I than most readers here know that the pencil business is only a joke - the "space pen" was developed for less than that on private money and happily bought by cosmonauts wen available so they wouldn't have graphite bits entering the electronics. I stand in admiration of what was done to continue the Russian program - but that may be a special case, since most of thew people involved did not have immediate options to do something much more remunerative.

That said, I think that almost all of us who have had occasion to seriously review mission budgets are shocked at how fast costs mount up. Operations for Hubble have been north of $250 million a year until the final servicing mission flies, since so many engineers have to remain on the payroll until their services cannot possibly be needed (and of course for in-orbit support, some need to be on call and involved a good fraction of the time afterward). A full-time senior scientist is going to add up to maybe $250K on the budget including benefits and overhead costs, an engineer of comparable experience rather more due to market forces and fungility of expertise. In addition, several years ago NASA brought in accounting rules that require a mission to, for example, pay for its use of the Deep Space Network (a merry set of accounting headaches ensure for missions that were budgeted and in progress then).

You will often see mission staffs trimmed as the mission enters a mature phase after the "phase A" operations - less frequent communications, support desks fully staffed only on weekdays - trade cost for efficiency and risk.