The Mercury-Redstone programme, a subset of the Mercury human spaceflight programme, used a modified version of the Redstone missile for suborbital flights in space. It culminated in two manned flights in 1961 (Mercury-Redstone 3 [short form MR-3, alias Freedom 7, carrying Alan Shepard] and Mercury-Redstone 4 [short form MR-4, alias Liberty Bell 7, carrying Virgil “Gus” Grissom]), the first by the United States; earlier flights had been unmanned (except for, in one case, a chimpanzee).
Oddly, both MR-3 and MR-4 carried a retrorocket pack which was fired just after apoapsis to slow down the capsule. This would obviously have been needed on the later Mercury-Atlas flights, which did go into orbit, but the Mercury-Redstone missions were all brief suborbital hops that couldn’t have gone into orbit even if they’d tried (the Redstone was a single-stage alcolox rocket without anywhere near the performance to put anything into orbit), making the retrorockets unnecessary. They did slow the capsules down, and, thereby, cause the thermal loads on their heatshields to be lower and their splashdown points to be closer to the Cape, but the Mercury heatshield could easily withstand far higher heating loads than these (given that it would have to protect the Mercury-Atlas capsules during reentry from full orbital speeds - and already had protected three unmanned Mercury-Atlas flights by this point in the programme, two of them from orbit), and the U.S. Navy wouldn’t have had any trouble at all going a bit further east to pick them up (especially when one considers the distances to which they would later go to collect the astronauts and capsules scattered all over the Atlantic and Pacific oceans from the orbital Mercury, Gemini, and Apollo flights).
If anything, their inclusion on these two flights would seem to have been a disadvantage, as it introduced an additional potentially-lethal failure mode (what if a retrorocket blew up instead of firing?). Although Wikipedia claims that
[...] At the top of the curve, the spacecraft's retrorockets were fired for testing purposes; they were not necessary for reentry because orbital speed had not been attained. [...]
I can’t imagine that the retrorockets were included merely to test them for service on the manned Mercury-Atlas flights, since they could easily have been tested on the earlier, unmanned Mercury-Redstone flights (and, for that matter, the early unmanned Mercury-Atlas flights) without potentially endangering a human crewmember.
The only possibility that makes any sense to me would be if, after MR-2, with Ham on board, flew a much higher and steeper trajectory than intended (due to a booster malfunction that triggered the rocket's abort sequence at the end of the Redstone's burn),2 resulting in the capsule (and its Pan troglodytes occupant) experiencing G-loads much higher than intended (although still completely survivable, albeit quite uncomfortable), the retrorockets were added to the manned Mercury-Redstone missions in order to limit the height and steepness of the capsules’ trajectories (and, thus, the Gs felt during their reentries) should a similar malfunction occur during a manned flight.
Can someone help clear this up for me?
2: Other malfunctions to occur on that flight included an unplanned rapid decompression (Ham, wearing a spacesuit, was unaffected).