Negotiating a (faculty) job
So, you got a job offer in the physics department at the University of Someplace (UoS). Congratulations! Here are some tips about negotiating.
There will be an informal phase of negotiations before anything is written down. You will probably talk about your needs and they will talk in general about their facilities. Then they will email you a formal offer letter, which you should expect to be mostly correct. You should email or call with some small remaining negotiations, after which they will send you a second offer letter that you will need to accept or reject. Usually there are two weeks between offer letter and deadline.
Salary is calculated from what the institution "normally" does, not your personal special needs. BUT! More data is always better, and you can argue for more salary if you have a lot of experience. ALSO! Make them say the first number. Two things might happen: it is way lower than your research indicates, and you can negotiate it a lot higher. Or it is higher than your research indicates, and you can negotiate it a little higher. They have probably built into their budget a raise of between $1k and $5k no matter what, and if they are substantially lower than your research indicates they can ask the dean for more money.
You can find a lot of salary information at the Chronicle's salary data site. It will be aggregated across all departments (except the med school if there is one) and separated by job title. The average assistant professor salary is about 10% too low for starting physics salary, in general. I don't know about education faculty.
A little more digging on the web might yield salaries for each specific faculty member from a few years ago. These are misleading because they are total salary and therefore include summer money for some people and not for others. But they'll give you an idea of what the range is in the department.
AIP publishes salary data for physicists aggregated across institutions, but the last data available is 2006. It's basically worthless for salary negotiations.
If UoS is a public institution, it must publish faculty salaries. The HR department is also required to release salary information to you (or anyone else that asks). They might not want to release names with the numbers, or they might take too long to do that. I would ask for 9-month salary for all assistant professors in your department as a function of years of service. You want a starting salary equal to or better than the salary of someone in their third year of service. This is because cost-of-living increases are probably about equal to inflation.
There are two major costs in PER: space and personnel. (remaining costs: travel, memberships, supplies, in about that order. Perhaps these are $30k [depending on how much you want to travel and buy] and therefore not worth mentioning.) Startup is unfettered funds, so you don't need to budget in advance for how you spend them. Similarly, while you should be able to talk in general about what you need, you don't need to itemize.
For space, you CERTAINLY need to negotiate for research space in addition to your office and teaching space. In the negotiations, you should be prepared to talk about how much space you need (interview table that seats 2? 6? Data-crunching workstations for two grad students? A classroom devoted to your needs where you teach but can renovate to fit your project?). Your research space does not need special electrical or hood hookups, so sometimes they can repurpose office space (not YOUR office) or classroom space.
You can expect the department to pay some or all of the cost to renovate the space to be research-appropriate (take out an old hood, asbestos remediation, whatever), but your startup will need to pay to outfit the space (tables, computers, cameras, etc). If you want to use a specific classroom to gather data, you can expect to pay for equipment for research but not teaching. So they should pay for clickers, but you pay for video cameras.
You should ask for a flyout before you need to accept the offer so that you can see the space they offer (unless you've already seen candidate spaces on the interview?).
For personnel, a (low) six-figure startup means you can hire a postdoc and perhaps a couple of graduate students. At this stage in the negotiations, you can talk about this in terms of people instead of money directly, but you should ask how much a grad student or a postdoc costs yearly. Expect a graduate student to cost on the order of $30k and a postdoc $100k (including tuition/benefits/salary). An undergraduate is about $5k for the summer and $3k for the academic year. You can also use your startup for summer money for you (probably 2/9s of your academic year salary).
They probably expect that you will use it for two months summer salary for each of two years, plus 1 graduate student for two years, plus a postdoc for one year. Feel free to raise their expectations in terms of number of people hired, but don't plan on using the money for more than about three years.
Liberal Arts School vs. Large Research University vs. Small Comprehensive
Here's the short version: you need to adjust your research needs and expectations to both the kind of school you're considering and the specific school you're considering. Why both? Because schools tend to normalize their expectations across their "peer" institutions, subject to whatever local constraints they have.
A small liberal arts school (SLAC) doesn't have graduate students, so you shouldn't expect to pay them with your startup. But they do expect that their faculty will engage undergraduates in research. You should ask if undergraduates are normally paid with credit or money for research, and then plan to pay at least 2 undergraduates with money for the boring parts of research that aren't worth credit. SLACs vary greatly by how much money they have available for research, from as low as $10k to as much as $250k. In contrast, large research institutions (or ones that aspire to be) have graduate programs and a larger expectation of postdocs. Expect startup packages from $100k to $400k.
In any case, you should to articulate what you want to spend the money on, in general terms (what personnel, what facilities, etc) and let them come up with a number. If you name a number, it will probably be outside of what they are expecting (too high or too low), but if you give a rationale, the argument changes to one about facilities, not dollars.
What happens if you already have a grant?
You lucky dog, you.
Your existing money (from prior grants or what-have-you) is not on the table for adjusting your startup. You should negotiate as if there is no existing money at all. In your head, later, you can think about how the two pots of money interact.
If you will be transferring a grant in, you need to tell them at this stage (especially if the overhead rates are different between the two institutions). Ask to talk to someone in their sponsored programs office about transferring the grant.
If you're not transferring a grant (i.e. you're paid as a subcontractor or consultant through someone else's grant), then you should still tell them about it. It still does not impact how much startup you get, but it's nice to tell them because it makes you worth more in their eyes.
I do not have a two-body problem because my husband is not an academic. I have heard that now is the time to ask for things. Once you are there, they will do nothing.
At minimum, if your spouse is an academic, you should ask for institutional affiliation for your spouse (this is free and entitles her to a desk, library privileges, and access to the sponsored programs office) and information on the campus daycare for the kids (if any). You might also consider adjuncting or part-time researcher positions for your spouse, depending if his expertise and interests. I hear it is easier to go from research scientist to instructor than from instructor to research scientist.
They will probably turn down anything that involves giving money to your spouse, but probably accept anything that doesn't. You should still ask. Be prepared to talk (succinctly) about his experience as faculty and as researcher, especially grants acquired and courses taught.
They probably have a spousal hire program or office of family life or something. The office probably has no money and no power, so be wary if this is all they offer you.
Teaching and timing
At my first faculty position, my year 1 teaching load was 1:1, adjusted to 0:2 (this is pretty common). This turned out to be a bad idea. I wasted a lot of time in the fall semester trying to get my research lab set up, waiting for equipment to show up and authorizations to go through. It would have been much better if I had taught 2:0, so that I could get to know the students better (for interviews and hiring research worker purposes) and start research in the winter with a fully-stocked lab.
Other stuff to consider
- Moving expenses. A typical amount is about $2k, sometimes $4k. These are not startup and often come from a different pot of institutional money. If they can't give you moving expenses, see if they will start your appointment early (i.e. early summer) and pay you summer salary in lieu of moving expenses.
- Extra trip to the town for you to find a place to live, either for just you or for you and family. This is typically scheduled at your convenience a couple of months before you plan to move in. Asking for this signals that you are serious about the job and are already considering how to move your family.
- Sometimes your office computer comes out of startup. Sometimes it comes from a different pot of money. Ask.
- Ask for a copy of the tenure and promotion requirements and guidelines. Ask if there have been any failed tenure cases in the department; ask if other positions up the tenure chain of command have recently changed hands.
- Adjusting the tenure clock. If you have experience in a TT job before, you might be able to credit this towards your tenure clock. They might advise you not to do this, but note that you can go up for tenure early. On the other end of the scale, ask about their policies for stopping the tenure clock in the case of childbirth/adoption/temporary disability. They should (at minimum) stop the clock for one year per event.
Scientopia has a great aggregation of many job search things.