[Edited to add: This thought experiment presupposes, of course, that labour is equally organised in all areas - otherwise, labour mobility can be used to undercut labour union power, which is what we're seing in the EU today.

Thanks to Mikhail from SF for pointing this out.]

The problem for labour is that capital and goods will almost always be more mobile than workers: Any infrastructure that is capable of transporting bulk quantities of people will also be capable of transporting bulk quantities of goods, but the inverse is not true (you can pump oil through a pipeline, you cannot do the same with people). The same is largely true for goods vs. capital; since capital can (at least in economies that use fiat money) be translated into information, and since information can be transported with high compression rates compared to goods, capital will often be the most mobile of the triad, and workers the least.

Unions, in this view, work in a twofold manner: First, they negate the advantage that mobility of goods and capital provides to the capitalist by ensuring that the capitalist is met with the same demands everywhere he turns. The union's demands travel at the speed of information, which is the same as, or faster than, the speed at which capital travels. Second, unions negate the disadvantage that the workers suffer due to their relative immobility, by preventing punitive withdrawal of capital: Even if workers everywhere demand the same, the capitalist can undercut those demands by using the mobility advantage of capital to move capital away from certain areas and into other areas (a.k.a. engaging in selective lockouts), thus breaking the resistance of workers in that area and thereby breaking the uniformity of union demands (a kind of inverse strike screw). The unions prevent this by pooling capital in strike funds and similar constructs. Since the strike fund is a pool of capital, it can do anything and everything that a capitalist can do with his pool of capital, thus effectively neutering the ability to engage in selective lockouts.

The principal effects of globalisation have been to increase the ability of goods and capital to move across national borders. Labour unions, however, are still largely operating within traditional national and legislative borders. This gives capital a decisive advantage - for the first time since the rise of organised labour, labour is unable to negate the mobility advantage enjoyed by capital, because it lacks the organisational infrastructure to do so. The result has, predictably, been exploitation, gangster capitalism, tax evasion and a general return to a 19th century balance of power on the labour market.

The logical response to the advantage that global mobility of capital gives the capitalists is for labour to organise globally, which also answers the question of what 'collective' means in collective bargaining: Collective bargaining is bargaining that covers everyone on the labour market. Since the labour market is now global, genuine collective bargaining will have to take place on a global scale. Global collective bargaining will not happen by itself, because it is not in the interests of the trans-national companies, so someone will have to force the trans-nats to accept it. With world government not being anywhere in sight, that someone would have to be organised labour. Globally organised labour.

This raises two obvious questions: What should a global labour union be able to do, and how can we create an organisation with those capabilities. The former is easier to answer than the latter, of course:

- A global labour union should be able to comprehensively target a trans-nat.

One of the features of 21st century capitalism seems to be the increasing consolidation among companies. Partially this takes the form of horizontal consolidation, which aims to control, or at least have a foot in the door of, a given sector of business in multiple countries, or even globally. Partially it takes the form of vertical consolidation, in which a company seeks to establish control of as much as possible of the chain of production - from resource extraction through manufacturing and transport, all the way to the customer's hands in the retail store (and, for certain products like vendor-locked hardware and software, even beyond).

This poses both challenges and opportunities for a global labour union: On the one hand, horizontal integration means that in order to effectively threaten the profits of a trans-nat, labour unions must be able to block production in the same sector in several countries at the same time. Further, vertical integration poses the additional challenge that even blockading a single sector in multiple countries will not be guarenteed to hurt the targeted trans-nat sufficiently to drive it to the negotiating table, since it can largely use the other links in its change of production to diversify its profits for as long as the conflict is ongoing.

Therefore, organised labour must first be able to identify all (or at least a majority) of the trans-nat's income sources, which is in and of itself not that easy, since most trans-nats maintain a deliberately opaque company structure in order to fool national authorities (and/or provide plausible deniability to national authorities that want to be fooled). Having then mapped out the structure of the trans-nat that it wishes to hurt, the union must be able to bring overwhelming pressure to bear at a sufficient number of leverage points to decisively break the trans-nat's will to fight. Keeping a million workers striking for several months is not cheap, so it is desirable to make a sufficiently strong impression that both the trans-nat hit and its fellow trans-nats in general will think twice before provoking an all-out conflict again.

