When might Space Fighters be practical? Note: Majority of arguments below are based on a realistic hard scifi setting. In softer settings you can probably invent any Applied Phlebotinum or Minovsky Physics needed to support or refute the plausibility of starfighters. Also, per Standard Sci-Fi Fleet, we are using "space fighters"/"starfighters"/"strikecraft" as a shorthand for all combat-capable Small Craft. There is no need to wrangle over the differences between bombers, fighters and other subtypes.
A1. Glass Cannons. Glass Cannons everywhere.While an inductive argument is not foolproof, there is some evidence in military history that defense will often lag behind offense. Armor Is Useless, in other words. Look at, say, how infantry armor was abandoned for quite a while due to the impracticality of the thickness needed to protect against advanced guns, or how modern carriers need to use active defenses to intercept incoming missiles rather than being able to just weather them. Extending from this, future space combat scenarios may involve spacecraft firing at each other with weapons they cannot survive. If relatively small starfighter weapons can continue to lay the hurt on capital craft, it may be more practical to let relatively expendable strikecraft sortie than risk capital spacecraft whose loss will cost heavily in money and manpower. There have been times throughout history when defense did outpace offense, however, at least for a while. Arguments B4 and B6, below, address potential problems with this idea.
A2. The Tech Level is low, and orbital combat is a top priority.Quite simply, space fighters are easier and cheaper to build than large ships. If the setting has a tech level close to what we currently have in Real Life, building a Standard Sci-Fi Fleet of capital ships may simply be impossible, or at least prohibitively difficult and expensive, but small, single-person spacecraft could be realistic enough. In such a setting, fighters would be launched directly from a planet, and combat would take place in orbit or otherwise in nearby space. Space fighters may be the most practical way — or the only practical way — to get any combat capability into space at all in such circumstances. This is the alternative that has come closest to becoming Truth in Television so far. On the other hand, if the tech level is a bit higher — high enough to allow for a Space Elevator or other comparatively low-cost means of getting into orbit, for example, and to allow manufacturing and resource gathering to take place directly in space — then the construction of larger ships becomes much more feasible. In a setting with such a tech level, it is possible for one large warship to be more economical than an equivalent squadron of fighters.
A3. Weapons are projectile-based and slow/ungainly.If the dominant weapons are solid projectiles, interceptors would be useful as a complement to the point defense screen, thinning out incoming fire so the PD on capitals would have an easier job. Obviously, you can't do the same with energy weapons. This would also be contingent on the projectiles not being too fast, else the interceptors would have difficulty running them down. An easy way to make this happen is to create something like a smoke screen yourself. By firing a specially designed projectile which will be inflated and/or spread out small particles that are magnetically attracted to a centre zone, one can form an area that can block incoming DEW for a short while.(similar to whats being used in Starship Operators) Smaller ships(i.e. Space figters) can use smaller versions to shorten the distance and use solid projectile weapons, or simply move to another angle to attack. See B5 for the opposite.
A4. The manoeuvrability of space fighters versus larger ships makes them worthwhile.Space is not an ocean, but space fighters would still be more manoeuvrable than larger spacecraft thanks to the Square/Cube Law. The larger a spacecraft is, the harder it is for its structure to handle the stress of rapid acceleration during maneuvers — see this web-page for more on this. In this case, rather than being a space version of an aircraft fighter, the space fighter would be more analagous to a PT boat or other "fast attack craft". The main question is whether this added manoeuvrability would be enough of an advantage to make space fighters sensible — missiles and robotic drones would have the same strength, after all, and likely even more when you take out the mass of the pilot or life-support gear. Still, some critics of space fighters do occasionally argue that larger ships have no disadvantage at all when compared to space fighters in a realistic setting, and this is one plausible counter-argument.
