Shahrukh Wani

Published by the International Growth Centre (IGC) on the 3rd Dec 2018.

Pakistan tax-to-GDP ratio is about 12% . In comparison, OECD countries raise taxes equivalent to about 34% of their GDP. This limits Pakistan’s capacity to fund public investment. Where it does collect taxes, it principally relies on indirect taxes on goods and services, which account for 6.3% of GDP. The other 4.2% of the GDP that make up direct taxes are collected mainly by businesses withholding a percentage of economic transactions for the government. Hence, they do not require voluntary tax compliance from individuals, which, as a result, remains extremely low.

Start with improving enforcement

There is a relatively strong consensus that effective enforcement can lead to more tax revenue. Weak enforcement ranges from a property tax collector taking a bribe to not enforcing the tax, or a just a simple lack of capacity of the tax authority.

Improving enforcement may also result in more productive firms in Pakistan. Ilzetzki and Lagakos (2017) created a model of Pakistan’s economy to understand the impact that increased enforcement could have. They find that increasing enforcement will not only allow Pakistan to raise more tax revenue from firms in the short-run, but it would also result in higher growth by bringing these firms into the formal sector. By formalising, they would be able to make productivity enhancing investments and as a result be able to grow.

Enhancing enforcement requires Pakistan to do two things. First, hire better trained staff who have access to the right technology and resources. Piracha and Moore (2015), using a case study of Punjab, Pakistan’s largest province, find that each property tax circle needs to collect on average about $350 a month themselves to cover operational costs of their offices.

Furthermore, Pakistan will also have to focus on aligning the incentives of the bureaucrats associated with tax collection, with the incentives of the government. This is the classic principal-agent problem: politicians delegate tax collection power to bureaucrats, but cannot monitor them perfectly. At the same time, the front-line bureaucrat will most likely have better information. Due to this information asymmetry, tax collectors need to be effectively incentivised to fully and fairly enforce taxes. To this end, Khan et al. (2016) collaborated with the Punjab government to run an experimental study by assigning different incentive

schemes to property tax collectors. They found that over two years of the study, performance incentives encouraged tax collectors to add new properties to the tax collection roster, significantly increasing revenue.

Information is the key

Most developing economies have large proportion of their economies based on informal, cash-driven transactions, which means they are not captured by the tax base. There are a few ways Pakistan can attempt to create better information. Evidence suggests that implementation of a value added tax (VAT) helps raise more revenue by generating a paper trail between firms through cross-reporting of liabilities. This is evident in the paper by Pomeranz (2015) in the case of Chile.

Building on this line of thought, Pakistan’s VAT structure has certain challenges. First, intuition suggests that VAT should be applied uniformly by a single agency. In Pakistan, it is however split between federal and provincial governments, creating administrative and coordination challenges.

Second, by providing extensive exemptions in the VAT, Pakistan distorts the information trail; VAT is levied at 17% on the majority of goods, however, several goods have a either a lower rate or are exempt. One example is the exemption of red chillies but not green chillies.

Empower cities to tax

Similar to other countries, Pakistan’s cities are its principal engines of growth. Due to rapid urbanisation about 38% of Pakistanis are estimated to live in cities, yet they contribute to about 55% of the GDP.

These cities both require extensive public investment, and are sources of revenue potential.

Presently, Pakistan does not have financially empowered local urban governments, instead, most urban taxes are implemented by one of Pakistan’s four provincial governments. These provincial governments have large jurisdictions, with populations ranging from 12 million to over 110 million. As managing cities is not the central function of these governments, most of them have not developed effective urban administration mechanisms.

As noted through the synthesis of research conducted by the IGC’s Cities that Work initiative, land and physical properties are a major source of untapped revenue for most developing country cities. Punjab, for example, despite being home to nine cities home to over a million people, collected only Rs. 10 billion, or about 6% of its total tax revenue, from property taxes. Other parts of Pakistan have not fared better. Sindh, which is home to Karachi, Pakistan’s largest city, has not had a revaluation of land and property since 2001.

Yet there is large potential to increase this. For example, an estimate from the IGC (2011) shows that Punjab could more than double revenue from property taxes if it undertook comprehensive administrative reforms.

