MDGs, taxes and where tax revenues come from...

As part of the build up to its 50^th Anniversary, the OECD Development Centre has released a new report giving an updated estimate of the cost of meeting the MDGs. Entitled "Revisiting MDG Cost Estimates from a Domestic Resource Mobilisation Perspective" and financed by the Gates foundation, the report plays a valuable role in re-focusing minds towards the costs involved in meeting 6 of the MDGs by 2015, the target set back in 2000.

The paper also plays an important role by bringing a domestic resource mobilisation perspective. In these times of fiscal austerity and increasing attention on how developing countries can increase their own resources through taxation and savings, the report builds on the OECD’s own work on taxation in the 2010 Africa Economic Outlook to gauge the degree to which domestic resources might be used to fill the “financing gap”. By bringing the topics of MDGs and domestic resource mobilisation together,  the paper serves as a reminder that underlying many of the problems faced in developing countries is the need for strenger states and improved tax systems.

However, in bringing the two topics together, and despite the caveats given, it may oversimplify the  complexity of raising additional resources through taxation.    

A major element of the analysis of the scope for countries to cover MDG expenditures  domestic resources is based on an analysis of "tax effort". This measure comes from estimating the "average" tax performance across a group of countries, taking into account some of their characteristics such as income levels, the share of agriculture in the economy, and the level of trade. Countries with tax performance below what average performance would predict for them are considered to be capable of exerting more tax effort. Estimating tax effort therefore assumes that those are the only determinants that determine tax performance - that a country is gathering less or more than the average, based on the chosen country characteristics.

But tax performance also depends on a lot of other aspects. So can we really assume that the fact that some countries gather more revenue, even taking into account their income levels, means that other countries should be able to do so? For one thing, as the above brief explanation shows, the tax capacity predicted by the analysis will depend very much on the countries included in the analysis. Income per head may be correlated with a lot of characteristics that might affect tax performance, but other institutional and structural characteristics are also likely to be important so including EU countries, for example, may distort the prediction of what revenue levels are "feasible". Being resource rich also seems like a characteristic worth taking explicit account of in the analysis in this context, for example. Indeed, the 2010 African Economic Outlook report itself highlighted that a major part of the increase in domestic revenues across Africa has occurred in resource-rich countries. 

In addition to this, averages can be easily skewed by outliers. Countries such as Lesotho or Swaziland appear as having high tax effort, but in reality this is because (at least until recently) they receive a substantial part of their revenue through transfers from the SACU regional revenue sharing pool, largely financed by South Africa. So although they may be better placed to finance MDG-related expenditures, this is not necessarily related to tax effort or their taxation system. 

Although the estimation of the financing gap is explicitly recognised as a simplification, increased revenues are likely to be increased through increased economic growth. This is also overlooked by the study to some degree. To estimate the financing gap, the study estimates the additional resources required to achieve a growth rate that would  reduce poverty by 50 percent. This is then taken as the financial gap that must be filled to achieve this growth rate. But achieving the growth rate itself would imply economic activity, that itself generates tax revenues and potentially lowering the gap to be filled.

This highlights a borader point. The relation between taxation and the financing of the MDGs really needs to take account of where tax revenues actually come from and how they are collected. 

A gathering body of research, some of which was summarised in our recent Discussion Paper, highlights the institutional implications of tax policy and tax policy reform and their fundamental importance in determining how successful a country is in raising revenues. How the revenues are collected can then impact on growth through its influence on private sector behaviour. These are by no means simple issues to resolve. 

Further, on the specific question of how the MDG "gap" could be financed, increasing domestic resource mobilisation and increasing ODA are discussed as two substitutes. But as the (more critical) literature on the impacts of aid suggest, the institutional demands and implications of increased aid are very different to those of increased DRM. The timescales are clearly different too.

What this all highlights is the vital importance of paying more attention to the detail of how resources are and therefore can be raised. But of course, the analysis completely ignores the potential increased revenues form stemming capital flight. Only recently, Global Financial Integrity provided an updated estimate of capital flight in 2009, giving a figure of $600bn in 2009. Put in those terms, and given the estimated need for an additional $120m to meet the MDGs, maybe domestic resource mobilisation could do it after all....  

