God, Immortality and the Futility of Life

William Lane Craig has a pretty dispiriting take on the atheistic view of life:

If there is no God, then man and the universe are doomed. Like prisoners condemned to death, we await our unavoidable execution. There is no God, and there is no immortality. And what is the consequence of this? It means that life itself is absurd. It means that the life we have is without ultimate significance, value or purpose. 

(Craig 2008, 72)

Embedded in this short quote are a number of important claims. The first is that in order to avoid futility and meaninglessness we need our lives to have ultimate significance, value and/or purpose. The second, perhaps more important, is that we cannot have these things unless two conditions are met:

Craig’s Two Conditions for Meaning: Our lives are without ultimate significance, value or purpose unless (a) there is a God (who, among other things, determines objective value, purpose and significance); and (b) we are immortal.

Over the years, I have written several pieces that call into question Craig’s claims about futility and the atheistic view. In particular, I disagree with his claim that God determines objective value, purpose and significance. I also disagree with the notion that “ultimate” significance is absent on the atheistic view, or that we need it in order to avoid futility.

Today, I want to consider another criticism of Craig’s view. This one comes from Toby Betenson who has just published a nice little article in the International Journal for Philosophy of Religion entitled “Fairness and Futility”. In it, Betenson mounts an internal critique of Craig’s view of God and the futility of life. In short, he argues that if we take Craig seriously in his views about God, it turns out that God actually undermines the meaningfulness of our lives. Betenson then uses this internalist critique to suggest that Craig’s focus on ultimate significance (as opposed to relative significance) is misconceived.

I want to go through the main phases of this argument. I have my own thoughts on it and I will insert those into the discussion as I proceed.


1. How might our lives be futile?
Betenson’s argument works by paying close attention to what Craig himself says about the futility of life under atheism and the meaning of life under theism. This close attention is what allows him to hoist Craig on his own petard. This means that several portions of Betenson’s article are given over to quoting Craig verbatim and then drawing out the implications of what Craig is saying. I don’t want to follow the same approach here, since that would end up with me simply repeating what Betenson has to say, but I’ll have to do a little bit of it in order to illustrate a key point in the dialectic.

This key point relates to Betenson’s claim that Craig’s account of meaning under theism implies the existence of a “causal relevance” condition for meaning. In other words, in addition to the two conditions stated above (viz. the existence of God and personal immortality), Betenson thinks that Craig’s account of meaning implies a third condition. This third condition stipulates that in order to live a meaningful life, one’s actions must make a causal contribution to events that are of ultimate significance. Betenson cites some passages from Craig in support of this. In these passages, Craig speaks repeatedly of the need for one’s life to ‘make a difference’. Here’s an example:

The contributions of the scientist to the advance of human knowledge, the researches of the doctor to alleviate pain and suffering, the efforts of the diplomat to secure peace in the world, the sacrifices of good people everywhere to better the lot of the human race—all these come to nothing [on the atheistic view]. In the end they don’t make one bit of difference, not one bit. Each person’s life is therefore without ultimate significance. And because our lives are ultimately meaningless, the activities we fill our lives with are also meaningless. 

(Craig 2008, 73)

Craig is here talking about the difference between a life filled with actions that make some relative difference (i.e. that have value relative to a particular person or time frame) and those that make some ultimate difference (i.e. have value from the point of view of the universe). The latter are what matter when it comes to meaning. So this gives us Craig’s third condition for meaning in life:

Craig’s Implied Third Condition for Meaning: Our lives have meaning (and they avoid futility) if our actions make some causal difference to events of ultimate significance, value or purpose.

It’s important to realise how this affects Craig’s complete account of meaning in life. It means that in order to live a meaningful life, you must have three things. First, you must have a God that brings into existence objective values. Second, you must live forever. And third, your actions must be causally relevant to events of ultimate significance (or ultimate “value” if you prefer).

There are already some peculiarities in this account. For instance, one could dispute the role of God in bringing into existence objective values. One could also puzzle over the importance of personal immortality in this account (why must I hang about forever in order for life to have meaning? Isn’t it enough for my actions to make a difference to something of ultimate value?). I puzzle over these things myself and as we shall see such puzzles reemerge in Betenson’s argument. But set them to the side for now.