There is some hope here. Reforms passed earlier this year granted government the power to forcibly acquire any property that a citizen holds by paying 100% over the price they have declared in their tax returns (hence creating incentive for them to declare the true value of property). However, it is unclear whether these reforms will be implemented in practise.

Increasing taxation is possibly the most fundamental challenge Pakistan faces today, not only to adequately finance public investment and services, but also to establish a fairer society. To do so, it is useful to bridge the gap between research and policymaking by integrating research findings into policy decisions. This article points towards three avenues which can act as a point of departure for such a discourse.

 Originally published in Dawn.com on October 16, 2018 here (where it was the third most read article of the day, to my surprise). It was subsequently cross-posted at Scroll.in here. Header image by Dawn.com.

Pakistan’s formula for economic growth is as flawed as it gets: borrow foreign currency-denominated loans, build some large-scale infrastructure, get a minor growth spurt in the process, and wait until this growth spurt fades so we can repeat the process again.

This is what the previous government did. And, the one before that. It could have worked if, while borrowing to build infrastructure, it did not ignore the underlying constraints to growth and productivity.

Because they did not do that, Pakistan has ended up with an increasing level of debt, a balance of payment crises, and a government struggling to keep the growth spurt going.

When these challenges become dire — Pakistan often ends up getting a loan by the International Monetary Fund (IMF).

This time, if we’re successful in persuading them which looks to be the case, will be the 22nd occasion we will be loaned capital by the fund since 1958.

And, if our public discourse and policies remain the same, we will without doubt keep knocking at IMF’s door every few years (or some other lender for that matter).

The logical argument made by analysts in Pakistan here is that the government needs to bring meaningful reforms to our economy.

So, in due course, we are in a fiscally sound enough condition that we not require bailouts like the ones we get from the IMF.

This is a perfectly accurate demand. But, it often masks the politicalcauses to our economic despair.

The problem with talking about the economy divorced from politics is that we end up with superficial reforms.

This is because any meaningful reforms are impossible if the political structure does not allow them.

For this to change, our public discourse needs to take a holistic overview of our institutions. This is a contribution to that end.

Digging deeper

Let me explain. Take the much talked about balance of payment crisis as an example.

The most direct culprits are lack of exports and the increasing cost of imports. Pakistan imports nearly twice as many products and services than it exports.

In turn, there are many causes for low exports, some are macroeconomic determinants. The unsound infatuation of the previous government with an appreciated rupee is an example of this.

Dig deeper, we are exposed to the fact that Pakistan has developed little comparative advantage over the years. Which means that we mostly export basic textiles, cotton and rice and other related products.

Most of them are low-value items in the global value chain, so we earn little revenue from exporting them, and are hence unable to cover our import bill.

Dig further, we find that even in products which we do export, we face structural problems  —  such as lack of capital, whether it is human or financial.

Hence, exporters struggle to grow, move up the value chain, compete with foreign firms, and boost productivity.

But, why is this? Why does Pakistan fail to provide an environment which is conducive to developing globally competitive enterprises?

There are many explanations here — but one persuasive thought is that our institutions do not create the set of incentives needed for the growth of a competitive market.

Instead, they encourage a reliance on state patronage even if it comes at the cost of the larger industrial growth.

Pakistan’s public discourse has been rather good at ignoring these underlying causes. The result of this has been that, whenever Pakistan has found itself in such crises, it has been able to get loans to sail, or crawl, through the crisis.

But, all those reasons why this crisis emerged in the first place will remain — waiting to fuel another crisis down the road.

It is like putting the fire out but not fixing the leaking gas socket in the basement.

What are institutions?

A good place to start is understanding what institutions are, and how they influence our collective behaviour.

Simply put, institutions are the ‘rules of the game’ as Douglas North popularly put it in his 1990 book.

Think about cricket — we have certain rules under which everyone plays the game. There has to be a fixed number of players, they have a number of balls to play with, and everyone has a consensus on how we determine which team wins.

Now, apply the intuition behind this analogy to the broader institutions which shape our daily life.

Like the rules which govern the game of cricket, we have humanly devised rules which govern our lives in a more consequential manner.

