Life expectancy is on the left. It is a linear scale, though this can switched to a log scale. Income is on the bottom defaulting to a log scale (adjustable to linear). And the country size is plotted according to population. This can be adjusted to a fixed size, or a host of other variable such as aid received, education, energy use.

The year can be adjusted, or you can watch the changes unfold over the last 200 years. Countries can be highlighted for comparison.

And you can track changes of several countries over time.

What is of some interest is that life expectancy increased before income in many of the countries. I also found it interesting to note the slow economic progress or even regress of countries with poor governance compared to reasonably rapid gains in countries where there is rule of law and economic freedom. Compare Zimbabwe to Botswana and note the rise of Singapore and Hong Kong, neither of which have many natural resources, or even land or water.

Of course one is limited to the accuracy of the data, data from some countries is far more accurate than others.

Responsibility and costs for healthcare

There is much in the news about healthcare due to the current political tinkerings in the US. I have my ideas about structuring a health system, somewhat based on my experience and what I have learnt about the Singaporean system.

There are probably several reasons why costs are so high in the US (and elsewhere). One significant reason is litigation. Until the legislators are willing to cap "compensation" payable then things are unlikely to get cheaper. The argument that someone is entitled to such extreme amounts of money because he suffered harm neglects that these costs will be recovered from future patients. Indemnity insurance in Western, non-US health is significantly less. Consider perhaps $NZ 1000–2000 per annum in New Zealand to > $US 100,000 in the United States.

This aside, if one is to judge healthcare proposals there are several issues around health and economics that need consideration.

Who pays?

I care little if you wish to spend your money on Kopi Luwak? I am content with the brew at the local cafe, or even instant coffee. But I care a lot if you spend my money on such. It doesn't matter if nationalising health is cost neutral. If that cost is now paid by government and not individuals, then the government obtains that money from individuals via tax. Thus I am forced to cover your health costs, even if I made health choices intended to save me money long term. But the government payment of costs is even more complicated. There is no fixed amount of money. Individual payments to healthcare have different effects on the economy to government payments. Government removal of money from the economy decreases productivity; thus even a cost neutral program will leave the economy poorer in production, thus individuals worse off.

Who gets sick?

It is said that 75% of health expenditure is spent in the last 5 years of people's lives. Now that is not every person, but it is a lot of them. We don't know when we are going to die, and the increased medicalisation of our lives means that much is spent as we age and become increasingly infirm and near the end of our life.

This may be money well spent. It may maintain people's independence. And if it is your own money then do as you will. But as one becomes increasingly unwell he should be assessing his life and making plans for when he dies. And increasing medical possibilities potentially means increased costs. Yes you can spend the last 15 days of your life half conscious hooked up to a respirator in the Intensive Therapy Unit (ITU) with metastatic cancer that has failed 3 previous therapies, but I am not certain this is a gain over symptom control at home or in a hospice. Don't get me wrong, these things can be hard to predict, and ITU is often appropriate, even if ultimately unsuccessful. And these are decisions for individuals, and we all weigh different options differently. But people need to think about their own mortality.

How many people have high costs?

Following on from costs occurring at the end of life, it needs to be remembered that this is a cost that a high proportion of people will face. Rare expensive events can be covered by spreading the risk, that is how insurance works. But if events become common then every person has to cover their own cost. If we are spending $400,000 over the final years of life for say every second person (the other dying in their sleep after a long healthy life incurring minimal costs), that means we need to collect ~$200,000 per person to cover end of life expenses. People in countries with socialised medicine talk of how they have paid their taxes all their life, and this is true (though some goes to education and roading, and some is frivolously wasted), but I think they are unaware of how quickly the tax that they have paid is used up.

Personal responsibility

Our own health should be something we take responsibility for. It is true that many things happen that we have no control over: accidents, infections, diseases, poor genetics. But much is known about healthy lifestyle. Debates around increased risks of the order of 10–100% are difficult to confirm. But it is clear, for example, that smoking tobacco, morbid obesity, and base jumping do not usually improve one's health. Healthcare provision should take into account personal responsibility. (Though I think this intrinsically, I am aware that it can become politicised with claims that "politically incorrect behaviour" is "unhealthy").

The problem with this is people often don't take responsibility. And when they subsequently get unwell it seems unkind to leave them to suffer the consequences of their decision. Mercy is toward the undeserving.