Let’s focus, instead, on the consequences of Craig’s third condition. Granting its coherence, there are at least four possibilities when it comes to the state of the universe and the satisfaction of that condition. They are:

• (i) The universe could contain events of ultimate significance/value, and our actions could make a causal difference to them.

• (ii) The universe could contain events of ultimate significance/value, and our actions could make no causal difference to them.

• (iii) The universe could contain events of mere relative significance/value, and our actions may make a difference to them.

• (iv) The universe could contain events of mere relative significance/value, and our actions may make no difference to them.

These four possibilities are further illustrated by the diagram below.

For Craig, only the first of these possibilities allows for a meaningful life. The other three all lead to futility. Craig is well aware of the futility of (iii) — that’s what he is talking about in the passage quoted above, and that’s what he finds so problematic about the atheistic view. The fourth possibility (iv) is an example of what Betenson calls “futility overkill” since it involves a double failure. We can ignore it. The second possibility (ii) is the interesting one. Betenson wants to argue that Craig’s understanding of God actually supports this view. If so, then even by Craig’s own standards, the theistic view implies futility.


2. Betenson’s “Theism implies Futility” Argument
The argument for this is actually quite simple. It comes, once again, from paying close attention to what Craig says about God. As I have noted on previous occasions, one of Craig’s favourite critiques of the atheistic view is that it does not allow for “ultimate accountability” (or “ultimate justice”). In the atheistic universe, we can perform moral acts throughout our lives, including acts of great personal sacrifice, but in the end these acts count for nothing because they will not be rewarded or punished. We could have lived lives of debauchery and sadistic cruelty and the result would be the same.

In the theistic universe things are different:

[O]n the theistic hypothesis, God holds all persons morally accountable for their actions. Evil and wrong will be punished; righteousness will be vindicated. Good ultimately triumphs over evil, and we shall finally see that we do live in a moral universe after all. Despite the inequities of this life, in the end the scales of God’s justice will be balanced. Thus, the moral choices we make in this life are infused with an eternal significance. We can with consistency make moral choices which run contrary to our self-interest and even undertake acts of extreme self-sacrifice, knowing that such decisions are not empty and ultimately meaningless gestures. Rather our moral lives have a paramount significance. 

(Craig - quoted in Betenson 2015) 

Betenson thinks that there is something odd going on here. Craig seems to be saying that there must be ultimate justice in order for any of our actions to have moral significance (it’s the only way we can be sure that we “live in a moral universe after all”). He also seems to be saying that God is the only thing that can supply the necessary justice. He ensures that the moral scales will ultimately be balanced. The net result of this, according to Betenson, is that our actions don’t seem to be all that relevant on Craig’s account of ultimate significance. It is God as justice-giver, not us as moral decision-makers, who “makes the difference”. In short, if Craig is right about the importance of God in ensuring ultimate significance, he implies that our lives are not causally relevant to events of ultimate significance. But that was one his conditions for meaning in life. We have a contradiction.

To put this a little more formally:

• (1) If our lives are not causally relevant to events of ultimate significance, then our lives are futile (devoid of meaning).

• (2) If God exists, then God ensures that the universe has ultimate significance (by ensuring that it is ultimately just).

• (3) If God ensures that the universe is ultimately just, then our lives are not causally relevant to events of ultimate significance.

• (4) Therefore, if God exists, our lives are futile.

Is this a persuasive argument? Let’s consider some objections and replies.


3. Objections and replies
Betenson covers four objections in his article. The first two strike me as being fairly uninteresting. The last two are more important. I’ll briefly summarise what he has to say about each of them. Here’s the first objection (note: they are not set out exactly like this in the original article):

Objection 1: Even if God is responsible for the ultimate balancing of the moral scales (the ultimate Good), that does not mean that our lives are not causally relevant to lesser, but still significant goods.

This is an obvious criticism because Betenson’s critique seems to take an overly narrow view on the range of goods to which our moral choices must be causally relevant. It seems plausible to say that the decision of the doctor to cure a patient is a causal contribution to the good, even if it is not a contribution to the ultimate Good. Betenson argues that this objection is not available to Craig. This is because Craig repeatedly emphasises that a causal contribution to events of localised significance (like the curing of a patient) are not enough. Contribution to those sorts of events is possible on the atheistic view and Craig clearly doesn’t think it is enough.

The second objection is slightly more interesting:

Objection 2: On Craig’s view, our actions do have ultimate significance because our actions determine whether we go to heaven or hell, and since we live forever in heaven or hell, this means that our actions have everlasting moral implications.

