Politics and advice
Local elections are approaching. There are two novelties: First, there will be blocs and electoral agreements ahead of elections, a first for local elections. Ordinarily, parties may enter into electoral agreements in general elections, not in municipal ones. This shows there is at least a symbolic meaning attached to the coming polls. The electorate is expected to convey a signal of approval for the new presidential regime, and local elections thus gain a meaning that surpasses their usual importance. Second, again for the first time, local elections are being held after the general elections and the presidential vote. Hence, the party in power is known beforehand. Now, this is crucial because Ankara is still holds all the cards in allocating resources. Therefore, I expect a bias in favor of the incumbents. Because there is also an ongoing malaise – at least stagflation, at worst recession - some kind of economic voting will be important. There will be thus two counteracting dynamics at work. The upshot is I don’t think public finances will be as tight as pretended ahead of elections. This doesn’t imply there will be a renewed effort in vain to shore up private consumption in full, but there will be incentives working in that direction nonetheless. Why is that so?
Negat ve loan growth and vot ng behav or
This can’t go on until the municipal elections. Because they matter more than they appear, they have acquired a symbolic significance. Once more, this can be considered as ‘approval voting.’ As we can see from the graphic, for the first time after Lehman, deposit growth exceeds loan growth. ‘Muddling through’ requires at least some kind of rekindling, resetting, rebooting on the loan side. Why is that so? Loan data tell the tale of consumer and corporate lending in the ‘long view’, so to speak, i.e. over the last 17 years as percentages of GDP. If we take the ‘old GDP series’ in fact, calculated in 1998 constant lira, we see that from 2013 to 2016 – until the new series are produced - the real GDP has been up by a coefficient of 1.81. If we take into consideration 2016 and 2017, we can, without loss of generality, assume that the multiple is approximately 2. Total consumer loans –including mortgages, car loans, general purpose loans etc. - were only 1.77 percent of GDP in 2002, right after the 2001 Crisis, and they stand at 15.11 percent now. These are the amounts lent by saving deposit banks (commercial banks) only, which would make a difference when corporate loans are considered because nondeposit banks also lend heavily to that segment. Well, the outstanding consumer credit-to-GDP ratio has been up by a multiple of 8.5 over the last 15 years. If GDP has doubled up in real terms, that means a 17-fold real increase. The same applies to corporate loans and there, the figures are 5.8 and 11.6 respectively. Now this is unprecedented in Turkish economic history, especially the real growth in consumer lending. Even more astonishing is the continuous increase in the consumer loan-toGDP ratio over a span of 11 years. From 2003 to 2014 it always went up. Of course, it did go up in 2002 also because this was the immediate aftermath of the crisis. So, the whole process adds up to 12 years, of which 11 have elapsed under AK Party governments. Furthermore, with the exception of the Q4 2008-Q3 2009 period – 4 quarters immediately after Lehman - GDP has also grown.
This, I would call the “credit effect”. Now add to this the “exchange rate effect” of the first five years of the AK Party when the Lira appreciated. It might as well be labelled the ‘exchange rate illusion’ for two reasons but its impact on voters’ minds was real. This is also the period when mortgage and car loans skyrocketed – as they did by around 400 percent in 2005, a year when all Turkish assets had performed admirably, and the quality of cars had visibly improved due to supply chains. Combined, these developments created a “wealth effect”. True, both household and private corporate debt rose steeply during that period, but there was also the asset side of the coin. To complete the picture, we can also include infrastructure and real estate developments, transfer expenditures that consistently increased above inflation, health care reforms and suchlike. Now, some of them were perhaps only temporary and in part illusory but they were deeply felt by large chunks of the population.
Credit data also tell us that the consumer loans-to-GDP ratio stopped rising in 2014, and it fell a bit after that. So, if many voters vote ‘economically’, why don’t they switch to other parties? In fact, they did once in 2015. Let us not forget, however, that the AK Party has only once fell below 40 percent, in the 2009 municipal elections, but there were many factors affecting that outcome then. First and foremost, industrial production had fallen by 35 percent and GDP by 15 percent
in the run-up to the elections at that time. That was the aftermath of Lehman. Also, there was a oneoff reaction vote in Sivas. Almost everything could have worked against the incumbent party and helped boost MHP votes then. The second time when AK Party lost votes in a massive vein was in January 2015. Kurdish votes migrated, which in the main explains why the AK Party vote dropped down to around 41.5 percent. Historical data tell us that 40 percent is the minimum lower bound for the AK Party, the ‘granite threshold’, so to speak.