They can be formal which are written down, such as the laws of the land often codified in the constitution, or informal ones widely accepted by the population, such as kinship bonds.

These institutions are important because without them, society as we know it would collapse. But, these institutions differ from country to country, and a persuasive strand of economic literature argues that they are the key to understanding our development outcomes.

We can also split them between economic and political institutions. The former directly shape our economic incentives, such as ownership rights of property, and the latter determine the political structure, such as whether we’re a parliamentary democracy or not.

As a society, we decide which economic and political institutions to adopt, and these institutions shape much of our behaviour through shaping our incentives.

Incentives, any economist can tell you, are fundamental to understanding any society’s prosperity or lack thereof.

It’s not economics, stupid. It’s politics

The political institutions, I’d argue, are more important in determining our prosperity. As Acemoglu and Robinson argue, those who control the political power determine economic institutions.

So, if political power (which in turn determines the political institutions) is controlled by a small, extractive elite, they will set up economic institutions which benefit them, not the majority.

If the elite benefit from an economy underpinned by clientelism and patronage rather than a well-functioning competitive economy, they will choose the former.

It is important to remember that there is plenty of profit in poverty. It just happens to be controlled by few.

Now, look at Pakistan. Our political economy is defined by an embedded culture of rent-seeking and patronage.

This means we have a system which grants profits to certain players in our economy unfairly, hence undermining the central principle of efficient market allocation — fair competition — and creating a wrong set of incentives for businesses.

Our manufacturing sector is rife with examples of rent-seeking practices. For example, Pakistan’s automobile sector is dominated by a handful of Japanese manufacturers known for selling low-value cars while making a considerable profit.

Despite this, Pakistan provides them with extensive trade barriers to protect them from foreign competition. The recent finance bill (since amended) further shows the extent of their political patronage.

This is not by accident. Because Pakistan’s political power is controlled by an extractive elite, it has allowed for such political institutions to emerge which permit its government to provide these rents to certain car manufactures with immunity, even if this negatively impacts our shared prosperity.

Direct evidence of our political structure influencing economic outcomes comes from a paper by Asim Ijaz Khwaja and Atif Mian.

They show that politically connected firms in Pakistan receive loans from government banks in Pakistan at lower rates despite defaulting more than non-politically connected firms.

This is evidence of unaccountable political power translating into inefficient economic allocation.

Look at this from another angle. A large amount of economic history literature argues that one of the many reasons why some East Asian countries prospered in the second half of the 20th century was because of land reforms.

By undertaking substantial land reforms, these countries were not only to increase agricultural output but also raise living standards for their surplus labour, which, in the long run, subsidised their move towards industrialisation.

If land reforms are important for growth, would Pakistan ever adopt them in any meaningful way?

Despite attempts to do so, Pakistan has not gone through significant land reforms and over 10 million acres of land still remains under tenancy, while one estimate puts the average farm size in Pakistan at about six acres.

In South Korea, despite a significant increase in average farm size over the recent past, it still averages at about 3.5 acres (as of 2005).

Without making a deep dive into the history or merits of land reforms, if we agree for the sake of argument that more radical land reforms are needed in Pakistan — can we undertake such reforms when political power and institutions are so highly influenced by those who control large farm holdings?

In other words, if feudal lords control the political institutions, it shouldn’t be a surprise that economic outcomes will favour them.

To fix the economy, focus on the political discourse

What may seem to a passerby as a country which continues to choose poorly thought-out economic policies, sees rampant corruption and a failure to establish a productive industrial base, are in fact symptoms of the political institutional structure which benefits a narrow extractive governing elite at the cost of everyone else.

Our economic failure is a symptom of our collective political choices. Once we can allocate political power more fairly, we can make better economic outcomes.

Tweaking institutions at the margins does have some impact. Hence, the IMF’s stabilisation programme will provide some macroeconomic stability.

The stock market might recover, the fiscal deficit might get narrower. Taxes might increase a bit, so will inflation.

And, in due course, we will issue a statement saying goodbye to the IMF for few years. Before repeating the process again and again.

But, if the new government wants to break this cycle and make a sincere attempt at reforming Pakistan into some sort of an egalitarian, prosperous nation, it needs to start by looking at political power and the political institutions which rise from them, as they are the real constraints to our growth.