But there are problems with this too. For starters, one has to accept a very traditional view of heaven and hell to find it at all persuasive (if you are a universalist, this option won’t be available to you). For another thing, it assumes that a causal contribution to an event of everlasting personal significance is equivalent to a causal contribution to an event of ultimate significance. This equivalence is dubious. As Betenson puts it, when assessing events in terms of their ultimate significance, you have to take the “point of view of the universe” (i.e. a depersonalised point of view). From that point of view, it is not at all clear that whether you get into heaven or hell is ultimately significant. It is important to you, for sure, but not from the universal point of view. All that matters from that point of view is that the moral scales are balanced: that good is rewarded and wrongdoing is punished.

As I say, I don’t find these first two objections to be particularly important. It is with the third objection that things really start to heat up:

Objection 3: Betenson’s argument assumes that ultimate reward and punishment are what matter from the universal perspective; but this is wrong. What matters is whether objective good is done. There are other objective goods and we can, as moral beings, contribute to the total amount of objective good.

This is a modification of the first objection, but it is an important modification. It highlights problems in Betenson’s original discussion of the ultimate good. It suggests that, from a universal point of view, ensuring that people are appropriately rewarded and punished is just one good among many. There are other potential, universally significant goods (e.g. pursuit of knowledge and understanding, or, if you are a Christian, the worship of God). We can causally contribute to the total sum of those other goods.

As it happens, this is pretty much my view of meaning in life (except I do not think the worship of God is an impersonal good). The problem with it is twofold. First, it is not at all clear why immortality would be required on this account. I could make my contribution to the pursuit of knowledge and understanding, then shuffle off this mortal coil and drift into eternal oblivion. The total amount of good in the universe would still have increased through my actions. Second, it is not at all clear why God is needed on this account. Craig thinks He is needed because God determines what is and what is not good, but if like me you think that standards of goodness must be independent of God the need for His existence dissipates.

The problems with this third objection lead to the construction of a fourth:

Objection 4: In line with objection 3, what matters from the universal perspective is that good is done and this conception of good extends beyond ultimate justice. But it just so happens that the most important objective goods require that you live forever.

This is effectively a slight modification of objection 3 that tries to explain why personal immortality is needed in order to avoid futility. It may seem a little ad-hoc right now, given the order in which I have presented these objections, but there is some independent motivation for it. There are those — and I believe Craig is among them — who think that eternal salvation or eternal worship of God are the most important objective goods. And it just so happens that both of those things seem to require personal immortality.

The difficultly with this objection is twofold. First, it doesn’t rule out the possibility of there being other objective goods to which we could causally contribute and which could provide some degree of meaning to our lives. Second, there are serious questions to be asked about whether these objective goods really are objective goods. I and others have written about the dubious nature of the obligation to worship God. It also seems pretty strange and arbitrary for God to create us and then decree that the ultimate objective good — the one which will allow our lives to have meaning — is our willingness to worship him forever. As Maitzen points out in another article on this topic, it is doubtful whether such divinely ordained purposes could rescue our lives from futility. They will always seem questionable and unsatisfactory. This leads to some doubts about the very notion of “ultimate” significance.

I tend to agree, but that’s where I shall leave it for now. To briefly recap, Craig holds that God, personal immortality, and a causal contribution to events of ultimate significance, are needed if we are to avoid a futile and meaningless existence. Betenson counters that on Craig’s conception of God, we make no causal contribution to events of ultimate significance. This is because God is the ultimate justice-giver, who balances the universe’s moral scales. There are responses to this, but they tend to create more problems for Craig’s account.

Should prospective parents have to apply for licences? An Ethical Debate

Should prospective parents have to apply for parental licences? The argument seems obvious. Having children is a serious business. Negligent or irresponsible parents risk causing long-term harms to their offspring, harms that often have spillover effects on the rest of society. A licensing system should help us to filter out such parents. Therefore, a licensing system would benefit children and society at large. QED

Of course, I’m being somewhat facetious here. The idea of prospective parents applying for parental licences will strike many as both absurd and offensive. But there is no idea so absurd and offensive that at least one philosopher has not defended it. And when it involves something as contentious as parent-child relationships, you can rest assured that there will be more than one.