Nonetheless, highlighting the need for more resources to finance the MDGs is a good way to rally support and increase efforts in understanding the complex issue of raising domestic resources in developing countries.   

An edited version of this was posted at ECDPM's Talking Points here

"A loan? What are your assumptions, sir?"

This may be a question answered in Economics 101, but what are the assumptions behind offering loans to developing country governments for public investment projects?

The question arose for me the other day while attending a roundtable discussion of the EIB and how they intend to measure and monitor the impact of their investment projects. There must be some underlying assumptions since they expect repayment and they calculate an economic rate of return! And how come I've never wondered about this before?

All I could think of, is that public investment is expected to raise economic growth which is therefore expected to translate into higher tax revenues, some of which will be used for loan repayments. Is that really it though? Of course there will hopefully be additional other positive externalities which one should take into account to reflect the "true" value of the investment. But at base, if you assume the loan will be repaid, you also have to assume some mechanism for that to happen, which means you have to have some idea of direct impacts. But I'm pretty sure there's no ear-marking of revenues from these kinds of projects for loan repayments in budgets, especially since we can't measure the impact unless we really make an effort. That then raises the question whether or not  the ex-ante loan calculation is actually related to the specific investment, or just to a general expectation of ability to repay based on other factors?

This then relates back to the discussion being held, which was about how the EIB intends to measure the impact of its investments, as required under the renewed mandate for ex-EU financing given by the European Parliament and Council. What is expected from this measurement of impact? How will this differ from whatever is already used in the economic rate of return calculation? Do they really want to understand the impact or just to have some figures to put in the annual report?

Because the reason for the indicators and monitoring is likely to determine the nature of the indicators. If you really, genuinely want to measure impact in order to learn lessons for future development-related investments, then presumably the process needs to be pretty rigorous. To me that would imply pre-investment surveys of households and firms, with periodic follow-ups, something which could feasibly be carried out by local agencies in conjunction with local statistics offices - it could even use existing surveys taking place in a country or region. Then you could say something about the real impact of the investment project, and better understand the mechanisms at work. If not, you risk producing a CSR-type report of figures which are pretty much meaningless. With a more rigorous approach you could also get a potentially reasonable estimate of the impact on tax revenues and get a better handle on expected returns to investments.

But again, I may be missing something. Can it really be just through tax receipts that we expect a financial return? The discussion of taxation in developing countries highlights the huge amount of evasion, the weak administration or even the weak incentives to raise taxes, so it makes me wonder how reliable an assumption that would be. Maybe I just need to go and ask them. As a colleague of mine likes to ask, "what are your assumptions"? It's a good question I think.

As for the EIB's indicators, until there are some actual indicators to discuss rather than broad conceptual frameworks, it's hard to say how useful they will be.

Sutton and I on firm-size (he's probably right...)

I've been on the LSE podcasts once again, this time listening to John Sutton talking about industrial capabilities in an International Growth Centre Talk. It's interesting stuff but that is perhaps also a slight confirmation bias story....

For one of my PhD thesis papers, I tried to uncover a firm growth threshold effect from some Mozambican firm-level data. The idea was that small firms grow to a certain small size, below the government tax radar, or they start off considerably larger where they can handle being on the radar and dealing in some way with taxes, regulations, bureaucracy etc. It's related to the missing middle story, where the huge micro sector manages to escape, while the small "large" sector manages to get either incentives and benefits, or has political/government contacts, and in any case has enough scale to make certain compliance costs appear relatively very small.  Anyway, partly due to a tiny sample size, and partly perhaps because my underlying theory wasn't quite full-proof, I had a bit too many firms jumping the threshold to really make it a genuine threshold (at least, that's what the anonymous referee said when I submitted to a journal). 

Anyway, Sutton says the same story in terms of size, but has a different explanation - management capabilites. What he says also confirms what I saw in interviews in Mozambique so I have to stand aside, take my hat off and all the rest. As he points out in his talk, the lack of growth from micro to any reasonable size is indicative of a lack of ability to manage a medium large firms (particularly in the kind of conditions which I refer to above - so actually, we're not really saying different things in a way). But what is more, medium firms tend to emerge from existing medium firms, even if this is in other sectors. So, citing cases from Ethiopia and Tanzania, he talks about how trading firms gained insights into markets, prices, supply-chains etc etc before then entering totally different fields, including manufacturing. 

