Political business cycle in Bangladesh

Posted in economics, governance, political economy, politics by jrahman on October 10, 2015

While looking for something else, I recently discovered that in 2013-14 financial year, Bangladesh government’s current expenditure — that is, money spent on salaries, procurement, subsidies and loan repayments — increased by 0.8 per cent of GDP.  On top of that, government investment also increased by 0.7 per cent of GDP.

That is a lot of spending.

Now, I haven’t been following economic and policy developments in as much detail as in the past, so I may well have missed great policy initiatives that underline the apparent spending binge.  But according to the work of IMF’s Christian Ebeke and Dilan Olcer of Riksbankrise in government expenditure of that magnitude is not uncommon in election years in developing countries.  More worryingly, they show that this kind of binge usually ends badly.

The authors study how public finances vary over electoral cycles in 68 developing countries during the 1990-2010 period.  Their key finding is captured by this figure:

imf paper

In English, this shows that in the year of the election, government spending on salaries, subsidies and procurement — that is, stuff that can help government get a popularity boost or facilitate patronage — rises by about 0.8 per cent of GDP.  After the election, however, the time of reckoning arrives.  That expenditure has to be paid for, and it is done by cutting public investment and raising taxes such as tariff and import duties.  Not only is the initial expenditure typically wasteful, the way it’s paid for is usually harmful for the economy’s long-term growth prospects.

Again, I am not sure how the spending is being financed, and what has happened in 2014-15 financial year.  I do, however, have data for the 2006-07 election cycle.  That year, current expenditure actually fell by 0.7 per cent of GDP, public investment falling by another 0.7 per cent of GDP.

Why the difference?  Didn’t the then BNP government want to win?

Of course they did.  But remember, that government was turfed out halfway through the financial year, drying up the patronage pipeline.

That led me to think about political business cycle Bangladesh style.  Other countries have election cycles.  In the quarter century since the fall of the Ershad regime, Bangladesh has had something more, whereby the parties vie for the control of the street before the election.  That street politics of hartal-oborodh-oshohojog andolon should have an impact on the economy, right?

More specifically, to the extent that the point of the opposition street politics is to force the incumbent to change election ‘ground rules’, the enforced ‘economic shutdown’ should be clear in the data.  A stylised example might be that the opposition takes control of the ports, disrupts highways and rail networks, and the export of garments come to a halt,  the big garmentswallahs start working the corridors of our narrow incestuous pool of the establishment, and before you know it uniformed men politely inform the ruling lady that things have to change.

Is this stylised story evident in data?

Ideally, we would regress the impact of andolon on growth in ready made garments export after accounting for other factors such as global conditions.  But this is a blog post, not a research paper!  Let’s try something else.

This chart shows through-the-year growth at exports of readymade garments over four episodes: 1995-96, 2006-07, 2013-14 and 2015.  To the uninitiated, it might be a bit counterintuitive, so bear with me.


For each episode, there is a ‘zero month’: February 1996, January 2007, January 2014, and January 2015 — these are depicted by the horizontal axis in the chart.  We might expect the opposition andolon to gradually escalate to the zero month, so the chart starts three months earlier.  If the stylised story is right, exports growth would slow as the zero month approaches, and would eventually reach a ‘crisis point’.

That’s pretty much what happened in 1995-96.  RMG exports grew by 15 per cent in the year to November 1995, but growth ground to a halt by February 1996, and then plummeted by over 20 per cent in the year to April 1996, quite likely as a result of the continuous strikes.

The story doesn’t seem to work in the two subsequent episodes.  Remember all the horror stories of the economy being destroyed by the ‘battling begums’?  Well, in both 2006-07 and 2013-14, there doesn’t seem to be any discernible impact of political uncertainty on the RMG export growth.

Let me postulate another stylised story then.  In 1995-96, our garmentswallahs didn’t know what to expect.  In 2006-07 and 2013-14, the political shock was fully anticipated.  Everyone knew the ‘zero month’ of January.  Everyone knew the opposition demands.  Everyone knew the government stance.  Everyone knew the equation.  And stakeholders took appropriate measures.  Wonderful thing modern market economy — it’s not that easy to shut it down, let alone destroy, specially not if you shout your intentions in advance.

Of course, the latest episode — the dotted black line — wasn’t anticipated.  As 2014 drew to a close, no one expected BNP to enforce a nationwide shutdown in the new year.  And yet, that’s exactly what happened in January-February 2015.  And by April, RMG growth slowed sharply.  Perhaps the regime was on the brink after all, and the city council elections saved it.

Or perhaps not.  There is a global trade slowdown — this is why we need a proper econometric assessment.

Now, if a regime was indeed on the brink, one might expect its cronies to siphon their money out of the country — one hears all sorts of tales about neighbourhoods in Canada full of Bangladeshi begums, or properties in Malaysia, or suitcases full of gold to the Middle East.  If there is any truth to these tales, they would show up in the taka-dollar exchange rate.  They have to, since taka is worthless overseas.

The chart below shows monthly movements in the taka-dollar exchange rate over the same four episodes with same zero months.  In this chart, a positive number means depreciation of taka.  Keep in mind that since December 1971, taka has on average depreciated by 13 poisha a month against the dollar.

Look at the dotted black line or the green line — hardly any movement.  Taka held remarkably steady during the first months of 2015 or 2014.  If the regime was on the brinks, its cronies had no idea.  Just in case you are wondering, Bangladesh Bank’s foreign reserves have been consistently on the rise in recent years — going from five months of imports in 2013 to over seven now.  That is, there is no evidence of large scale capital flight in recent years.

Contrast this with the two earlier episodes.  In November 2006, despite strong exports growth, taka depreciated sharply.  Taka also depreciated in April-May 1996, after the BNP government was toppled.


It would appear that while Awami cronies the establishment (that is, the informed elite) have stuck by their patron the current regime, but BNP’s cronies  abandoned theirs fecklessly the incumbent in 1996 and 2007.  Why does Awami League have more loyal cronies?  Is it because the current regime is deemed more competent than the alternative?


Comments Off on Political business cycle in Bangladesh

%d bloggers like this: