Dani Rodrik's weblog

The battle is renewed: state capitalism, mercantilism, and liberalism

My new Project Syndicate column deals with the age-old battle between the liberal and mercantilist models of capitalism.  While liberalism has won the intellectual fight, the real-world battle is still on, and is likely to shape the future of the world economy.

[I]t is more accurate to think of mercantilism as a different way to organize the relationship between the state and the economy – a vision that holds no less relevance today than it did in the eighteenth century. Mercantilist theorists such as Thomas Mun were in fact strong proponents of capitalism; they just propounded a different model than liberalism.

The liberal model views the state as necessarily predatory and the private sector as inherently rent-seeking. So it advocates a strict separation between the state and private business. Mercantilism, by contrast, offers a corporatist vision in which the state and private business are allies and cooperate in pursuit of common objectives, such as domestic economic growth or national power.

The mercantilist model can be derided as state capitalism or cronyism. But when it works, as it has so often in Asia, the model’s “government-business collaboration” or “pro-business state” quickly garners heavy praise. Lagging economies have not failed to notice that mercantilism can be their friend. Even in Britain, classical liberalism arrived only in the mid-nineteenth century – that is, after the country had become the world’s dominant industrial power.

Coincidentally, my good friends Daron Acemoglu and Jim Robinson pre-empted me by a couple of days with their own Project Syndicate piece on “state capitalism”: 

state capitalism is not about efficient allocation of economic resources, but about maximizing political control over society and the economy. If state managers can grab all productive resources and control access to them, this maximizes control – even if it sacrifices economic efficiency.

As this excerpt makes clear, their take is complementary to (but also partially at odds with) mine.  The two pieces make great companions and should be read together.

Turkish court provides (lack of) reasoning behind Sledgehammer verdict

The Turkish court overseeing Turkey’s Sledgehammer case has finally issued its much awaited reasoning behind the verdict of last September, which found 330 officers guilty of plotting a coup in 2003 against the then-newly installed AKP government.  The fabrication of evidence and the manipulation of the judicial process in the case were so blatant and grotesque that there was considerable curiosity as to how the court would justify its decision.  How would it wiggle out from the impossible position it had placed itself?

The reasoning runs into nearly 1,500 pages, most of it just a copy-paste from the indictments.  The section that deals with the evidence is a spectacular monument to illogic and contains much misrepresentation.  But the biggest whopper comes when the court tries to dismiss the forensic reports that showed the coup documents, allegedly from 2003, contained digital elements from Microsoft Office 2007.

First, a bit of background.  All the incriminating coup documents in the case came on CDs, of unknown provenance, that were apparently written in a single session in 2003.  That is at least what the metadata on the CD indicate.  The metadata were the principal evidence used to charge the officers on duty at the time. Yet when American, German, and Turkish digital forensics experts looked at the documents inside the CDs more closely, they discovered remnants of Microsoft Office 2007 – references to new fonts and schemas that did not exist in the earlier versions of Office. 

Since Office 2007 was first introduced in 2006, three years after the coup documents were allegedly written, there was only one possible conclusion.  The coup documents and the CDs containing them were in fact prepared years later and then backdated on computers whose system clocks had been set to 2003. (Further details about the forensic analysis are here).  We were able to identify additional anachronisms within the documents themselves, which pointed to August 2009 as the earliest date on which the CDs could have been produced.  (The CDs were delivered anonymously to a reporter on January 2010.) 

The defendants pleaded repeatedly with the court to appoint its own experts to confirm or falsify these claims. The court refused.

So what does the reasoned verdict say about the presence of Microsoft Office 2007 in documents from 2003? It makes an astonishing -- and astonishingly inaccurate -- claim, namely that when a document is opened using a later version of Microsoft, the computer “transforms” the document to make it look like it could have been written in the later version .  Essentially, the court argues that the coup plot documents were “updated,” and that Office 2007 elements found their way into them when the original documents were worked on in later years.

Never mind that there is no shred of evidence of updating in the documents except for the occasional anachronism, or that the court says nothing about who may have done the updating.  The problem with the court’s argument is more basic. 

For any change in a document to be retained, the document must be saved after being opened. This goes also for the introduction of any elements from the new version of Office.  Once the document is saved, the updating date becomes the “last save” date.  If this document is subsequently burned onto a CD in a single-session, the “creation” date of the CD similarly corresponds to the updating date.  There is simply no way a CD or document can contain Office 2007 and yet appear to be last saved/created in 2003 unless the clock on the computer on which the operation is performed has been backdated.  Unless, that is, someone manipulated the metadata on the documents and the CDs to implicate the defendants. 

The judges know this full well of course.  That is why they refused to appoint technical experts to scrutinize the matter.  (To compound their dereliction, they write that they obtained their information from Microsoft “public sources.”)  The judges were determined to get wrong what is essentially a simple technical matter.

The case now goes to the Court of Appeals, and then most likely to the Constitutional Court.  It will be years before defendants can go the European Court of Human Rights to get redress on matters of substance.

But the significance of this case goes way beyond the injustice done to the defendants.  This is the most important court case in Turkish political history since the 1960 trials in Yassiada (another show trial, in which the generals tried the government which they had deposed in a coup).  What it has done is expose in its full splendor the monster that the Erdoğan government has helped create within the judiciary and police. 

We now can see that what appeared for some time to be the civilianization and democratization of Turkish politics was in fact the empowerment of a new “deep state,” the tentacles of which now extend into Erdogan’s residence and threaten his trusted intelligence chief.

As I wrote earlier, Erdoğan has now painted himself into a corner out of which there is no easy escape.  The path of least resistance is to let the fraud run its course and stand by as a sham legal process drags on through various stages of appeal.  That, however, is the surest way to undermine his future political legacy.

UPDATE: The Gulenist daily Today's Zaman has published a column attacking me and attempting to defend the verdict. My response is here. The exchange is worth a look as it reveals once again how desperate the Gulenists are to defend the indefensible, and how willing they are to distort the facts of the case to do so.