He also mentions the issue of large multi-product firms which can diversify risk through various products, thus offsetting losses in some areas while learning takes place in terms of how to improve standards or productivity in other areas. So the big multi-sector firms one sometimes comes across may not be so bad - an argument also made by Ha-Joon Chang in yet another LSE lecture, of course talking of the chaebol...

This is all well and good, and quite inspiring even if the idea is pretty simple. In fact, when I was discussing firm thresholds someone pointed out to me that even in a developed economy, it's very different to manage 3, 4 or 5 people, than 15. So perhaps in any country there is a threshold when you go from being a "successful entrepreneur" to a "successful company", and where the skills are very different. And while this is likely to remain true for a developing country, where markets are dispersed, consumption of anything beyond basic goods are low, and the bureaucratic burdens are high, the skills to be a successful, medium-sized, formal firms are likely to be very different from being small, or running the local subsidiary of a multi-national. Does that mean we need more management consultants in development? (yes!)

So, I like the idea.

BUT. That said, Sutton moves on to discuss "what needs to be done" and this leads him to suggest the need to look for "doable" policy issues to remedy to allow some of these firms that have the capabilities to thrive. A little along the lines of Rodrik's search for the "binding contraint",  he suggests that the diagnostics of what is required for a good business environment are relatively easy, but doing something about them is not. So, choose some which are "doable" and then "just do them".

This is a little unsatisfactory. Rather than a "just do it", this is where I was waiting for Prof. Sutton to say, "ask the question, why has it not been done already". He talks about land titling as an easy win, for example. But surely the question is, "who gains from the current land titling system?". Why has nobody sorted it out before? since many of these things have been discussed for a long time. It may indeed be, as he says, lack of capacity, but when someone mentions that along with "lack of political will", there is a need to ask the question where the political will does lie, and why. Maybe then you can actually work out what is doable.

In the meantime, the firms which are willing to take the risk will be learning capabilities to get around the current system instead of improving their business itself. So I think it's a great point about firm capabilities. It made me think I should resurrect that thresholds model some day... 

Testing de-Soto's hidden capital...

Now here is a paper which I think is potentially brilliant - finally someone is testing Hernando de Soto's hypothesis that if only people in developing countries could used the capital stored in their (even meagre) housing, they could use it as collateral to get credit and get themselves out of poverty. The lesson for poor governments was (if I remember correctly), sort out your property rights and the rest will follow. So what does the evidence say? Well,

the effects [on profits and interest rates] are likely to be non-linear and heterogeneous by wealth group. They also depend on the extent of competition between lenders. There can be significant increases in profits and reductions in interest rates when credit markets are competitive.

Hmmm. As far as I understand it then,  there is large variation, presumably relating to the type of house being used as collateral, and actually it all depends on the credit market anyway. As developing country credit markets are renowned to be tight for pretty much all but large companies (a fascinating paper here on that - actually more self-promotion but anyway) and government bonds,  then to some extent the granting of property rights to use as collateral is good, but perhaps isn't actually the "binding constraint". And anyway, the record for micro-credit is also pretty ambiguous I believe. Does this mean the de Soto effect was a bit of a red herring? Maybe I'll have to actually read the paper....

Moving tax bases...

When I wrote a blog post on the EC's proposed financial transactions tax a month or so ago, I tried to be relatively balanced about it. I may have failed though, as I couldn't hide a certain sense of scepticism -  whatever tax you invent, certain people, companies, and especially large companies will manage to get around it.

So the other day when I heard someone talking about how a bunch of historical taxes had been dreamt up, served a purpose, but then been "retired", it reminded me think that tax bases change through time due to adaptations of people's behaviour. Therefore, perhaps taxes administrators can't really continuously target the same tax-base but have to continuously change to keep up.

A few historical tax-base shifters that had (predictably?) interesting behavioural effects from the UK include: hearth/chimney tax (1662-1689), window tax (1697-1851), dice duty (1711–1862), wallpaper tax (1712–1836), brick tax (1784–1850), hat tax (1784-1811), glove tax (1785–94), hair-powder tax (1786–1869) and perfume tax (1786–1800) (thanks Wikipedia, which always tells the truth).