Even if it can make marginal changes on the economic front, they would not unlock the kind of transformative shift we need for widespread prosperity.

I just completed Joe Studwell’s new book “How Asia Works: Success and Failure in the World’s Most Dynamic Region.” Here are some thoughts I tweeted about after reading it.

1) Size of farms matters! Studwell argues that having a small (3ish acre) family-owned farms boosts agriculture productivity (read: land reforms!). There are various loopholes countries (like Philippines/Indonesia did) can fall in while doing this.

2) This will create a surplus which can be used to set up an export-oriented manufacturing base. (exporting makes companies productive). This is pretty much in line with broader evidence: the state protects certain industries, but also punishes losers if they are not productive.

3) Meanwhile, these countries set up a financial sector which promoted these export-oriented industries. But, here is the important part: these Asian countries made sure that the financial sector’s incentives were to assist long-term dev. goals, not speculative short-term ones.

4) Throughout this journey promoting exports is important. Because this is the best way to push your industries to become more productive. In other words: competition=developing comparative advantage. Let your industries fight with each other, and more productive ones globally.

5) Now think about Pakistan. We didn’t go through the kind of land reforms these countries did, so we don’t have the egalitarian, productive base. We protect inefficient manufactures despite them to not being globally competitive. We even had an over-valued exchange rate.

6) Having an equalitarian land ownership structure is good for development trajectory – I didn’t know that it is good for productivity too! Despite economic intuition arguing otherwise: small farms are good. So, it might be wise to talk more about agrarian reforms in Pakistan.

7) Another thing for us to think about is that the state vs market paradigm is flawed – there is no such binary relationship in the real world! Gov.’s need to work to promote market efficiency not protect patronage industries (I’m looking at you the Pakistan’s car industry)

8) Lastly, this is a great book and you should read it (no technical stuff) But, at the end of the day some of this stuff will work for Pakistan, others will not. Every country has its own path – but there is some stuff I’m willing to bet will work, agrarian reforms is one of them.

This is an archived version of an article written for Dawn.com, published on Aug, 28th 2018. If you want to share it, please do so the version on Dawn.com. 

Pakistan has a new leader at the helm who, if his first address to the nation is to be considered at face value, is set to make a radical departure from the conventional trajectory we have taken as a nation to create what he describes as a Scandinavian-style Islamic welfare state.

Keeping aside the intellectual ambiguity of coupling liberal Scandinavia with Islamic theology, if Pakistan is to soberly contemplate such a goal, it is useful to critically engage the feasibility and the implications of such a proposal.

What makes a welfare state?

While welfare states vary in characteristics, the term is used to indicate a type of social contract under which the state undertakes extensive intervention in the society to protect or assist those it considers to be disadvantaged or going through some form of a shock.

In other words, providing a form of social protection.

In practice, this could be through a number of interventions. Providing free or subsidised education and healthcare is a common one. So is unemployment insurance for people who get laid off or providing pensions for the elderly.

Most countries, including Pakistan, already have some welfare policies. For example, the Benazir Income Support Program is a welfare programme for those living in poverty.

What makes the Scandinavian states unique is the breadth of these interventions. For example, Denmark spends 29 percent of its GDP on social welfare spending (this does not include education), while South Korea spends about 10pc.

The countries which make up the European Union generally have a high welfare spending. Cumulatively, they dedicate about one-thirdof their GDP on social welfare — the largest chunk on pensions and other forms of old-age benefits.

Here, most people spend their work life contributing to the welfare state, and once they retire or go through a shock, they benefit from it.

By whatever means you measure, this is a lot of investment.

What will it take to turn Pakistan into a welfare state?

The first thing which is common among the Scandinavian welfare states is they collect a large proportion of their GDP in taxes.

Norway collects 38pc, Denmark about 45pc, and Sweden about 44pc. This means that these states have a large and extensive tax infrastructure — and people pay a large proportion of their income in taxes. The rich pay more: in Denmark, the top marginal tax rate is about 60pc.

Compare this with Pakistan — we collect only 12pc of our GDP in taxes. Denmark spends over twice as much of its GDP on social welfare alone. After recent tax reforms in Pakistan, the top marginal tax rate is just 15pc — significantly lower than Scandinavian countries.