In this post, I want to review the philosophical debate about parental licensing. I start by looking at Hugh LaFollette’s now-classic argument in favour of parental licences. This argument was originally published in 1980 in Philosophy and Public Affairs. It has since been the subject of much commentary. After outlining LaFollette’s argument, I want to consider a recent critique of that argument from Jurgen de Wispelaere and Daniel Weinstock. As we shall see, they claim that while there is something to be said for educating prospective parents and avoiding risks to children, it is not at all clear that licensing is the best way to go about doing this.

Let’s see how the argument develops.


1. LaFollette’s Argument for Parental Licensing
LaFollette’s argument for parental licensing was based on a simple set of analogies. Modern societies routinely rely upon licensing regimes. The most obvious example is the licensing regime that governs the driving of cars. If you want to drive a car, you must earn a driving licence. This involves passing a test of knowledge and skill that establishes a minimum level of competence. The same holds true in other areas. If I want to practice brain surgery, I have to earn a medical licence. If I want to practice law, I have to earn a law licence. And so on.

What is the rationale for all these licensing regimes? The answer seems pretty straightforward. Activities like driving and brain surgery carry significant risks to the rest of society. A car is a dangerous weapon in the wrong hands. Likewise a scalpel. We can reduce these dangers by having a testing and credentialing system that establishes basic levels of competency. The benefits of such policies more than outweigh the risks. This chain of reasoning may not hold true in all cases. The cost of licensing lawyers, for instance, may not outweigh the benefit. But that doesn’t matter right now. What matters is that it holds true in some cases.

That is all that LaFollette’s argument requires. Because if it does hold true in some cases it should also, by analogy, hold true in the case of parental licensing. Having and rearing children is a risky business. Incompetent parents risk causing significant long-term harm to the their children. A licensing system would train parents to obtain a basic level of competence and would filter out those who pose a serious risk to children. The benefits would outweigh the costs.

In other words, the following argument from analogy seems to hold:

• (1) Licensing systems are desirable/justifiable in the case of driving, doctoring or lawyering on the grounds that (a) those activities pose significant risks of harm to others; and (b) a licensing system would help to reduce those risks in a cost-effective manner.

• (2) Reasons (a) and (b) would also be true in the case of parental licensing.

• (3) Therefore, parental licensing would be desirable/justifiable.

Like all arguments from analogy, this has some holes in it. Two are particularly important. The first is the strength of the analogy itself. It could be that there are important differences between parenting and driving that justify licensing in the latter but not in the former. The second has to do with implementation. It could be that there are significant differences between driving and parenting when it comes to the practical implementation of a licensing system. In particular, the costs associated with licensing parents might be significantly higher than the costs associated with licensing drivers.

Let’s consider both of these problems in turn.


2. The Non-Substitutability of the Parental Good
Consider first the important disanalogies between the activity of driving (or performing surgery or whatever) and parenting, particularly in relation to the value of those activities to those who are unjustly denied their licences. No licensing system is perfect. Every system is likely to issue “false positives” and “false negatives” (i.e. likely to deny licences to those who are genuinely competent and grant licences to those who are not). The question is whether we can tolerate those imperfections. To answer that question we need to think of the nature of the licensed activities.

Driving is, for most people, a useful (occasionally necessary) convenience. It gets them about from A to B. This may have considerable value in their lives, but if they really need to get about, there are often reasonable substitutes to driving themselves. They could ask someone for a lift, or they could use public transport, or they could cycle or walk. Some of these will be more reasonable than others, depending on the circumstances. The important point is that — with some limited exceptions — there are always other ways of getting from A to B. If someone is denied a driving licence, they are not damned to permanent house imprisonment.

Something similar is true of performing surgery or practicing law. Those activities will be valuable to many people. It is true that some of those people may have had their hearts set on a career as a doctor or lawyer — that it is their dream job — but there are, nevertheless, alternative forms of employment available to them, many of which might end up being equally (if not more) fulfilling. So, once again, there are reasonable career substitutes available to those who are denied their medical or legal licences.

Is the same thing true of parenting? De Wispelaere and Weinstock argue that it is not. They argue that there certain aspects of the parental good (i.e. the relationship between parent and child) that render that good non-substitutable. If people were denied this parental good thanks to the vagaries of a licensing system, they would lack the reasonable substitutes available in other cases. So what is it about the parental good that makes it non-substitutable? They put it like this:

First, children are utterly dependent on parents for securing their basic needs and wellbeing, and therefore vulnerable to the choices and actions of parents. Second, children typically have no right or power to exit the relationship. Third, children’s love for their parents is spontaneous and unconditional; at times it simply defies ‘rationality’. Finally, the parent is explicitly charged with the responsibility of taking care of the child. Because of the form and quality of the intimacy associated with raising a child, many individuals develop a very strong interest in experiencing such a relationship. 