What's interesting is that these were all designed with a view to taxing i) something which was pretty visible; ii) something associated with wealth and which was therefore progressive; and iii) something which would inevitably raise revenues - if nothing changed.

But all of them were also open to declining revenues from changes in behaviour through time. So the chimney tax didn't last long, apparently due to the intrusiveness of the inspections to see how many fire-places people had, while people of course began blocking up chimneys.  Over time, the window tax led to people boarding up their windows and living in darkness, leading to health problems (but it nonetheless lasted 150 years!). And apparently wallpaper tax led people to buy plain paper and then stencil on designs. The brick tax led to an increase in size of the standard building brick, while presumably the hat and glove tax led people either to buy less hats and gloves, or to a black market in those goods.

We still tax certain goods through duties and excises, a little like the hat and glove taxes then, but still, the point remains that the tax base on a good deal of items is constantly changing and I wonder how much that is really taken into account in the short-term tax policy design that goes on. Particularly in developing countries where the capacity for tax enforcement is relatively weak.  

On another note, at a workshop on taxation I attended last week in Bonn, one study presented was examining the increase in revenue collection performance when autonomous revenue authorities start up - the anecdotal evidence is that there is generally a surge in revenues, followed by a roll-back to more or less previous performance. Could this be because people have adapted to the new enthusiasm of the tax collectors and new systems so spend a little more effort hiding their money? Perhaps we simply have to accept it - people will always wear less hats and board up their windows, so if we want a steady tax revenue base, we need to keep inventing....

National accounts and tax is the real frontier!

National accounts? Jesus, can it get more boring?

Well, actually, after yet another podcast today from the LSE events series (this is how I stay connected with the metropolis while driving to Maasrticht!), the point was made by Tony Atkinson that very often economists are happy to use and work with data on GDP etc, but very often have no idea about where it comes from. And I agree. And not only from working in Mozambique....

In my first real job in the Scottish Civil Service, I worked on producing the Scottish Input-Output tables, basically the tables behind a social accounting matrix, which are used to reconcile national accounts data. Since the border with England is pretty porous, as you can imagine data on exports and imports from the rest of the UK are pretty ropey. And what about the rest of the data? Well, there were periodic surveys carried out of the expenditure and sales patterns of specific sectors in the economy, but obviously there were too many sectors to do a survey every year, so some sectoral data was just scaled up versions of data collected a while ago. A long while ago in some cases. Scaled up based on what you ask? Based on the sectoral growth rate data, which came from the production surveys, also a sample-based dataset from which GDP growth rates were estimated, themselves based on some pretty subjective weights.

Just from that brief description, you can probably see some room for errors. Low response rates on surveys might cause a few problems. Major changes in inputs and/or sales may make a few changes too which can only be picked up through a survey. But even the overall growth rate of the sectors is based on sample estimates, again subject to selection bias through non-responses. And yet, this is about as good as it gets. While the various data and measures of expenditure, production and incomes were then made to reconcile through "RASing", I remember someone telling me that in Germany (perhaps pre-RAS) they had balanced their tables through a kind of auction in the cafeteria, with those with excess numbers offering to offset with those with deficits in their column. Arbitrary? A little bit. Yes. 

So imagine what goes on in a developing country, where the survey design, response rate, data entry, analysis, and balancing are all even more open to error. I recall one IMF mission discovering a mistake in an excel file which meant that in fact GDP growth forecasts (themselves ropey) were being over-estimated due to the transport sector, which in fact was expected to grow at 7 percent not the 17 percent the excel error had led us to believe!


This is something dealt with in a recent blogpost by Shanta Devarajan from the World Bank. Although his focus is on the implications for poverty estimates, his point is not limited to that.  

This also has clear implications for tax and tax policy. If your revenue target is estimated as an (increasing) share of GDP, if you increase your growth forecast, you also increase the actual amount of revenue which the tax authorities must collect. What will they do under that pressure? My guess is they might become a bit arbitrary and over-enthusiastic with their revenue collection, and become a little more coercive perhaps...   

