Paying the Piper: Per Diems and Other Allowances in Tanzania

In my daily digest of news around the world, I receive articles from the New York Times, Foreign Policy, the Economist, among others.  On Africa and Tanzania, the majority of these articles are usually “big picture”  pieces about the state of affairs in the continent or Tanzania like most recently, the New York Times’ piece on conflict minerals or Foreign Policy’s piece on China‘s involvement in Africa‘s construction boom.  So it was a pleasant surprise this morning when I saw the Center for Global Development‘s William Savedoff highlighting Twaweza‘s policy brief 9.09 which looks at the government of Tanzania‘s scheme of allowances.

The brief highlights the different types of allowances civil servants get and what shares these constitute of the total allowance budget.  Twaweza go on to decompose the difference between remunerative and duty enhancing allowances.  The latter being allowances “thought to improve productivity and service delivery, while [the former]… allowances have a less clear association with quality service delivery.” So in practice, “duty enhancing allowances comprise expenses for training, while remunerative are expenses for housing.”  The distinction, however, is not so clear and some of the duty enhancing allowances, such as housing only accrue to higher level officers and thus would constitute a remunerative allowance rather than a duty enhancing allowance.

The brief goes on to argue that the high levels of allowances relative to wages skew incentives away from quality service delivery.  For instance, “the average civil servant working in one of the Ministries of Central Government receives more than five times as much in allowances as the civil servant delivering services at community level.”  Allowances are also higher than what is necessary.  “A civil servant spending a day in the rural areas easily pockets a per diem to cover expenses for food and beverages that is 10 to 20 times larger than the food poverty line for a family of five.” (Emphasis mine)

The nature of this allowance system only creates perverse incentives where people seek to get more training, attend more meetings and so forth.  What is worse is that donors seem not to have caught on and by their finance, perpetuate this system.  The brief provides a particularly illustrative example of this when it details a generic example of how government would go about booking a capacity building meeting in Dar es Salaam.  In looking for evidence for excessive and unnecessary travel, the brief goes on to cite the most embarrassing example.  During the 2009 International Monetary Fund or IMF and World Bank annual meeting in Istanbul, Turkey from 2nd to 8th October Tanzania‘s delegation included 25 individuals.  Although this is far less than the Wu-Tang Clan‘s entourage, officially known as Killa Beez, Tanzania‘s delegation was one short of the combined total of the Kenyan, Ugandan, Rwandan and Burundi delegations.  Tanzania‘s delegation was 5 times the size of the Rwandan delegation, just under 278% the size of the Kenyan delegation,  5 times the size of the Burundi delegation, and 357% the size of the Ugandan delegation.  This is why Rwanda will be first to overtake Kenya in the race to middle income among the East African Community or EAC countries, and has increased per capita income by 62%.  Tanzania will get there, but at a much slower pace.

So, what is the solution?  The brief concludes beautifully and I provide the full text here because my paraphrasing will not do it justice.  “If allowances are reduced to a (still respectable) 30% of the wage bill, it would free Tshs 285 billion in the Government’s budget. In addition, another Tshs 100 billion could be saved from reducing use of business class tickets, transport costs and accommodation, bringing total potential savings to Tshs 385 billion or almost $300 million.  This is an enormous amount that could be used to reduce taxes and borrowing, to hire additional teachers and nurses, to buy vaccines and bed nets or to invest in roads, bridges, water, and electricity generation.”  The brief‘s concluding paragraph is also worth quoting in its entirety and I provide the full text below as well.

“The saved resources could also be channelled into higher salaries for civil servants, who because of the cost savings would become better off than they are now with allowances. And as lower allowances are likely to improve incentives for service delivery, reducing allowances can be expected to lead to better public services, in particular when salary enhancements are used to align staff incentives with service delivery. In other words there is scope for a broad coalition of citizens and civil servants to change the allowance culture. The rationale is there: now it is time to act.”

Mind you, neither Twaweza nor I is trying to advocate scrapping performance-based pay for civil servants in Tanzania.  A lot of empirical evidence has repeatedly shown the link between performance pay and improved performance.  In education, for instance, research in Israel has shown that raising the principal’s wages improves school outcomes, while research in India has shown that paying bonuses to teachers raises students’ test scores.  We also know that aid, when tied to performance can be very beneficial.  Research in Indonesia shows that aid provision based on prior performance improved health and education.

In short: incentives matter.  We also know, however, that incentives are not a panacea for solving problems.  There are instances where payment schemes do not provide beneficial services to clients.  Another research in India shows that insurance agents in seeking high commissions tend to sell products to consumers that are not necessarily beneficial to those consumers.  Less sophisticated consumers are most likely to succumb to being sold these less beneficial products, highlighting the need for more informed consumers (insert citizens here in our case).  Citizens still bear the burden of seeking and getting information so that they can hold government responsible.  Tanzanian citizens must demand greater accountability that will see allowances fall in line with performance.

Although the longevity of this status quo of allowances makes it quite unlikely that policy will change, civil society should encourage and empower the citizenry to demand greater accountability.  Now, am not a proponent of government intervention (see blog posts here, and here for greater details of my views on government intervention) but evidence shows that paying taxes makes citizens more likely to hold politicians accountable.  If taxation is the only way Tanzanians will hold their politicians accountable, then let us increase the long reach of Tanzania’s Revenue Authority or TRA into people’s pockets (They already take about 29% of my income, and seeing my involvement in the discussion about the government’s accountability, I think am set).  It seems that not only is taxation without representation tyrannical but also it seems that taxation is representation.  Let us incentivize the Tanzanian public to hold Tanzanian politicians accountable, or pay the piper and face the music.