That is so because even ‘economic voting’ pure and simple isn’t that pure and simple. The migration is towards a substitute or ideologically neighboring party, and it is limited to at most 10 percent. From an economic viewpoint, there is also some kind of memory dependence. For an uninterrupted period of 11 years, the three middle Gini 20 percent bands had access to credit, for instance. The slowdown is not that marked compared to the path from scratch to the peak either. The 2040-60 percent Gini slices – the lower- and upper-middle classes - look backward with an ARMA (1, 1) and attach little weight to what may happen in the future. Arguments to the effect that debt has increased by such and such, technological innovation is lacking, investments and privatizations weren’t efficiently conducted or designed etc. don’t strike the ‘median voter’ as important at all. This is of course if and only if there are signs of, an albeit modest, recovery. Hence, I expect a hefty pay increase for government employees and this will spread over from there in January and February.
Mortgages and hous ng
There is a real feeling of uneasiness though. The uneasiness comes from business conditions, FXdenominated debt, currency depreciation, high interest rates etc. Most important of all – probably, although I am not sure - the price, rent and sales figures have been busy nosediving recently in Istanbul and Ankara’s real estate markets. Is this enough to change the mind of the ‘median voter’ from a purely ‘economic voting’ perspective? Well, so long as ‘muddling through’ becomes the main scenario, possibly not. Is it probable? Yes, of course, especially because the outside world in helping. The interest rate outlook has already begun to change globally. From a sectoral viewpoint, we have been facing services-cum-public administration plus industry-based growth over the last couple of years because construction and real estate have been down already, beginning from 2016. From the viewpoint of expenditures, consumption would explain most of it. However, consumption is conditional on credit because earnings are not go up – except administered salary increases that may be due ahead of elections, but the electoral cycle will be short-lived. We are already in the electoral cycle, and that is all there is to that: more public spending, pension increases for retirees, public credits going rampant. Since the argument that has been running in the street to the effect that ‘growth was nowhere felt’ was clearly erroneous, these normal ingredients of an electoral business cycle should suffice. Because 2017 growth mostly benefited profits, wage-earners may not have felt it much but then there are other means to appease larger layers of the population. After all, in 2016 exactly the opposite happened. In 2018 there will be little growth, and definitely nothing or a negative print is on the agenda for H2 2018. We are already halfway through the downturn. Therefore, in 2019, at least in H1, there will be some incentives provided not only to corporates but also to consumers. Fiscal discipline will come after that.
The problem is with the construction-based capital accumulation scheme. To some extent and for some time, this could work and indeed even in advanced countries, real estate has partially replaced land. As one might recall, in the old way of things, land, labor and capital were the main arguments of the production function. However, if one over-stresses one factor, even in the presence of technical progress, the over-invested factor will be subject to diminishing returns. I don’t see how construction and real estate can further push the economy towards a high-growth equilibrium. These days seem to be gone. As such, a temporary mortgage rate cut can only help the sector and those who are intending to buy their first homes in order to live there. Home buying for investment purposes can only be triggered via price expectations. A full-scale real sector restructuring can be attempted, and the MBS issuance aims at that to a certain extent. Therefore, I expect sort of a ‘muddling through’ on that front also. Not a full-scale spending drive, which is virtually impossible, but some kind of help via credits and pay increases might be helpful ahead of elections.
Elect ons
The MHP lost over five percent to the new party IYIP (Good Party), but its voters who had migrated to the AKP in November 1, 2015 elections amidst worries about security and stability were back. IYIP couldn’t penetrate the Central Anatolian strongholds of the MHP - i.e. its traditional voters - but managed to appeal to the Aegean and Mediterranean coastal areas plus Eastern Thrace. There, the MHP voter profile is different. Furthermore, IYIP could attract more than two percent from CHP voters, a normal outcome because in such regions the profile of MHP voters and former center-right voters now voting for CHP over the issue of secularism are quite similar. Finally, IYIP got over 2 percent from AKP and the rest. Clearly, this is an explosive or implosive combination. In local elections, which will anyway be held after the general elections for the first time, IYIP voters might easily migrate back. MHP voters will seal the fate.