This means that, if we were to make a move towards becoming a welfare state, we would need to radically expand the tax system — not only taxing more people but also taxing more the (higher income) people who do pay taxes.

As welfare states require significant redistribution and government provision of public services, these tax revenues would be used to finance the growth in the size of the state, both by spending more on welfare services but also hiring more people to manage such a far-reaching infrastructure.

This is the opposite of austerity.

But, as those who have been at the helm before know very well, raising taxes is hard everywhere, but significantly harder in Pakistan.

This is because we have a cash-based economy with a large informal sector which makes it hard for the government to know how much exactly people owe. So, a large number of people are able to evade taxes, and evidence suggests that they do.

One way to solve could be to bank on some form of intrinsic motivation for them to voluntary comply. However, the evidence is unclear on how much taxes can be raised from the voluntary channel alone.

What is clear is that many people in developed countries, including most welfare states, do not have the opportunity to evade taxes.

This is because of third-party reporting of income thanks to high formalisation of the economy and a strong paper trail which makes getting caught likely, and costly.

So, in order to become a welfare state, Pakistan will need to raise a lot more taxes – which can be achieved in part by making it harder for people to evade them.

Along with taxes, what makes Pakistan different from Scandinavia is that they are significantly wealthier than us. The per capita income in Pakistan is about $1,500. In Norway, it is over $70,000.

Now, you can make the argument that what makes Scandinavia rich is its welfare infrastructure, or that Scandinavia being rich allows it to have a welfare infrastructure.

But, what is clear that their economy is far more productive and people are far wealthier than us which allows them to spend heavily on social welfare.

And, if Pakistan wants to have public services on par with theirs, it would need to drastically expand the size of its economy.

Take the southern European nations, for example. They set up extensive welfare systems, promising people gangrenous pensions, but their economies were not robust enough to back such a system. History is a witness to what happened next.

But setting up a welfare state is also more than taxes and income. It is about establishing a social contract. A contract between a person living in a farmhouse in Islamabad and a person living in a chaunra in Tharparkar.

It is about convincing people that their national identity trumps their affiliations with social class, creed or kin.

Consider this. Norway, Denmark and Sweden are by and large ethnically homogenous countries. That means most people share the same ethnicity which, evidence suggests, feeds into the social preference for redistribution.

People tend to be willing to help other people who look like them.

Pakistan is not a homogenous country, and has, perhaps as a consequence of that, strong informal networks. For instance, if a villager suffers from an illness they are often helped by people of that kin group – biradari.

It means that even when there is no formal welfare infrastructure, people tend to create pockets of an informal support network to help each other out.

But to convince people that a formal institutionalised welfare state is the goal would require to build an overreaching identity which triumphs local ties.

Demography, broadly, is also important. In many welfare states, an aging population is a big challenge — too few working-age people contributing to the state with taxes, and too many old people taking pensions and healthcare.

A Pakistani welfare state, however, is likely to have the opposite problem. We do not have an aging population but have a very young one.

Spending on education is likely to take a large chunk of social investment, and an increasing size of the population means that the welfare state would need continuous expansion to cater for a growing population.

This could be managed if the move towards a welfare state is combined with measures to control population growth.

So, is Welfare-istan possible?

Yes, if we can build an extensive tax infrastructure, a significantly more productive economy, a stronger national identity with preferences of redistribution and control rapid population growth to make this all feasible.

The pitfalls are also clear – if you expand welfare spending without improving the tax system, you might end up in a budgetary crisis (and end up taking debt to manage it).

If you fix the tax system, but the economy does not grow fast enough, you would not raise enough revenue and would end up promising people something you cannot afford.

This, as our new leader will learn soon, is easier said than done.

This article was commissioned by Alif Ailaan – a DFID funded, project implemented by DAI Pakistan. You can read it here: https://elections.alifailaan.pk/the-case-for-a-culture-of-learning/. A version of this is also published in Urdu here.

---

For what it’s worth, Pakistanis have created a lot of noise concerning its education crisis. Enough to get the government’s attention. To some extent, it has paid off. It is the centre – or near the centre – of our public debate. Despite all this, over 20 million are still out of school. Those in schools aren’t doing much better either. As the debate becomes noisier, it is useful to take a step back and understand the true cost we’re incurring as a nation of raising a generation in the dysfunctional education system. The solution is not only to improve our schools but also to do something bigger, something more ambitious: develop a culture of learning. A culture which encourages its young to be curious, push them to learn for the sake of learning; to think, to write, and to debate. Anything short of this would be setting Pakistan to fail.

Imagine the potential each of us has. If you’re reading this, it means you’re literate. You already have the most critical skill one needs to access knowledge. But at least 4 out of 10 of our fellow citizens can’t read a sentence in any language, let alone the number of Pakistanis who can read an article with ease. This means that if you’re able to read this and comprehend the argument I’m trying to make here, you’re likely to be a minority in Pakistan. We have heard numbers to this effect before, but  have we understood the implications of this? Majority of Pakistanis, for no fault of their own, cannot do the very basic act. This societal failure represents one of the most tragic losses of potential both at the individual and national levels.

It shouldn’t come as a surprise that this loss of potential is a manifestation of something we have spectacularly failed to develop, a culture of learning. A culture which encourages people to think about the world around them, to question others, to seek out new ideas, and above all, to stimulate a healthy social debate which drives innovation and progress. Take libraries for example. The overwhelming majority of us don’t have access to one. The instrumental value of this is clear, those who cannot afford to buy a book, don’t have access to books and, by extension, information. But the intrinsic value is also critical. The absence of libraries means that pleasure and happiness in pursuing knowledge for the sake of itself become limited to those who can afford to do so. What greater tragedy could there be.

It is not just about building libraries, but also how we sow the seeds of curiosity. In our schools, students are forced to rote learn, instead of thinking critically about what they are being taught. Any student who spends enough time looking at an information can memorise it and forget shortly after the exam, but the challenge is to  use that information to understand the world around them and come up with solutions to problems. This requires encouraging students to question, not staying quiet and simply agreeing with the teacher and regurgitating exactly what was fed. It is important to encourage them to work in groups, thinkg about the practical implications of what they are learning, and at the end of the day instilling in them  a life-long desire to continue to do so when as they grow up. Once these children become parents, they can do so to with their children. And a culture of learning is born.

We need to improve the quality of schools, build infrastructure, and hire more teachers. We need all of this but, at the end of the day, we also need something broader. We need curious students, life-long learners, libraries which can enable that. For education reformers today, it’s important to understand the future society that we want. It requires setting an ambitious long-term vision of the role schools have to play in social change. One way this can be done is  alongside teaching core skills such as maths and languages, encouraging students to read, speak and write clearly, and share their thoughts with their peers. This demands greater focus on curriculum design. Right now, Pakistan focuses on forcing students to memorise a lot of content and learning little.

“We have inherited a house,”Abdus Salam once put it, “which has no windows and its walls are very high, and it’s very difficult to know whether we have inherited a house or a prison.” This prison is not a literal one, it is, at least in my reading of this quote, one of imagination and curiosity. If we want to crawl out of this prison, we need to start at schools. As we go out to vote in general elections this summer, remember that we need to be ambitious in what we ask. We’re not children of a lesser God.

The article was published by The Guardian.

With more than 20 million children out of school, Pakistan has, at last, begun talking about its education crises. Our media and civil society routinely grill politicians on a lack of funding for public schools. Opinion sections of national newspapers usually publish a few articles a week on how the lack of quality education is becoming an existential threat to Pakistan’s social cohesion. Foreign aid funded projects take primetime television ads to tell parents about the importance of educating their children.

It has had some impact; education has become a key talking point in political debates. The government regularly boasts about the growing education budget with promises to provide an “excellent environment” to students. But what is lacking in this increasingly noisy debate on Pakistan’s education crisis is the experience of parents and students on the ground.

The lack of nuanced policy is leading to an alarming trend. Education spending on the ground is being translated into schools as a means to provide jobs, rather than to provide children with a quality education. One of the leading research papers in this field is Pakistan’s Education Crisis: The Real Story by Wilson Center fellow Nadia Naviwala, and Ahmad Ali from the Institute of Social and Policy Sciences in Islamabad.

There is a very strong political element to education spending, as legislators are customarily elected on the basis of how many jobs they can provide to their constituents, and hence hiring new teachers takes priority in budget allocation, particularly when close to a general election. For the government, this preoccupation appears to kill two birds with one stone; an easy fix for the education crises, and sought after permanent government jobs for their constituents. As a consequence, education departments are typically the single largest employers in most provinces.

Increasing the number of teachers across the country has also been an easy policy for everyone to get behind, especially since the public discourse on fixing the education crisis has largely been focused on the need to spend more on education. In 2016, Pakistani provinces spent between 17 to 28% of their budgets on education, while the global average was 14%. Combined that’s $7.5bn spent on public education nationally, with most provinces doubling their budgets within the past five years.

But as Naviwala has argued, Pakistan can’t simply spend its way out of its education crisis. On the ground, most of the spending is being used for hiring non-performing teachers or providing salary hikes for existing teachers. A small section of the spending is set aside for new education infrastructure however about half of it, on average, goes unspent by provinces every year. In fact, the proportion of spending on much-needed education infrastructure has decreased, as salaries take a larger than ever proportion of the spending total.

The problem is that this rapid rise in spending isn’t translating into education for all. School enrollment nationally has continued to stagnate. Even if enrollment drives are able to get students into schools, evidence shows that only one in four children who enroll in the first grade remains in school by the 10th grade. Even of those students who remain in school, most aren’t learning basic skills like literacy. Studies have shown that over half of all 3rd graders, children aged 9-10, in government schools are illiterate.

The 4% GDP public education spending target – which is promoted as the gold standard by donors and campaigners – seems to provide a simplistic solution to Pakistan’s education crisis. Take for example, Sindh, Pakistan’s second largest province. In 2016, the province spent 12 times more on teacher salaries than it did in 2010; despite this increase the majority of its fifth graders can’t read a 2nd-grade level text. Sindh has been on the forefront of a general trend of turning public education departments into employment agencies.

Ironically, Sindh’s provincial government has itself admitted that as many as 40% of its schoolteachers are ghost employees, meaning that these teachers don’t show up for work, and in some cases they don’t live in the towns where they are posted to teach. But the same teachers keep getting pay rises. Unsurprisingly, this toxic combination of political patronage and ineffective policies has seen a decline in net school enrollment rate in the province.

Parents rightly ask what’s the benefit for them in sending their children to school when the children are unlikely to learn? Especially when these children are economically more beneficial for families when they are working.

Pakistan has to build more schools and train its teachers. That requires it to “spend better, not simply more”. Something international donors need to understand. Pakistan’s private education sector – which educates about 40% of the country’s students – can serve as an example here.

Private schools on average pay five times less to their teachers than their public school counterparts. What’s even more fascinating is that private school teachers also have lower academic qualifications and are half as likely to receive training than teachers employed in government schools. However, despite this, the students at these schools are on average two grades ahead of their government school peers.

Why is it then private school teachers deliver better? At the core, private school teachers feel more accountable to parents. Parents can pull their fee-paying children out of school if they believe that their children aren’t learning the key skills. This makes teachers and private school administrators more efficient at delivering quality education. This isn’t the case in government schools who are answerable to usually an education department located often miles away.

This means that public school teachers in Pakistan are rarely fired – even if they aren’t delivering in classrooms. How teachers are hired and promoted is also crucial. Public school teachers regularly report that politics play a strong role in these decisions, at times hinting towards outright political patronage.

Ultimately, Pakistan needs to combine any spending increases with critical reforms that develop accountability between teachers and parents. It needs to look into the curriculum it is teaching in classrooms, and develop an intensive monitoring system that makes sure that teachers are showing up for work and delivering quality education. For too long, our leaders have been focusing on expanding the size of the country’s education budget and falsely marketing it as showing the government’s growing commitment to quality education.

For change to happen the quality of the conversation around Pakistan’s education crisis needs to improve. The people need to demand better teachers who are accountable to the communities they teach in, while donors need to rethink about their fixation with unrealistic funding targets and instead support reforms which make schools more efficient.