(De Wispelaere and Weinstock 2012, p 199)

When you add the distinctive nature of the parental good to the fact that any licensing regime with a sufficiently large reach is likely to issue a high number of false positives, you have a situation in which those false positives start to look less tolerable. For example, if a licensing regime successfully authorised 99% of all competent parents and only denied 1% of competent parents, you still have a significant problem if the regime is used to licence every potential parent. Say you have a population of one million potential parents. Even with a 99% accuracy rate, you would be denying parenthood to 10,000 competent parents. These parents would not be like unlucky drivers who can hop on the next bus.

To sum up:

• (4) Every licensing system will be imperfect, i.e. it will deny licences to those who deserve them and grant them to those who don’t. The question is whether this imperfection is tolerable.

• (5) If the goods that deserving people are denied through the imperfections of the licensing system are reasonably substitutable, then some imperfections are tolerable; if the goods are not reasonably substitutable, then the imperfections are less tolerable.

• (6) The goods associated with traditional licensing schemes (e.g. driving, performing surgery) are reasonably substitutable; the good associated with parenting is not reasonably substitutable.

• (7) Therefore, the inevitable imperfections of a parental licensing system are less tolerable than the imperfections of typical licensing systems.

I have relatively little to say about this argument. I agree with the De Wispelaere and Weinstock that the parental good is non-substitutable. Nevertheless, I would make two points. First, it is not clear to me that the good associated with particular careers are as reasonably substitutable as they let on. In particular, I think there may be unique goods associated with, say, the act of saving someone’s life through surgery that are not easily replicated by other career paths. Second, I think the properties of the parent-child relationship that make it non-substitutable cut both ways. In other words, the dependency and vulnerability of children might be part of what makes parenting so special, but those properties are also reasons to worry about the harms of incompetent parenting. Indeed, they may even provide stronger grounds for licensing than would apply in the case of driving or performing surgery.


3. Implementation Problems
Let’s set aside the disanalogies for now and assume that the prima facie argument for licensing goes through. What would such a licensing system look like? What would be its practical costs? Even in his original paper, LaFollette acknowledged that it would be very difficult to implement a restrictive licensing regime. Unless you are willing to allow mass sterilisations and or invasive clampdowns on procreative sexual behaviour, the regime would probably be highly ineffective. People who would not pass the licensing tests would still be allowed to conceive and bear children.

Once the children are brought into existence, various options would arise for taking them away from irresponsible parents, but this may well have undesirable effects. The adoption system is unlikely to have the capacity to absorb such children, which will put pressure on institutional or foster care systems, which are often bad for child welfare.

These practical difficulties have led LaFollette to revise his proposal. In a 2010 article, he suggests that a licensing regime should not focus on punishing or restricting unlicensed parents but should, rather, focus on rewarding or incentivising licensed parents. In particular, he suggests a system of tax breaks. Parents who are willing to undergo the training and pass the basic competency tests should be rewarded with some tax break. This would avoid the significant costs associated with a punitive regime, while still retaining some of the benefits of a licensing regime.

The problem with this, as de Wispelaere and Weinstock point out, is that it leaves the licensing proposal in tatters. If the idea is that educating prospective parents has social value, and that it is not possible to punitively filter out irresponsible parents through a traditional licensing regime, then its not clear why we should retain a voluntary licensing regime. The most irresponsible prospective parents would have very little reason to engage with such a regime. Surely there are better ways of supplying the necessary education. For example, why not simply include parenting classes and tests as part of the mandatory school curriculum? That way everybody would still require some basic knowledge and skills of parenting and you’d avoid all the negatives associated with a licensing regime. This would come with its own set of problems, but they would be no more significant than the problems associated with all other forms of standardised education, and they might do some (comparative) good.

This sounds about right to me. Indeed, I am more persuaded by this critique than the former one. I think a restrictive licensing regime would be morally and practically unworkable (not unless we are willing to tolerate costly invasions into our private sexual lives). Given these practical problems it seems likely that there are other, less costly, means to a similar end.


