Singh begins by laying out the commonly believed tropes of military coups:

1. Coups as elections
2. Coups as battles
3. Coups as coordination games

In "coups as elections", the model follows that coups are effectively a referendum on the popularity of government, if a government is deeply unliked, coups are an effective way to oust those governments without the ballot box. The most recent example of would be the Arab spring or Iran's Green movement, where a popular, public driven revolution should lead to a change in government.

"Coups as battles" states that coups are inherently a military affair, and the party with greater military preponderance is crowned the winner. Thailand or South Korea's military coups would be an example of these, wherein a force with overwhelming military power takes over the government. This model does not accurately capture all coups, however.

"Coups as coordination games" is when the better coordinated party emerges as the winner. Singh notes that the antecedent to this involves that in the event of a military coup:

• Military forces will engage in aggressive amounts of fence-sitting to avoid the armed forces from being drawn into a fratricidal conflict (possibly leading to a civil war)
• The general public is unable to stop a coup, but may help it succeed.

This, of course, rests on the notion that a civil war is deemed to be too costly for the majority of the armed forces, and that the general public is themselves not armed.

Singh supports his hypothesis by walking the reader through Ghana over the course of 5(!) coups in 20 years. Ghana, a nation of 11 million with a largely unarmed public, experienced coups that succeed largely because the incumbent government was uncoordinated and expected the military and public to mobilize in support of the incumbent. These coups succeeded because the bulk of the military refused to engage, preferring to wait out the coup to see the winner, and the messaging failed to engage the general public (if there was messaging at all). Singh specifically cites the importance of seizing the radio stations, as well as military assets that could be used as a show of force (i.e. the coup plotters flew an unarmed military jet over Accra during the 1979 coup in Ghana). While this may seem to lend credence to the idea that coups are elections, since elections are largely driven by messaging as well, Singh notes that the purpose of these messages was not to stir the public into a frenzy, but rather present to the public (and other military members) that the coup's success was fait accompli. Once the notion of the coup's success spreads, it becomes a self-fulling prophecy, thus guaranteeing the coup's success.

The coups in Ghana provide an interesting real life example of the Byzantine two generals as well. During the 1979 coup in Ghana, Singh sites multiple cases in which Rawlings managed to coerce the security forces into coordination by letting those who supported the coup make public displays first. By conjuring the notion that the military services are all aligned, the bulk of the security forces that were engaged in fence-sitting start start to believe that the coup is bound for success, thereby ensuring the success of the coup. Rawlings had effectively bypassed the two generals problem by not forcing all the generals to commit at once, rather, by starting off with generals who supported the coup, the other generals were more inclined to follow.

It's important to note that this idea relies on significant amounts of information asymmetry, the general public must not see the coup plotters as weak, and the incumbent government (as well as the coup plotters) must believe that the coup has a chance of being successful. Without coordination on these fronts, coups are hard pressed to succeed, Singh argues, and provides the Ghana coup in 1967 as an example.

The role of the public is highlighted during the USSR coup in 1991, where Soviet journalists framed the coup plotters as weak and uncoordinated, completely against the reality that the coup plotters had the entire Soviet apparatus under their control. The information asymmetry was again pertinent here, as Boris Yeltsin's "courageous stand" in the Russian Parliament building was beamed to every major news channel in the world, and Yeltsin's famous image of standing on a tank played on every Russian television. This was compounded by the coup plotter's inability to move in lock step with each other, the coup plotters originally believed that Gorbachev would acceded to the coup once arrested, and when Gorbachev refused, the coup plotters had no contingency plan to control the media. The confusion in the ranks only grew when Air Marshall Shaposhnikov broke ranks and warned that he would bomb the Kremlin if the coup plotters moved forward with their plan to attack Yeltsin in the Parliament building. Even though Shaposhnikov was bluffing (he only ordered two unarmed bombers to fly over the Kremlin), the coup plotters had no coordination in forcing Yeltsin's hand, resulting in their eventual collapse.

The most clear conclusions one can draw from Seizing Power are:

1. Coups are not steady state, the longer a coup is drawn out, the better the chances are for the incumbent
2. General public, when roused, may help the coup succeed, but are unlikely to stop a coup
3. The bulk of the armed forces, unless drawn on deep sectarian lines (and even then), will refuse to engage until a single side is presumed to be the winner

Viewed from these lens, the failure of many Arab spring revolutions (as well as the Green movement in Iran) becomes clear. Coordination among citizens were at an all time high with social media, but social media also dispelled the information asymmetry that coups so desperately rely on. Furthermore, without a clear nucleus to coordinate, as in the case of many Arab spring revolutions, the coups are prone to fail. The refusal to engage until a winner is presumed among the military is displayed as well, with the armed forces refusing to violently put down the protests until it was clear the protests would fail. Amusingly enough, these points all validate the basic guidelines of success that Edward Luttwak laid out in Coup d'État: A Practical Handbook.

I'm struck by the nature of the case studies in here. Only two countries are examined, Ghana and the USSR, and Ghana in far more detail. Perhaps this is because Ghana's revolutions are far less well known, making their records less likely to be tainted by the court of history, but one has to wonder how correlated coups are in Ghana, given they happen so often, and whether the nature of coups in Ghana are simply peculiarities. Turkey seems to be a curious omission from the book as well, given Turkey's rich nature of military coups.

The primary accounts in this book are fascinating, but I would hesitate to draw insight from them. Since coups are quickly airbrushed by history after their success or failure, primary accounts may be useful for determining what happened during a coup, but seem to be far too biased to accurately draw inference.

My last quibble is with Singh's overtly broad interpretation of "coordination games". To Singh, coordination games are binary, either they "succeed" or "fail". Yet, this ignores all the nuances of coordination games, particularly with parallels that can be draw to Byzantine fault tolerance. I would be extremely curious to know if there is a "quorum" floor for the success of coups with regards to the coordination games.

By no means do these quibbles make Seizing Power a bad book, however, Seizing Power provides insight into a field that requires far more statistical analysis applied to it, and it's a delightful read for those interested in coup tactics. I hope that in the future there are more books like Seizing Power, covering coups across different regions.