Thank you to all of those who joined the webinar "Harmonization of Contributory and Non-Contributory Programmes"! Here is a summary of the questions raised during the webinar and some insights from the webinar's panelist, Dr. Stephen Kidd.
Q: Question for Dr. Kidd: When you argued that the poor are composed by population coming from a variety of consumption groups (e.g. below 2$ a day, 4$ a day etc), how exactly are you defining poverty?
Answer: All poverty lines are relatively arbitrary and, if we were to use a sensible definition of poverty, we’d be using poverty lines of $5 or $10 per day: anyone living under $5 per day is certainly in need of additional social protection. Furthermore, as I argued, people’s incomes move up and down frequently. So, when you use panel data sets, you can often see that the majority of the population in developing countries spend some time under the poverty line over a relatively short period (see this paper on Georgia for an example). So, my basic argument is that, in developing countries, most people are living in poverty – or are highly insecure – and would benefit from accessing social protection. Choosing a low poverty line gives an unrealistically optimistic picture of real levels of poverty. See also this paper challenging the concept of “the poor” and this blog on why “the poor” is a pejorative term.
Q: How do you identify (and thereby target the poorest) if the majority are in the informal sector and you don't use proxy means test?
Answer: Perhaps the question is, why target the poorest if almost everyone is living in poverty or insecurity and is in need of social protection, and access to social security for all is a basic human right?. And, in particular, when it is impossible to identify the poorest, who are, in effect, a fictional construct, with membership changing continually. Countries can afford to provide schemes to the majority of children, older people and persons with disabilities, if they are really committed to tackling poverty. As I mentioned, 3% of GDP invested in social protection would deliver an effective system that really tackles poverty and inequality, builds social cohesion and generates economic growth. Arguments about targeting are really arguments about how much to invest in social security. Countries target “the poor” when they are not committed to investing in social security (see this paper for further discussion).
The proxy means test is a weak system of targeting, generating errors of over 60% when targeted at the “poorest” and selecting people in a highly arbitrary manner. Read this paper for more information. Also read this blog for a comparison of universal and poverty targeting.
Q: How social insurance systems could be implemented in the informal sector bearing in mind the disincentives it may cause in terms of decreasing the net income of the workers?
Answer: social insurance is meant to be mandatory, if it is going to function. But, informality means that governments cannot enforce payroll taxation. So, the informal workforce need to be encouraged to participate in social insurance schemes, perhaps by allowing contributions to be made on a less rigid basis or providing access to other benefits, such as loans. Subsidising access to informal economy workers – such as by government providing matching contributions – is a means of privileging those that are better-off and governments would be better off investing in tax-financed schemes, which benefit everyone.
However, the decrease in income is really just people saving or investing in insurance: they can undertake consumption smoothing and access benefits when they are in need. So, it is to be expected that they pay for it. The disincentive comes when the payroll taxes are too high, because schemes have been mismanaged and/or poorly designed.
Q: Why not advocate for a hybrid functional system? Is it really feasible to have a universal social security system without actively involving the employed who contribute to social security?
Answer: My point is that we should look to have both, but we can’t have a universal system unless the basic platform of the system is provided through tax-financed schemes. The employed contribute taxes and, therefore, also finance tax-financed schemes, from which they can also benefit. As I said, both non-contributory and contributory schemes should be regarded as contributory. We pay the taxes for non-contributory schemes and contribute to the country in many other ways, so we should also benefit from these same schemes that we pay for.
Q: Is the classification of schemes based on the dichotomy Social Insurance/Tax financed schemes still relevant when most developing countries have adopted a hybrid approach (with contributory component and subsidies, partial or total)?
Answer: It is a simplistic dichotomy when the reality is more complex. As I said, many contributory schemes are, in fact, often financed from general taxation, especially so-called health insurance schemes. Depending on the level of formality in the labour force, the balance between financing from general taxation and from social insurance contributions will vary. But, tax-financed schemes are an essential platform for the broader social security system.
Q: State tax revenue can be a mix of VAT, personal income tax and corporate tax. Any comments on the ideal mix and enforcement mechanisms for low and middle income countries? Thanks.
Answer: You should probably ask others about this. Generally, though, more progressive tax systems will depend more on income taxes and corporation taxes, and less on VAT (unless VAT is on luxury products)
Q: Universal coverage schemes require substantial government funding, and increasing the tax base. What strategies would you recommend to extend the tax base, given the size of informal economy, and without negatively impacting formal workers purchasing power and overloading small and medium enterprises?