4. Conclusion
In sum, the argument for parental licensing is provocative, controversial and initially quite plausible due to the seductive analogies with similar licensing regimes. In short, it has everything that you might want from an applied ethical argument. It is only when you delve a little deeper and consider the distinctive nature of the parental good and the costs of practical implementation that it starts to unravel.

But there may be broader implications here. At the very end of their article, de Wispelaere and Weinstock point out that this critique of the licensing proposal might have unforeseen implications for traditional adoption regimes. After all, adoption regimes already tend to involve the de facto licensing of adoptive parents. But if licensing is unwelcome in the case of natural parents why is it any more welcome in the case of adoptive parents? There may be answers to that question, but I won’t pursue them now. Something to take up in the comments section, perhaps?

Three Practical Hurdles to a Universal Basic Income

The campaign for the introduction of a universal basic income (UBI) has been gaining ground in recent years. What was once a slightly obscure proposal, beloved by certain political theorists and welfare reformists, is now being embraced as a potential solution to the threat of technological unemployment. I myself have written about it on several occasions, mainly focusing on different political and philosophical arguments in favour of its introduction. These arguments focused on the normative/political grounding for the UBI. They rarely, if ever, focused on the practicalities of the UBI. How would it be introduced? Would this be an easy thing?

In this post, I want to take up some of those practical questions. In particular, I want to consider a pragmatic argument in favour of the UBI, one that is often trotted out by its supporters. Then, I want to consider some potential pragmatic hurdles to the introduction of a UBI. In doing so, I draw upon an article by Jurgen De Wispelaere and Lindsay Stirton entitled ‘Practical Bottlenecks in the Implementation of a UBI’. In that article they look at three specific hurdles to the introduction of a UBI, ones that supporters of the policy should consider.


1. The Under-consumption of Welfare and the Simplicity of the UBI
Most welfare entitlements are selective. That is to say, only certain people qualify for those entitlements. For example, jobseekers allowance is a selective entitlement: you are only entitled to it if you are an adult who is actively seeking work. Similarly, childcare benefits are selective: they are only available to those who have dependent children. Sometimes, there are multiple dimensions to the selectivity. For instance, childcare benefits might be means-tested, meaning that only those with children and with incomes below a certain level, are entitled to the benefit. These selective policies can be justified on various grounds, but their selectivity can give rise to certain practical problems. Chief among these practical problems is the problem of under-consumption.

This problem is widely discussed in the literature on welfare entitlements. It arises whenever people who are genuinely entitled to a welfare payment fail to receive it. This is a significant failure of policy: the rationale behind selective welfare programmes is that they provide benefits to people who are genuinely in need of them (who genuinely deserve them). If they fail do this, then there is a significant “justice gap” associated with their implementation.

Under-consumption may be attributable to any number of causes, but it is generally believed to increase whenever the programme includes complex sets of rules and guidelines, whenever the entitlement criteria are vague, whenever the programme requires the applicant to pass a means test, or whenever the criteria of entitlement are associated with some negative prejudice (e.g. the “unemployed”). Selective entitlement programmes tend to exhibit these properties to greater or lesser degrees, which means they often suffer from the problem of under-consumption. Furthermore, there is a limit to how fast and effective bureaucracies can be at tracking people as their life circumstances change (e.g. their employment status, marital status, parental status). Since these life circumstances often determine eligibility for selective entitlements, it follows that selective programmes will always have to contend with a significant risk of under-consumption.

There are other practical problems associated with selective programmes. As you might imagine, the bureaucratic costs associated with policing the boundaries of entitlement are often significant, and sometimes entail humiliating and morally costly intrusions into the lives of those trying to claim the entitlements. It would be better if we could minimise those costs.

This is where the pragmatic benefits of the UBI are typically touted. Irrespective of its moral, political or economic credentials (which are nothing to be sniffed at) the UBI seems attractive because of its administrative simplicity. If it is truly universal, then there is no need to police the boundaries of entitlement. There is no need for costly bureaucratic structures to assess those who make claims, no need for complex rules and guidelines, or prejudicial criteria of entitlement that might dissuade people from making a claim. The problems of bureaucratic cost and under-consumption can be addressed in one fell swoop.