A5. There is significant electronic warfare in the setting.Retaining human pilots for Space Fighters can provide survivability in works with significant electronic and information warfare — a robotic drone is of little use if the enemy can control it with a software hack, and missiles can be led off-course with decoys of one sort or another. Human pilots may prove more resilient against such threats. There is some real-life precedent for this; notably, there have been actual cases where militaries have claimed the ability to bring down sophisticated remote-controlled and AI-equipped spy drones with nothing but radio signals and hacking. Of course, space is the sort of environment where even a human pilot will likely need to rely on computers and electronic equipment to a significant extent — but still, keeping a human in the loop may give a significant advantage. Another, related problem with drones (and possibly, though not necessarily, missiles) could be the absence of an AI capable of doing the job of a human pilot — see point B3 below for more on this.
A6. Cost.Moving and utilizing a large ship is expensive. All the fuel spent is not free in a realistic setting. If sending in a smaller ship or a set of smaller ships can do the job, real life militaries will not send the big guns. A fleet with a carrier can indeed send out a few space fighters to perform smaller tasks rather than sending out the smallest possible ship like a frigate. The smaller ships may not be able to survive a long journey by themselves, but the mother ship can. This is particularly the case if there are any FTL systems that requires really big ships to perform, so the fighters cannot travel at FTL themselves, but require a larger ship to bring them to a closer range where they can then perform a small task with their mobility where the big ship cannot. B7 arguments are only true when space fighters are used against large ships, but this is not always the case in real life. You have smaller or weaker targets, like transport or civilian ships, patrol ships if there are pirate behaviour, and as long as you need to take care of those, you always have some sort of space fighters that are easy and cheap to launch, and the fighters that can fight against this kind of small enemy ships. Another thing about cost, is economy. You cannot support a fleet that is above your economical capabilities - if you only have an income of X, you cannot support a fleet that spends 2X. In a fictional or ideal world, yes, the bigger the better, build the largest ship your technology can give you, and build zillions of it, and you have the strongest fleet. In the real world, physics is not the only limiting factor to the ships - your political concerns, your enemy, your economic strength, your ability to gather troops, etc. are also limiting factors.
A7. Because there are planets.If you enter the gravity well of a planet, it is hard to get out with a big ship or the big ship simply doesn't have the power to fly in the atmosphere, but the small fighters can enter it and be back. Arguably not space fighters, but if the enemy intercepts in space, than these need to have space combat ability as well. Also arguably that the big ship can perform orbital bombardments, but it will be hard to aim a few hundred km away, with clouds and with orbital defence shooting at you while you are zooming through your target at 7.8km/s.
A8. Stealth.There is no stealth in space, for large and long operating ships. But the same may not apply to smaller and short operating vessels. If a vessel operates with cold propellant, have an onboard system that cools down its exterior and traps the heat inside for only a few hours, it will be virtually stealthy, or at least much more stealthy than any big ships that inevitably heats up faster.
A9. If large ships became man operable with small crews.You cannot really distinguish a space fighter and a space frigate if it is operational with a 2 man crew, or even less than 10 like the heavy bombers of WW2, while having much larger ships in your fleet.
A10. When you are cheap or poor and space is huge, 1 single capital ship is rather useless in defending all your bases.No matter how big your enemy's ship is, there is almost always a slightly smaller design that can still damage it, but is a cheaper alternative when you don't have enough money to build and constantly send out such a large ship. If against a similar enemy, both will then build smaller ships to increase strategical mobility of the ships for better fleet management. At a certain point, only small ships can be mobile enough for your needs, and space fighters might be able to damage them, so sending out skirmishing fighters may become a good way to stall your enemy, and protection against this kind of enemy is necessary.
A11. If you don't have Deflector Shields.A1 will still be valid countering B4 as long as no force field technology is developed. Physics tells us that no matter how thick your armour is, each attack is still going to chip away the armour bit by bit. Also, Hyper Velocity Impact tests(http://ares.jsc.nasa.gov/ares/hvit/hit.cfm) shows that a thick armour is not an ideal defense against KEW at really high speeds. Countering B4, a capital ship with too much armour (i.e. mass) will be a sitting duck under super long range sieges that are slower but are also massive.(e.g. asteroid bombs) since it is hard to turn and change course. When might Space Fighters NOT be practical?
