It is clear that the possibility to go on strike is an effective power asset for employees. However, it is also clear that every strike that is carried out is a lose-lose: the employer loses money immediately, the company gets in a worse economic position which will presumably reduce also its ability to pay wages, and the labour union loses money on paying the employees without any productive activity coming out of it.

So it would seem in everyone's good interest to negotiate in such a way that minimises the number of strikes – those should only be an extreme action, reserved for the odd case every couple of years when the negotiators fail their job to come to an agreement in time. (Which should be especially seldom in the case of big companies, where the negotiators are professionals specialized at estimating how much the opposite party can be made to budge.)

Yet, at least in central and northern Europe, major strikes are a very regular phenomenon. In Germany, the transport sector alone upsets the whole country (yet another lose!) almost annually by going on strike.

Why? What's going wrong there?

  • $\begingroup$ "Yet, at least in central and northern Europe, major strikes are a very regular phenomenon." Can you please support this claim with a reference? $\endgroup$
    – Giskard
    Mar 28 at 6:43
  • $\begingroup$ @Giskard hm, it's easy to find anecdotal references, but I can't find solid, up to date statistics right now. Wikipedia quotes a table that lists five countries with more than 100 days of strike over a period of 5 years for the average worker, which I daresay confirms the premise. It's rather old though (1996-2000). $\endgroup$ Mar 28 at 8:17
  • $\begingroup$ Hi! This table does not measure days of strikes per worker, but "average number of days not worked for every 1000 employees", so the average day not worked (due to strike action) per worker is less than 0.3 days over 4 (5?) years in all these countries, which are the countries with most strikes 1996-2000 (according to the Wikipedia entry). This seems to refute your premise, depending on your definition of the word "often". $\endgroup$
    – Giskard
    Mar 28 at 8:31
  • $\begingroup$ Ah, well that certainly paints a very different picture if, this is how the table is to be read! It definitely doesn't match up with my experience though. Perhaps the question should be refined to “why are strikes so common in the transport sector of central European countries”. But ideally, an answer would pinpoint where the premise holds, where not, and why. $\endgroup$ Mar 28 at 9:14
  • $\begingroup$ IMO it would be nice if you were to edit your question to reflect these; i.e. what is it exactly that you expect from an answer and what your assumptions are. $\endgroup$
    – Giskard
    Mar 28 at 10:17

2 Answers 2


Strikes last a few days usually or weeks - indicatively per https://www.euronews.com/next/2023/03/07/industrial-action-in-france-and-the-uk-which-countries-have-the-most-strikes-in-europe

the highest national average per 1000 workers in certain European countries is something between $50$ and $150$ days, meaning less than a day per worker annually. On the other hand, a wage increase tends in most cases to stay.

To see that purely income considerations makes a strike really likely, let's consider a greatly simplified exercise. First, let's think in terms of weeks.

Let $W_0$ be the weekly wage without a strike and $E(W_s)$ be the expected weekly wage that the employer is expected to concede after a strike. Let $\beta <1 $ be the gross discount factor but the weekly one, which is a critical aspect of the situation. The discounted stream of income without a strike is $$I_0 = \sum_{t=0}^{\infty} \beta^t W_0 = W_0\cdot (1 + \beta + \beta^2 + ...) = \frac{1}{1-\beta}W_0 \tag{1}$$

while the discounted stream of income given a strike that will last $S$ weeks (so $S$ weeks without income), is $$I_s = \sum_{t={S+1}}^{\infty} \beta^t E(W_s) = E(W_s)\cdot \big[\frac{1}{1-\beta} - (1 + \beta + \beta^2 +...+ \beta^S)\big]$$

$$=E(W_s) \cdot \left[\frac{1}{1-\beta} - \frac{1-\beta^{S+1}}{1-\beta}\right] = \frac{\beta^{S+1}}{1-\beta}E(W_s) \tag{2} $$

In terms of discounted income only, a strike gives more if

$$I_s > I_0 \implies \frac{\beta^{S+1}}{1-\beta}E(W_s) > \frac{1}{1-\beta}W_0 \implies E(W_s) > \frac 1 {\beta^{S+1}} W_0 \tag{3}$$

Now comes the crucial aspect that $\beta$ is a weekly discount factor, so it will be very close to unity. For example to be consistent with a yearly $\beta_{yearly} = .9$ (which is considered a high degree of discounting the future, even for individuals), we must have $\beta = 0.9979758875215$ looking at $52$ weeks.

Suppose the strike lasts 1 week so $S=1$. Then the strike is beneficial for the workers if

$$E(W_s) > \frac 1 {(0.9979758875215)^{2}} W_0 = 1.00406054930652 \cdot W_0 \tag{4}$$

This means that even if the expected increase in the wage is just $0.5 \%$ (namely half a percentage point), the strike is beneficial - income wise. Note that this would be also the overall increase in the yearly income (this percentage does not compound week-by-week).

Considering striking for two weeks $S=2$, still does not raise this wage-increase-threshold above $1\%$. One needs to consider a strike that will last a month ($S=4$) to obtain that one needs to anticipate at least a $1\%$ increase in the wage for the strike to be beneficial in discounted income terms.