If only things were that simple. Although the UBI certainly has some practical advantages over its more selective peers, it also comes with its own practical hurdles to implementation. As de Wispelaere and Stirton point out, the sheer size and scale of a UBI would be unprecedented. Setting up the infrastructure for its administration would be no small feat. Indeed, there are at least three hurdles that would need to be cleared. Let’s examine them all.


2. The First Hurdle: Maintaining a ‘Cadaster
The first practical hurdle for the UBI is:

Hurdle (1)- Identifying all eligible claimants: If the UBI is to be made available to everybody (or at least every adult or citizen) in a given society, then it is essential that we have some way of identifying all these people. In other words, it is essential that we have some universal register or ‘cadaster’ for eligible claimants.

Many people don’t perceive this as a problem. This is because they think the main problem with selective entitlement programmes is that they struggle to identify the ineligible, i.e. to exclude those who don’t meet the criteria for the programme. Consequently they tend to think that the UBI, because it removes most criteria of ineligibility, solves the problem of identifying eligible participants. But this is incorrect. Just because we no longer have the same number of criteria of exclusion, does not mean that we will magically succeed in including everyone entitled to the UBI. Identifying every person (or adult or citizen) in a given state is itself a significant administrative task.

So how could we do it? De Wispelaere and Stirton look at two possibilities. First, we could try to construct a register from scratch. In other words, introduce some new scheme for tagging and identifying every person (adult or citizen) in an administrative area. This is likely to be a costly endeavour. When the UK proposed to introduce a ID card scheme in the 2000s, the original government estimates were that this would cost approximately £5.6 billion. Later independent estimates suggested that the true cost could be anywhere between £10 and £20 billion (these figures are cited in De Wispelaere and Stirton’s paper). Furthermore, there may be great cultural differences in the willingness of a populace to put their names on such a register. People may not trust the government, and people may worry about how their personal information would be used (beyond simply determining eligibility for the UBI).

The other possibility is that we make use of pre-existing registers. Obvious examples would include registers of social security numbers or voting registers. The problem with these is that they are often far less universal than people suppose. Social security registers tend to work pretty well for people whose births are registered in a particular state but not for those who move to a state later in life. They usually need to apply for a number, and some are either unaware of the need to do so, or simply cannot be bothered to do so. The same holds true for voting registers. I can speak from personal experience on this front since there have been several times in my adult life that I failed to register to vote and so my name was missing from the register. Use of overlapping, but pre-existing, registers might be a possibility but this too would come with costs and would be likely to contain gaps.

I would make one point about this practical hurdle. While I agree with de Wispelaere and Stirton that we shouldn’t ignore the practical hurdles associated with the UBI, I also think it is important not to overstate them. As with every policy, the assessment of the UBI should be conducted comparatively, i.e. by comparing it to its rivals. No welfare policy will completely avoid the problem of under-consumption, or be free from all administrative costs. What matters is whether it is preferable to its rivals and whether these problems can be minimised. The importance of this observation holds in relation to the other two hurdles as well.


3. The Second Hurdle: Ensuring Robust Payment Modalities
The second practical hurdle is an obvious one:

Hurdle (2)- Ensuring that everyone gets paid: It is key to the ethos behind the UBI that everyone receives their income grant, but how can you ensure that this is the case? In other words, how can you ensure that the payment modalities are sufficiently robust to include all eligible claimants?

As I say, this is a pretty obvious practical hurdle to implementation. There are many different payment modalities in modern societies. Not everyone is able or willing to access to all of them. In their analysis, de Wispelaere and Stirton consider a few options.

The first is to use the existing taxation system. This would involve disbursing the basic income grant as a tax deduction (or “negative income tax”). The advantage here is that you could piggyback on top of pre-existing administrative structures. The disadvantages are multifold. First, not everyone participates in the formal economy, or is capable of or willing to submit a tax return. Furthermore, this system is often disliked by feminists on the grounds that taxation systems are often set-up for households as opposed to individuals, and men have tended to control the household. This might impact on the alleged benefits of a UBI for women. Finally, and perhaps most importantly, there is the fact that taxation systems often operate on a “once a year” basis, meaning that the income grant may only be supplied in one large annual lump sum. It is arguable that a more regular system of payments would be more beneficial to vulnerable members of society.