B1. Reliable point defenses exist.If the universe has PD that can mow down far more aggressively-manoeuvring missiles like cavalry before Gatlings, only blatant Plot Armor can protect sluggish strikecraft from getting torn to pieces even more easily. If there is no Stealth in Space, and if combat takes place over large distances, it could give the defenders plenty of time to detect incoming fighters and try to take them out from afar. A9 above might be a way through this.
B2. Long-range missiles are a viable alternative.In real life, long-range missiles are an increasingly important part of warfare; the same may be true in space. Instead of fighters, large spacecraft could simply launch robotic missiles at each other from great range. These would have a few advantages over fighters. For starters, a missile doesn't need to make a return trip (or indeed decelerate relative to its target at all), which means it can either carry much less fuel (making it smaller and lighter) or it can carry the same amount of fuel, but use it for manoeuvres that a fighter could not afford to make. The missile could also accelerate more rapidly, both for this reason and because it wouldn't carry a pilot that could lose consciousness from excessive G-forces. All this can combine to make the missile harder for point defenses to hit — it could give the enemy less time to react as it approaches, evade point-defense fire more effectively, and present a smaller target. Unlike an Attack Drone (see below), a missile would not necessarily need advanced AI or remote control. It would simply have to track a target, accelerate towards it, and perhaps make some randomized evasive manoeuvres to try and dodge point-defense fire. A missile could also be cheaper than a fighter or an Attack Drone, meaning that more could be deployed — also making the job harder for point-defense systems. It may still be much easier for the enemy to shoot down missiles in space than it would be on Earth — greater distances mean more warning and more time to react, and no horizon or real limit on the range of point defense weapons means more chances to take the missile out. However, depending on the way Space Fighters would be used in the setting, they may suffer from the same weaknesses to an even greater extent (see above) — their one advantage would be if they could engage the enemy from a range great enough to make dodging defensive fire possible, while missiles obviously would not have this option. One might be tempted to say that there's no rule that says space fighters can't simply carry missiles. Today, most airborne missiles are carried by fighters (some of which are in turn launched from carriers). Depending on how technology advances, this approach might work in space, as well. Then again, defenses may be such that fighters and fighter-launched missiles don't have the punch to damage capitals; see B4 below. It's also possible that the same lack-of-horizon that will make missiles easier to shoot down in space will also reduce the need to have fighters carry missiles into action, instead of having them launch directly from larger ships — and a missile launched from a fighter is effectively the same as a multistage missile, with the key difference that the "first stage" is manned in the former case.
B3. Reliable and inexpensive drones exist.This depends on a variety of circumstances, including the range at which space combat in the universe takes place, the quality of AI available in the universe, and whether faster-than-light communication is possible, among other factors. Unmanned robotic fighters would need either decent AI or some means of remote control, and the possibility of the latter depends on either combat taking place at fairly close range, or the availability of a Subspace Ansible. However, some combination of AI and remote control could be practical — remote control (with a light-speed delay) could instruct the fighter as to its overall goals and priorities, and an on-board AI would handle moment-to-moment decisions that depend on "reflexes" and adapting to quickly changing circumstances. If these problems could be solved, drones would have several advantages over fighters, mainly not needing to carry a pilot. If computing technology is miniaturized enough, the pilot and life support could be replaced with a much lighter computer system — and in space, every gram of mass counts for fuel use and manoeuvrability; also, as with missiles, there would be no need to worry about a pilot blacking out from G-forces. Of course, no pilot also means not putting human lives at risk, and possibly faster reaction times. The main unknown variable here is the weight and cost of the pilot-replacing technology. A future can be imagined in which it is cheaper to train a pilot and build a piloted craft than it is to install fighters with advanced AI (indeed, several sci-fi writers, including Isaac Asimov, have imagined such scenarios). Some evidence for both sides of the debate can be taken from real life. On the one hand, human labor still turns out to be cheaper than automation in many applications, including industrial tasks that could be automated with present-day technology — using a human being instead of a machine is by no means always the more expensive option. On the other hand, looking at present-day space technology, robotic craft have certainly turned out to be far cheaper and more practical than manned vessels for a variety of scientific and commercial applications — though admittedly, there's no way to know whether this would still apply in the chaotic environment of combat. Then there's the fact that computers are still getting smaller and more powerful every year — but then again, many experts already foresee the end of Moore's Law and rapid advancement in computer power in the not-too-distant future. Looking at present-day military forces, robots and drones are certainly being used more and more — but so far it looks like they will work alongside human fighters, rather than replace them outright. The same balanced approach may prove viable in space combat. As for the Subspace Ansible, it is, of course, totally impossible to know how much such a hypothetical device would weigh or cost. Another potential problem with drone fighters is that they may fall prey to information warfare — hacking, fake radio signals, and the like. Piloted fighters would still likely need lots of computer equipment, and so may not be immune to information attacks either — but they may still be less susceptible than a fully robotic drone. See point A5 above.
B4. Highly resilient capital craft.Modern fighters and bombers are a threat to wet navy capital ships and land fortifications because they can carry weapons that effectively damage them. However, even in space where they do not have to fight against gravity to launch, engineering limits would prevent a relatively small fighter from carrying too big a weapon. As a counterpoint to A1 though, you can probably imagine a world where torpedoes and air-launched munitions never gained the punch to be useful against battleships or land fortifications, leaving battleship cannon as the kings of the battlefield. Alternately, future defenses such as Deflector Shields might scale with size such that even the heaviest strikecraft weapons barely damage capital craft, or not even that. Either way, strikecraft would be effectively useless in the offensive role. While a larger projectile would have more mass, it would also have more space for propulsion, fuel, payload, all the useful jazz. A good existing example of this is the Honor Harrington series, where until recently the aversion of Armor Is Useless meant that subcapital weapons lacked the punch to usefully damage (super)dreadnoughts. A12 is a reason to counter this.
B5. Weapons are energy-based or fast and agile.You can shoot down a projectile. You cannot do the same for directed energy; if it's already on-target, one can only attempt to absorb or deflect it. Sending interceptors out to aid PD would be worthless in this case. Furthermore, if missiles are sufficiently agile and fast that strikecraft cannot easily catch them either - reference the currently-in-development BRAHMOS II that does over Mach 6, impossible for any current Western air-launched missile, much less the carrying fighter, to catch in a tail-on chase - that would also eliminate their defensive role as a supplement to the point defense screen.
B6. Everything Space Fighters can do something else can do better.While there are advantages as well as disadvantages to space fighters when directly compared to larger ships, a good look at the concept from the very base upwards is necessary. The first question shouldn't be "What advantage does a fighter have over a big ship?" but "What can a space fighter do?". Because we're talking about military ships here, the answer is generally to bring some sort of weapon payload (bullets, lasers, blaster bolts, missiles, bombs) in contact with a target. But the conditions of combat in space make fighters pointless for that. On planet, fighters are needed to extend the range of whatever deploys them (an airforce base or a carrier). If the base were to shoot the guns or the missiles that a fighter carries directly, it wouldn't have nearly the range that a fighter can achieve. The horizon on planet prevents direct targeting beyond a limited range. The friction of the air slows down bullets and missiles so they drop to the ground short of the target when they have been slowed down enough or their fuel has run out respectively. The engines and shape of an fighter allow far more efficient travel in atmosphere than those of a missile (or bomb or bullet). Not so in space. There is no horizon, so everything can be targeted directly. There is no friction, so ranges are not limited. There is no need for aerodynamic design, so missiles are far more effective than fighters. For comparison: if one were to use a missile that is the same size as the fighter i.e. using the same engine and same