Or maybe not. One to investigate....

   

Collier on some of these political economy issues

I was listening to one of the Growth Week podcasts from the LSE site the other day. It was a discussion of state-building with Paul Collier, Tim Besley and a politician from Bihar. Since I found it interesting but at the same time had the feeling there was something missing, here is a quick summary....

Besley gave quite a simple exposition of state-building, based on the Weberian/historical interpretation of Western Europe, as states requiring taxes to fund wars (and now, a common cause of effective public spending), and then putting in place the legal structures required to go with it. According to his view, then, governments should be "investing in tax systems" and the expenditure system, and "investing in growth" through provision of a legal framework and security, purely on the basis of extracting more through taxation. So once you get a certain level of tax revenues, you improve the rest of your institutional structure, including property rights etc, and off you go. If they don't, this is because the pay-off from taxation is not enough, due to a lack of "institutional cohesiveness" and the lack of longevity of political actors (see here for the full underlying papers).

While that is an oversimplification, the point made by Collier was that sometimes those in power are not interested in state-building, so the investment-payback or the time-inconsistency issue of taxing now for longer-term payback are of less interest. There is plenty of evidence from plenty of places that this may be the case. This could be due to the escape route provided by natural resource rents, by aid, or simply be a minimal tax-base which can be reallocated according to the whim of the political class, meaning that according to the model they have no incentive to promote growth to raise revenues. So what does that mean for "state-building"?

According to Collier, the political instability which might lead politicians not to invest in tax etc, often comes from high unemployment among young men. This then implies a need to look at jobs.  And what are the jobs which might be appropriate in a poor country with a high-level of low-educated, urban unemployed? Construction! And construction of low-cost, basic housing in particular.

As an aside, this actually relates to a proposal I heard about quite some time ago from a Canadian economist named Lauchlin Currie. Based on a wide experience in Colombia but also elsewhere in South America, it also tied in with the kind of "leading sector" models advocated by "big push"-type theories of development (a la Rosenstein-Rodan). Get these low-skilled, urban migrants into construction, they create their own demand and off you go. But how to get started? Currie's plan was to use indexed mortgages and forced savings. That meant you could avoid the front-loading problem  caused by high interest rates on large loans, and provide the pool of money required. Obviously, it would also imply providing useful investments, particularly if construction also related to infrastructures. Collier also recognises the issue of the need for access to finance, but also housing standards that are too demanding which raises prices, and weak property rights.

Is Collier reinventing the Currie leading sector approach? Indeed, I have heard him advocating the need to study the construction sector more closely and the cement sector in particular (as a further aside, I recall that provincial budgets in Mozambique were often highly concentrated in 'building", whether it be residences for district administrators, prison walls, or schools, but often, there were very few construction companies actually around to do anything beyond petty construction, making these ambitious plans rather impracticable).

In any case, his second reason why the Besley investment model may not work is that a difunctional state may be what the government wants in order to make personal wealth appropriation easier (think Mobutu - and also the Chabal & Deloz book I talked about here), and the third is that poor public expenditures may mean weak incentives for taxpayers to comply.

On this latter point, this is something which is receiving more and more attention, relating to discussion of "fiscal legitimacy" and accountability. But what Collier also mentions is something about "effective organisations", and how civil servants often "haven't internalised the objectives of their institutions". From having spent some time in a civil service, this may indeed be the case, but I wonder if it is really to do with the civil servants, and not the leaders of government institutions who have sometimes not internalised the objectives of their institution, but rather imposed their own objectives on their institution. Indeed, one of the areas I want to look at on tax policy is how tax policy is affected by the lack of intra-government coordination on tax policy formulation which, to me, is all about turf wars, personal ambition at the head of a particular organisation etc.

How would that fit in with the above? Well, perhaps by breaking down "government" into its constituent parts, you can then also make the case that even if the ruler wants to "invest" in taxation, if the mechanisms for making that investment are disfunctional, if capacity is poor, if understanding of the implications of these things is limited, and some individuals at the heads of certain institutions do not want to promote this "investment", then it will fall apart. So I'm going with your last explanation Paul. With a slight twist...

Anyway...