But if there is to be a more regular system of payments, how would it be disbursed? Direct payments to bank accounts might be problematic on the grounds that some people may not have bank accounts, and direct cash payments would have significant administrative costs (i.e. the need to have large, well-staffed payment centres where people could go to pick up their payments on a weekly or monthly basis). A more innovative solution would be to introduce a system of basic income debit cards for all eligible persons. These could be used to make payments or to take out cash from approved ATMs. But problems arise when it comes to: (a) ensuring that people actually get these cards; and (b) setting up and maintaining the IT systems that would be used to administer this payment system.

Again, these problems could be overcome (with some inevitable errors or gaps), but the costs should not be underestimated.


4. The Third Hurdle: Maintaining an Effective Oversight Mechanism
The third hurdle arises even if we can clear the other two:

Hurdle (3)- Maintaining effective oversight: Once a UBI is implemented it will be still be important to ensure that everyone is receiving their income grant and that no one is being left out.

This is again about ensuring that we avoid under-consumption. You might wonder why the focus is solely on under-consumption. Surely, you might say, there is also a problem associated with potential overconsumption (or “double-dipping”)? We don’t want people to get more than their fair share of the income grant, do we? I would tend to agree with this but de Wispelaere and Stirton deliberately avoid this issue in their analysis. Nevertheless, it seems fair to say that an oversight mechanism would also be needed to avoid the problem of over-consumption.

So what kinds of oversight mechanism could we have? In the public administration literature, two general types of oversight mechanism are discussed. The first is a “police patrol” mechanism. This requires some centralised administration, specifically a group of enforcers, to investigate and ensure proper enforcement and compliance with the terms of the policy. This is typical of traditional selective entitlement programmes involving means-testing. Since the absence of such a centralised system of enforcement is usually thought to be an advantage of the UBI proposal, it is unlikely that UBI advocates would be keen to create such a system. Furthermore, given the size and scale of the UBI, it is pretty clear that any such system would represent a massive and unprecedented intrusion into our private lives. Maybe we are willing to tolerate that; but recent scandals about government data collection suggest otherwise.

That leaves us with the second type of oversight mechanism. This is the “fire alarm” mechanism. This doesn’t require a centralised group of enforcers. Instead, it relies on individuals self-reporting the problems they are having with the system (e.g. failures to receive entitlements). The complaints from individuals would then be processed by some agency, which could help to ensure optimal levels of consumption.

There are at least two problems with the fire alarm mechanism. The first is that it is unlikely to work in identifying possible incidents of over-consumption. People who are receiving more than they are entitled to will have few incentives to report this problem (though, of course, some honest souls might do so). Resolving this by getting people to “spy” on one another seems like it could lead to some unpleasantries. The second problem is that even though it might work better when identifying incidents of under-consumption, it is likely to face some practical costs. There will need to be some bureaucratic agency for processing complaints, and people will need to feel comfortable trusting this agency with their problems.


5. Conclusion
In sum, although the UBI may represent a gain in terms of social justice and economic efficiency, it will not be without its practical costs. Three costs, in particular, were discussed in this post. The first is the cost associated with identifying all the eligible claimants; the second is the cost associated with setting up some reliable means for disbursing the income grants; and the third is the cost associated with maintaining some ongoing oversight mechanism. Although every policy is likely to have some costs and some inevitable flaws in consumption, it is important that UBI advocates don’t ignore the costs that are unique to it.

Interview on Robot Overlordz Episode 150: Tech Unemployment and Enhancement

I have the good fortune of being a guest on the Robot Overlordz podcast this week. Some people might be interested in checking it out. The link is here.

Robot Overlordz is a biweekly podcast, hosted by Mike Johnston and Matt Bolton. As they say themselves:

Robot Overlordz is a podcast about the future. On the show, we take a look at how society is changing. Everything from pop culture reviews to political commentary, technology trends to social norms. All in about 30 minutes or less, every Tuesday and Thursday

I recommend checking out their website and listening in regularly to their show. In my episode, I discuss three things with Mike and Matt:

• Tech Unemployment: Are human workers going to be replaced by robots? Is this happening now? Are fears about tech unemployment misplaced? And doesn't this simply point to a tension or contradiction inherent in capitalism?

• Neuroenhancement and the Extended Mind: Could smartphones and wearable tech be a part of our mind? Is interfering with someone's iPhone ethically equivalent to interfering with someone's mind?

• Enhancement in Sports and Education: Should we allow the use of enhancement technologies in sport or in education? Are traditional methods of educational assessment fit for purpose?