Isn't health provision caring for those who are suffering?

Yes, and no. It is true that we should think of caring for those who suffer, at least from a Christian perspective. But previously this has been about providing food and shelter, giving clothes, comforting people in pain. This is something that most individuals can give to another. But modern health care is more extensive. We can diagnose and treat many illnesses, often at significant monetary cost. It is not something that most individuals can do, and it is hard for many individuals to cover the cost of doing so for others. I applaud hospital and charity work in the third world. I think these are excellent endeavours, frequently they do much good for minimal cost. But I am cautious when people say that everyone is entitled to healthcare irrespective of cost. Someone does have to pay and there are competing claims to money. The statement that no cost is too great for good health is just not true.

Health insurance?

People need to rethink this. Healthcare costs are somewhat foreseeable. We can budget for routine doctor visits, prescription costs, spectacles, regular dental care. It should be possible to assess approximately how often people and families need to attend to their health, and budget for this, with a moderate margin. Insurance is not about these costs.

You don't get insurance to cover the cost of your mortgage each month. If you did, the company would pay your mortgage, and your premiums would cover that plus an overhead. It would be more expensive. Health insurance should be insurance. It is to cover that which you do not expect to use it for. Insurance should have a high excess (save up the excess in discounted premiums and keep it in a bank account) and cover rare (and expensive) events that do not happen to most people. That way you have cheap insurance as an insurance, and you don't pay overheads for costs that you know are going to be regular.

Health care adherence and responsibility

A recent study published in PLoS Medicine confirms what has previously been thought by those working in African health field. Adherence to antiretroviral (Anti-HIV) medication in Africa is high (> 90%) and higher than the West. After several hundred interviews the reasons were documented. Summarising the responses,

[Medication] adherence through a number of deliberate strategies aimed at prioritizing adherence: borrowing and “begging” transport funds, making “impossible choices” to allocate resources in favor of treatment, and “doing without.” Prioritization of adherence is accomplished through resources and help made available by treatment partners, other family members and friends, and health care providers. Helpers expect adherence and make their expectations known, creating a responsibility on the part of patients to adhere. Patients adhere to promote good will on the part of helpers, thereby ensuring help will be available when future needs arise.

I am not certain how robust this study was, nor whether other factors (age, religion, sex, income, etc.) were adequately addressed, but it is interesting that the high adherence was due to perceived responsibility of action.

Relationships confer responsibility in addition to providing resources. Recipients of help must recognize what they receive and reciprocate. To ignore these responsibilities is to risk resentment on the part of helpers.

It is their sense of duty within the domain of immediate family, supporters and community. There is no disconnect created by involving the government as an intermediary. When others help directly there is a sense of reciprocality. Not necessarily that one has to do the exactly the same as others, rather that one should do his duty in response to others.

I believe that when antiretrovirals have been scarce or expensive, the choice to medicate one patient funded by other patients who were not receiving treatment meant that the survivor would likely care for the non-survivors' kin. So an expectation is on the treated person to comply so that he will indeed survive.

Interestingly another article in the same edition compared health outcomes in children for user pays medicine and free health care. There was no difference in rate of illness. The authors wrote,

This lack of any effect of being randomised to free health care on health outcomes is unexpected, since there has always been an assumption that increased access as a result of free health care improves health.

Though the free care group used significantly more services!

People are capable of caring for themselves (if they have the resources). They are able to be responsible for their care and respond to reasonable community expectations.

Government intervention is mostly unnecessary—at least in illness; however it is likely the attribute of self care and responsibility extends beyond health behaviour.

(Not to mention the government cash savings by leaving the funding private. And, as usual, private investment is more efficient: in this case same outcome for probable lower cost—I am using attendance as a proxy for cost.)

Health outcomes and inequality

A difficulty I have with socialist interpretation is their ideology is such that good outcomes are made to appear bad if they arise within capitalism. While they may claim their focus is different (equality), I have a hard time believing that socialists would abandon their political philosophy if history showed socialism was more wealth generating than the alternatives. Rather, the better living conditions would be evidenced that socialism is preferable to other systems.

If capitalism cannot be said to decrease wealth, attributing it blame for causing poverty can still be attempted.

An article evaluating the health system in Chile was published in April. It blames government policies for creating inequality in health care, and states that this causes

decreased solidarity between rich and poor.