On the other hand, however, vertical integration does provide an interesting opportunity for a less comprehensive way to put pressure on a trans-nat. As outlined above, more and more trans-nats own the product they sell all the way from factory to retail store. Furthermore, due to the increased mobility of goods and the disparity of purchasing power between first-world and third-world countries, the manufacturing frequently takes place on the other side of the world from the retail store in which the end-user purchaces the finished good. And as a commenter on ET remarked, Sun Tzu had a thing or two to say about long supply lines...

Or, to paint an even clearer picture: In a major docking facility in a place like Hamburg or Shanghai, thousands of tons of goods are transshipped every single day. Goods like iPods, which are almost worth their weight in gold. Or goods like cars, which while less valuable are far more urgently needed, because retail stores and wholesalers have smaller stockpiles. And they still belong to the same company that runs the factory where they were made. If a global labour union can identify the weak points of a trans-nat's supply lines and engage in surgical strikes - you should pardon the pun - against those weak points, a couple of thousand striking workers kan cause goods worth several tens of millions of € to be held in an inaccessible limbo for months, directly hitting the bottom line of an exploiting company as surely as if they'd blockaded the offending factory itself.

- A global labour union should be able to comprehensively paralyze an entire industry.

Sometimes, a labour union wants to change conditions less in an individual company than in an industry as a whole. Examples would include health standards for coal mines, work/rest schemes for truck and train drivers, safety regulations for construction workers, etc. In this case, hitting each company in the industry in turn gets rather cumbersome rather fast. If, on the other hand, the labour union is capable of hitting a large number of financially unrelated operations within the same industry, the leverage needed to gain collective agreements covering that industry may be obtained. Consider, if you will, the effect on the global economy if - say - a significant number of steel plants were to cease operation for a month or two. Or if long-haul shipping came to a screeching halt because dockworkers all over the planet simultaneously blockaded the transshipment terminals. Such an action would leave a profound impression, to say the least.

- A global labour union should be able to comprehensively paralyze an entire country.

One of the most powerful weapons in the traditional labour union's arsenal is the general strike: A comprehensive withdrawal of labour from the entire economy, at the same time, potentially lasting for months unless the government intervenes or agreement is reached on whatever points prompted the general strike. Historically, the use of general strikes has been limited due both to the collateral damage to the economy involved in such a move, the fact that governments usually move to quash general strikes after about a week at the outside, and the fact that such a move is extremely costly. Unlike a limited strike or a strike screw, a long-lasting general strike will drain the coffers of the labour unions just as surely as it will hurt their opponents.

The two former considerations are still valid for a global labour union. But the last no longer applies. With the ability to transfer money across national borders with relative ease, a near-comprehensive strike can be kept up essentially indefinitely (near-comprehensive because it must, of course, allow for the distribution of food, medical aid, etc. in order to last for more than a couple of days on the outside). In this fashion, organised labour can apply the strike weapon to entire countries on a scale never before seen. Such a powerful tool should, of course, be used with the utmost caution, but the very fact that the possibility exists is itself a potent bargaining tool.

As an aside, the ability to do Bad Things to entire countries on a scale sufficient to scare their rulers into submission is a privilege already enjoyed by the trans-nats: The annual turnover in your average trans-nat is several times greater than the government budget of a mid-sized EU country, nevermind the third world. So to those who argue that this capability is too dangerous in the hands of a global labour union should carefully examine the regulation of the ways in which the trans-nats can use similar nuclear options. Judging by the behaviour of some trans-nats, I suspect that a survey of such regulation would be a rather brief affair.

- The global union should be able to organise boycotts of targeted companies.

The infrastructure needed to organise a boycott is largely the same as the infrastructure needed to organise a strike, and a boycott can more easily gather support from people who are not affiliated with the mother organisation. This will be important in the West which is increasingly - and rapidly - becoming de-unionised. On the other hand, it would be illusory to believe that boycotts can by themselves replace strikes and blockades. At most they can be used in combination with other activities to turn up the pressure on the targeted companies.