Of course, things in real life are much more complicated -for example, the above exercise implicitly assumes zero probability of being fired/losing employment.

Even if we changed the "infinite" horizon to a two-year one, then we would get that we need an expected wage increase after a one-week strike of just ~$2\%$.

The point I want to stress is that, purely income-wise, the situation is inherently "in favor" of going on strike, because even small wage increases make it worthwhile, which also indicates why employers resist even small wage increases (either profit-seeking private firms or budget-constrained public sector organizations).



(for $\beta < 1$) $$A \equiv 1+ \beta + \beta^2 + \dots +\beta^{S} \tag{a}$$

Take $(1/\beta)$ as a common factor in the right-hand-side,

$$A = \frac 1{\beta} \left(\beta + \beta^2 + \dots +\beta^{S+1}\right) \implies \beta A = \beta + \beta^2 + \dots +\beta^{S+1}. \tag{b}$$

Subtract $(b)$ from $(a)$ $$A - \beta A = 1+ \beta + \beta^2 + \dots +\beta^{S} - (\beta + \beta^2 + \dots +\beta^{S+1})$$ $$\implies (1-\beta)A = 1 - \beta^{S+1} \implies A = \frac{1 - \beta^{S+1}}{(1-\beta)}$$

  • $\begingroup$ This would be a better answer if you compared the difference between the increase with a strike and without. $\endgroup$
    – H2ONaCl
    Aug 26 at 13:18
  • $\begingroup$ @H2ONaCl $W_0$ can be readily considered as the counter-offer by the employer before the strike. Say current wage is $W_c$ and $W_0 = 1.03\cdot W_c$. The moral will not change because the percentage increases are small in real world situations. $\endgroup$ Aug 26 at 17:28
  • $\begingroup$ "average per 1000 workers in certain European countries is something between 50 and 150 days, meaning half to one-and-a-half days per worker" Seems like the lower bound should be one-twentieth of day per worker. $\endgroup$
    – Giskard
    Aug 26 at 18:38
  • 1
    $\begingroup$ @Giskard Thanks for the correction. On your other comment, my answer provided a quantitative assessment of why the financial motives tend to favor a strike. This is per my long-standing motto "when in doubt, do the math". $\endgroup$ Aug 27 at 12:22
  • 1
    $\begingroup$ Ps. In your own answer, you write that the national average of striked days per workers per year is less than 0.15 in most countries, but you end your answer with 'the situation is inherently "in favor" of going on strike'. Do you not feel that something is missing here, do you really feel that you have answered the question? $\endgroup$
    – Giskard
    Aug 27 at 17:05

Your lose-lose argument only holds in the short-term and the issue is a longer term (dynamic) problem over time.

For both the employees and employers, it can be beneficial to accept a short-term loss, if it translates into a big gain in the long-term (by getting higher wages for workers and the opposite for employers).

Since wages tend to be sticky and not go down, the sum of all long term effect of the entire future (up to the firm’s or worker’s life span) can often easily outweigh the cost of a short-term strike.

Lastly, because of inflation, wages need to be negotiated regularly. So the more often negotiations are needed, the more often you have the potential for strikes.

  • 3
    $\begingroup$ This answer explains why wage negotiations are frequent, but not why such negotiations "frequently" result in strikes, instead of an instant agreement a la Rubinstein. $\endgroup$
    – Giskard
    Mar 28 at 6:45
  • 1
    $\begingroup$ “if it translates into a big gain in the long-term” – that's exactly my point: in the long term, money spent on a strike is just gone. If the employer (E) is so closefisted that the union has reason to strike, E loses a lot of money on that strike and still more later after having to raise the wages to end the strike. If the union (U) demands wages that E really can't affort, U loses money that the workers will perhaps get back in the short term because the company is forced to overspend, but on the cost of making the company unprofitable in the long run, which costs U/W even more money. $\endgroup$ Mar 28 at 8:02
  • $\begingroup$ What makes you think they have to raise wages to end the strike? Many strikes are unsuccessful? A company can lose 6 months worth of revenue on a strike, or raise wages and have higher wage costs for the next 100 years. Often, the cost of 6 months of revenue will be smaller than 100 years of wages, so you roll the dice and try to win. $\endgroup$
    – BB King
    Mar 28 at 15:00
  • $\begingroup$ 100 years is silly – but, sure, the company could just refuse to move. It doesn't change the point though: in this case it's the union that is confronted with exactly the same position as they were before the strike, so if it was justified for them to strike in the first place, then the consistent decision would be to not give in but just keep on striking. Of course this can go on only so long – at some point their choice will be between stopping the strike and driving the company into bankruptcy (or to move abroad or whatever). But how is that useful for anyone involved? $\endgroup$ Mar 29 at 8:14
  • $\begingroup$ Strikes very often don’t work, so workers very often do not keep on striking practice. Companies that do not give in to strikes don’t always go bankrupt and they profit if they can pay their workers more. So the “choice at some point “ you mention does not really exist always. $\endgroup$
    – BB King
    Mar 29 at 12:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.