While I disagree with the article, I will not do full justice to its subtleties here and it is best read to appreciate its point of view. Nevertheless, it apportions blame where there likely is little. It also seems to imply that had Chile not enacted its policies then the (perceived) negatives would have been avoided without having any significant effect on the positive outcomes of reform; a dubious claim.

Chileans have compulsory income contribution to health though they can choose public or private systems. A public system meant that all citizens had access to health care including the poor. The article gives examples of funds transferred from private to public which it views as inappropriate. This may be valid but without more detail I will reserve judgment. Other complaints include:

...access to care is somewhat equitable, but quality of care is not.

[Private] insurance policies for women of childbearing age may cost four times more than men's policies.

...adjusted mortality for the poorest decile is 6.0 out of 1,000, versus 4.8 for the richest.

[Private health care spends] ten times as much on administration per member and about two times as much on health care per member than [the public], even though [private] members are in better health and need less care.

But is this not to be expected?

In fact it is partially surprising that access to care is equitable. It is good that it is, but it is possible that things could be otherwise.

It is not surprising that quality is higher for a private system. It often times has more funds for this. All systems always have limits based on funding; if a private system has more funds it can do more tests, it can do more operations.

Policies for childbearing women will be more expensive. Men of that age have minimal health costs and therefore minimum premiums. Women have higher health costs due to childbirth. A health policy that excluded childbirth and related costs would be cheaper and possibly similar to men's. Do not insurance companies charge higher premiums to younger drivers as they have a higher risk? Anyway, if most women are married then the household costs for health care may be similar across families.

Mortality may be related to income even in countries which have predominantly public health, this may not be related at all to the private/ public differences; I will come back to this.

The paper complains about the money spent on administration and health care, yet notes this about the public system:

Access to specialists appeared deficient in the public service: there were waiting lists of up to four years in some specialties, such as ophthalmology.

Yet elsewhere claims the private system is inefficient. It may just be that prompt access requires a little more cash which the private system has. If the private system had waiting lists of 4 years despite more money then it is inefficient, but it doesn't.

And so enamoured with the state that anything that "sounds" like a private monopoly is bad.

Some [private] companies are in an oligopoly position: the main three firms share close to 80% of the market.

But government monopolies are good and could never be the source of inefficiency.

A logical recommendation for a health system counter-reform in Chile would be to move from a multi-insurer scheme to a single public insurer scheme, as in Costa Rica.

Despite all these claims in the discussion the authors are forced to admit,

Chile's health indicators are good, compared with countries with a similar gross national product, such as Colombia and Argentina. These results are due in part to Chile's high economic growth rate and a spectacular reduction in poverty. Further significant factors are the high proportion of the population with access to drinking water and sanitation, and the high adult literacy rate and education level.

They claim this is as the result of public policy:

But relatively equitable access to health care, mainly through public health insurance and public health services, which cover 80% of the population, played a major role in this achievement, as did maternity and child health protection programmes implemented through the public health network.

Presumably the achievement they are talking about is health indicators; not economic growth, reduction in poverty, water, sanitation, literacy. But of what evidence? Health outcomes on a population basis are better correlated to these latter issues than to individual procedures on sick individuals. As important as hospitals are it is unlikely they make nearly the same impact on life expectancy as more basic provisions and decrease in poverty.

And this is where the issue lies. If economic policy is to minimise government expenditure because that leads to economic growth, then this economic growth may be what has made the difference. Who cares how much people spend on private health care? It is a private expenditure. It is less money for the government to outlay which leaves more money in the hands of the producers. If the producers then create wealth, what does it matter that some of it is spent on health? Even if only 25% of the population contribute to the private system, that is still a large number of people the government no longer needs to pay for.

The bigger issues are:

1. Are the poor worse off than before?
2. Is the number of poor persons less than previously?

If the poor are no worse off (or even better off) and the middle class has grown as the lower class rises out of poverty then what is not to like? These are good outcomes.

To complain there is more inequality because not everyone has yet escaped poverty and some are excessively rich is a politic of envy. Yes the rich man has an individual responsibility to care for his poor neighbour, but income inequality is not a bad in and of itself. Oppressing the poor to become rich causes inequality but it is the injustice of oppression that is wrong.

It seems to me that socialists want equality of outcome so much that it is better that all live in abject squalor than most have a modest income, if that modest income means an associated variable income.