NGO-building

So, how do we go about creating a beast like the one I described above? I don't know exactly. I don't know enough about the relevant law, history or organisational theory and practice to even begin to formulate a comprehensive plan. But I do see several advantages and challenges compared to the last period in history in which labour attempted to organise. Taking the disadvantages first:

- Scale

The sheer scale of a muti-national, let alone global, labour union beggars the imagination. It is, quite simply, organisation on a scale not seen before in human history. Then again, so were the labour unions of the 20th century when they were first formed.

- Nationalism, racism and sectarianism

The labour unions of industrial-revolution Europe had the advantage of being reasonably homogeneous beasts. As a commenter on ET pointed out, a contributing factor to the failure of the labour unions in the United states was the way racism was used by the government and big bizniz to play different sides against the middle. It is to be expected that we will see similar attempts by gangster capitalists, like Rupert Murdoch to take an obvious example, to divide labour along ethnic, religious and national lines. Our opponents are not stupid, and the tribalism card has worked so well for them in the past (and present, for that matter).

- We do not control the flow of money

Part and parcel of the ease with which capital moves around in the globalised economy is the fact that it's largely moved around by banks. Tightly regulated banks. Should the banks en bloc decide that they don't like a global labour union, they can cause all kinds of grief. For that matter, should certain governments (or rather their corporate lords and masters) take a dislike to the idea of a unified global labour movement, it would be simplicity itself to declare it a 'terrorist organisation' - at which point governments all over the world will be all too willing to freeze bank accounts, confiscate funds, prosecute those who facilitate the transport of funds to support strikes abroad and generally be a pain in the ass. A way to compensate for or remove this weakness must be found.

I suspet that a global labour union will initially distribute its funds through various slush funds and quasi-independent networks. Much like the way an illegal partisan organisation would finance its operation. Or, for that matter, similar to the way the trans-nats evade taxes. There is a certain delicious irony to the idea of making a global labour union harder to quash by using the very money-laundering infrastructure set up by the gangster capitalists themselves.

A global labour movement today also has a number of off-setting advantages, however:

- Faster flow of information

When the current labour unions formed, information traveled at the speed of trains and/or horse-drawn carriage for the most part. Information now travels at a perceptible fraction of the speed of light. Once the basic organisation and mobilisation schemes are in place, blockaders and strikers kan be mobilised globally in the space of hours, using cell phones, SMS and the Internet. It would be foolish to think that new media can substitute for feet on the ground: High unionisation levels and a strong cadre on each workplace that a global union wishes to take action against will still be needed. But the cell phone mediated flash mobs used to evade police at e.g. anti-G8 demonstrations offer valuable proof of concept for swift mobilisation schemes.

- We've done it before

A global labour union will have a huge backlog of history and tradition from the European and American labour unions that have been at times successful, and at times been the next best thing to utter failures, throughout the 20th century. Their good ideas can be co-opted and adapted, their failures can be studied (and hopefully avoided) and their successes can inspire and serve as visible proof that change for the better is possible.

- Functioning labour unions exist in the West

European and American labour unions can provide training and advice to allied labour unions in the developing world. Further, they represent a reserve of both capital and cadre that can directly support the operations of their allies abroad. And in some countries they have not entirely insignificant lobbying resources.

- Western economic imperialism

Possibly the most important single advantage a global labour union can derive from the Western cadre is - ironically - the very disparity in purchasing power and wage expectations that was created by and sustains 21st century gangster capitalism. When the first labour unions were formed, all labourers were paid a sufficiently crappy wage to make it very hard to build respectable strike funds and generally limiting the unions' activities. However, due to the very exploitation that a global labour union would be created to combat, a worker in the first world earns at least ten times as much as a worker in the third world. And when you factor in the less-than-cursory relationship between exchange rates and purchasing power w.r.t. subsistence goods, the number of striking workers supportable by even moderate commitments from the war chests of Western labour unions becomes staggering. Again the delicious irony of using the very tools created by the gangster capitalists to break their power.

Of course that's just the list I can think of off the top of my head. I'm sure that my readers can add considerably to both the advantages enjoyed and disadvantages suffered by an attempt to unify the global labour movement.

- Jake