The border is not where the map says it is: regulation and customs on the Middle Corridor
By Emzar Gvarishvili, Deputy Head of Maritime Digital Services Management Unit at Maritime Transport Agency of Georgia
Every map of the Middle Corridor draws one line from China to Europe. The cargo that uses it crosses seven regulatory environments and restarts the paperwork at each. Seen from the Georgian end, the route’s real competitor is not another route; it is its own unpredictability.
Ask a freight forwarder what worries them about the Middle Corridor and the answer is rarely distance. The route is long, but length can be priced. What cannot be priced is a week that may or may not be lost at a border, and that is what every shipper really weighs against the alternatives.
Seven restarts
Follow a container from China to Constanța and count the jurisdictions. It is cleared for export in China, enters Kazakhstan’s transit regime, boards a ferry across the Caspian, enters Azerbaijan, enters Georgia, crosses the Black Sea or continues by rail through Türkiye, and arrives in Romania and the customs territory of the European Union. Each of those is a separate customs administration with its own forms and its own idea of what deserves a closer look. Twice along the way the rail gauge changes, and every change of gauge is also a change of document.
The container does not travel one corridor. It restarts the process seven times, and each restart is a small lottery. Trains run to timetables; borders do not. A crossing that reliably takes a day can be planned around; the one that takes a day this week and four the next sends cargo elsewhere, and on the Middle Corridor the variance in transit time is made at the borders and in the ports, not on the track between them.
Where the hours go
It is tempting to blame a single agency, and usually wrong. Watch a shipment through a port and the delay has no single owner. The documents required on one side of a border differ from those on the other. The customs check waits for the border check, which waits for the terminal, because nobody designed the three to run at the same time. Several authorities each hold a mandate over one slice of the process, each applies its rules with some latitude, and none is responsible for the total. And at every stop the same information about the same container is typed in again, because the last stop’s system does not talk to this one.
Only the last of those is a software problem. The others are matters of rules, sequence and coordination between institutions, which is worth saying plainly, because the fashionable answer to corridor friction is a platform. A platform will take the retyping out of a port. It will not make two agencies share a queue, and it will not release a guarantee faster. Digitalization treats the visible symptom; the cure for the rest is duller and more political.
The treaties are already signed
Nor is the problem a shortage of agreements. The WTO’s Trade Facilitation Agreement commits its members to pre-arrival processing, risk-based inspection and single windows. The IMO’s FAL Convention standardizes a ship’s arrival and departure formalities and, since January 2024, requires port states to accept them through a single electronic window. TIR and its electronic successor eTIR give road transit one guarantee and one document across borders, and the CIM/SMGS consignment note does the same for rail across both gauges. The instruments exist; they are applied unevenly, there is no single transit regime from China to the EU border, and there is no shared data layer, so what one administration already knows about a container, the next one asks for again.
The view from the gate
At a Georgian port, seven parties touch a shipment: the ship’s agent, the maritime administration, the border police, customs, the terminal operator, the ring of off-dock terminals outside the port where containers are stored, cleared and dispatched, and finally the railway and the road hauliers. None of them is slow on its own. The shipment simply carries the sum of their times, and nobody owns that sum. Before any new system is bought, the first saving is in sequence: which controls could run in parallel, and which data could be reused instead of requested again.
On the ship side, Georgia has done that. Since October 2024 the agent submits a port call’s information once, through the National Maritime Single Window, and every authority takes what it needs; the repetition is gone, and so is the counter. The cargo side is where the work now is, because for a container the border is not the coastline. It is the gate of the terminal where the last document is checked, and the discipline that now governs the ship’s paperwork has to be carried through that gate to the yard, the railway and the road.
Delay is a design choice
Waiting at a border is rarely an accident. It appears where three things overlap: paper, which lets the same information be asked for twice; discretion, which makes the answer depend on who is at the counter; and face-to-face contact, which turns a rule into a conversation. Remove one and the waiting shrinks; remove all three and most of it disappears.
Two further sources of delay have nothing to do with paper. Where lanes, yards or scanners are missing, the queue forms before anyone has looked at a document. And where two railways and two customs services share one line but keep their own working hours and timetables, a train that arrives after the other side has gone home waits until morning. The remedies are unglamorous and well known: a single submission instead of paper; risk-based selection, so that the system rather than the officer decides what is inspected; electronic payment and remote release; published clearance times; capacity where the queue forms; and border hours agreed at both ends.
What this means for the corridor, and for Georgia
Three things follow. Friction was designed, so it can be redesigned; procedures, sequences and opening hours are choices. Digitalization without harmonization only moves the queue: a national system speeds up one node, and the container still stops at the next border. And shippers pay for predictability before speed, which is why the corridor’s real competitor is not another route but its own variance.
For Georgia, the task now runs from the ship to the cargo and out through the terminal gate. For the corridor, the lesson is that no country can fix this alone: a container still arrives at the next border carrying the next country’s rules, and the route will be judged by its worst crossing, not its best port.