Answer: Actually, I don’t regard 1% of GDP or 2% of GDP as substantial. The average investment by developed countries in social security is 14% of GDP and some successful economies invest up to 20%. This brings rewards in terms of lower inequality, greater social cohesion and stronger economies. But, extending taxation is about building the social contract. And, this can only be done by offering entitlements to people and not by targeting “the poor” with methods such as the proxy means test which are highly inaccurate and arbitrary and cause significant opposition in communities. When people are convinced that government can deliver quality services – such as a guaranteed old age pension – they will slowly begin to be willing to accept higher taxes. But, of course, tax evasion needs to be addressed and punished. Indeed, it would be a good start if those working on social protection and advocating higher taxes were to pay taxes themselves (many consultants and development partner staff do not pay tax, but want to access the benefits).
Q: What role do you think corruption plays in limiting the effectiveness of these schemes? Do you think that private sector management of these funds and the involvement of stock/bond markets is optimal?
Answer: Corruption is more of a problem in poverty targeted social assistance schemes, and much less of an issue in entitlements schemes, which are much easier to monitor and manage. It’s also a problem in contributory schemes, including those managed by the private sector (with high charges another significant issue). All contributory schemes need strong and effective regulation by government and independent parastatal oversight bodies, with teeth to enforce penalties.
Q: The problem with talking about tax is that in many developing countries the tax system itself is in shambles with leakages, lack of implementation, corruptions, etc. so What do we do? Do we wait until they fix their tax system or find alternative ways to provide social protection, i.e. work politically????
Answer: Improvement will only come from ambitious and progressive politicians who can offer improved services and have the confidence to increase taxes. Again, it’s about building the social contract, slowly and over many years. But, the key is to offer entitlements and not arbitrary and low quality poverty targeted schemes with massive errors.
Q: Apart from gender, other factors such as caste (in South Asia), race, gender non-conformity and landlessness can contribute to insecurity, esp. within the informal sector. How do we account for these while designing social protection programmes? Sudheesh, Oxford Department of International Development, University of Oxford.
Answer: Many factors can contribute to insecurity and make it more difficult to access social security schemes, in particular social exclusion (see this paper for further discussion). However simple, entitlement based programmes make it much easier for those experiencing social exclusion to access schemes. Poverty targeting creates many more opportunities for exclusion.
Q: Great presentations. And great graphs from Stephen Kidd. Stephen, about the "gender bias" of contributory schemes (you showed a graph). Wondering if you'd have similar data of coverage by age (comparing youth -15-24- to adults)
Answer: Sorry, no. Best ask the ILO.
Q: Why are many countries across the globe not adopting the ILO-WHO driven social protection floors which are more stable compared to safety nets which have holes implying that some beneficiaries are likely to fall through at some point?
Answer: Many countries are moving towards this, and are gradually building more entitlement schemes. But, a neoliberal ideology is still strongly held by many of those working in social protection, with the underlying belief in low taxes, small schemes and poverty targeting. Many of these are in international institutions and many of the arguments they make are not evidence based. Look, for example, at the promotion of Bolsa Familia as an example of good practice when, in fact, it is a tiny , poverty targeted scheme in Brazil – which excludes around half of its target group – and has minimal impacts when compared to the Previdencia Social scheme (see this paper for more information). See also this paper on the myths perpetuated on social protection, to promote poor quality poverty targeted schemes. And, also see this paper on the type of poor quality analysis undertaken to promote poverty targeted schemes.
Q: I agree very much with the importance of Dan's point about the complimentary roles of SP and family policy. The trouble is that in too many countries the latter is weak or non-existent, and hence there is heavy demand for SP to fill the gap - is or is not the result distortion of objectives and outcomes? JW
Answer: in reality, in most developing countries, social protection is severely underfunded, as are broader family policies, social work, social care etc. We need to encourage much higher and more balanced investment across all areas, which means higher taxes and the rich pulling their weight and realising that they are citizens, not elites.
Q: For me it seems the contributory scheme usually has a different purpose than the non-contributory programmes. I mean, usually non-contributory programmes want to address hunger and contributory wants build social welfare. Then, they have different impacts in the Gross National Product. How do you measure their different impacts?
Answer: I was trying to argue that they are complementary, but with different and overlapping objectives. One key measure is coverage: are all those that are at risk and in need covered by either tax-financed or contributory schemes. And, secondly, are the value of the transfers high enough?