amount of fuel, it would have four times the range of a fighter, because the fighters needs a lot of fuel to brake and return to base again (and this is before you take into account the fact that using a missile instead of a fighter also frees up space that would be otherwise taken by the pilot and whatever equipment he needs to both stay alive and control his craft). So, unlike in an atmosphere, where mounting missiles on a fighter extends the effective range of the warheads, in space it would seriously limit it. As for guns, those are even less effective. Unless there is some sort of magical technology at play that makes 5 tons of gun components, propellant and bullets somehow capable of more destruction than just 5 tons of warhead (not the case with real physics) then carrying a small gun close to a target to shoot it is a colossal waste of time. Targeting is another thing that potentially looks like a reason for fighters to exist. But it is again not the case. Getting closer to the target does exactly the same thing as using a bigger lens (because there is no horizon) so the bigger lens wins. (does not get closer to danger, doesn't need refuelling, etc.) Intercepting incoming missiles works pretty much the same as launching attacking missiles, and attaching a space fighter makes it worse, not better. For that matter, anything that can destroy an incoming missile will probably be just as effective against a fighter, too. In the end, while one can point out plenty of advantages that a space fighter has over a larger ship (in a universe with real physics), there just is no task that a space fighter is best suited to perform. Either a bigger ship will outperform several small fighters, or one or several missiles will outperform one fighter.
B7 Cost is not an issueThis comes in two flavors. The first being the high-end SF post-scarcity society where construction resources are not an issue. Assuming some other kind of limiting factor (without at least one, offensive war itself makes no sense) that will be what determines ship sizes. For example if resources are infinite but the number of pilots is limited, ships will be designed in a way to capitalize on that i.e. the most powerful ships operated by the least crew. The second flavor being the erroneous but still perpetuated assumption that space fighters being cheaper than bigger ships is an advantage. Yes, a space fighter is cheaper than a space battleship. No, that does not necessarily translate into an advantage for space fighters. A single space fighter may be cheaper, but would not stand a chance in a fight alone, or else no one would build battleships. For space fighters to be a viable alternative to big ships, one needs to have enough of them to win against the bigger ships, so the question becomes what that whole swarm of fighters costs compared to the single big ship. And there is no reason why a whole group of fighters would be inherently cheaper than a single bigger ship. Maybe economics of scale make fighters cheaper. Maybe the greater efficiency of larger systems make big ships cheaper. There's no hard answer for which will be the case at the moment. What matters in the end is not so much cost, as cost efficiency. So: yes, small fighters cheaper. No, that doesn't mean anything by itself. A6 and A11 address the counter arguments.
B8. Space travel is slow.For combat over distances greater than, say, the Earth-Moon system, fighters lack the extended life-support and large fuel capacity needed to make the journey. If your pilot has to spend more than a week sitting in his cockpit just to reach the battlespace, he's not going to be performing at 100% effectiveness when he gets there. Which means either fighters are limited to action in planetary orbit, or they require large carrier spacecraft , which can provide all these needs, while not engage in combat directly — but that then brings us back to the question of why not use missiles instead. Another way to counter this is a dispensible living quarter, you bring it along for the travel, but separate it before combat like a stage in a dispensible rocket. If you win, go and pick it back and reattach it. If you lose, you don't really need to care about it anymore since you are dead.
B9. Space combat occurs at extended range.A sufficiently powerful and focused energy weapon travels at c while mass accelerators can propel projectiles at a signficant fraction thereof. Coupled with powerful sensors, space combat in this setting will be closer to submarine combat rather than surface navy engagements. Weapons strike at enemy capital ships light seconds away as predictive algorithms try to lead an enemy into a successful shot. The time taken to deploy strikecraft under such circumstances would render them moot, most engagements are settled in seconds by powerful capital ship grade weapons, and more importantly, computers and sensors. If the strikecraft require telemetry from their carrier to tell them where to fire, it is probably too late to hit anything.