Bucky wrote:so JH, did one of your boy's cronies really close down bridge lanes to get back at the mayor of ft. lee??

http://blog.nj.com/njv_editorial_page/2 ... cart_river

getting interesting

http://www.wnyc.org/story/cuomo-and-chr ... e-too-far/

The situation above is both fascinating and despicable at the same time. I do not think that Christie ordered the closures personally but would anybody really be surprised if it turned out that he did? All evidence points to him being an incredible self-absorbed and vindictive person and it is easy to imagine him sounding like Norcross on the Palmyra tapes: "...teach this jerk-off a lesson. He has to be punished."

And while some New Jerseyians revel in Christie's frankness and no BS approach, he often just comes off like a douche-bag and bully:

"I worked the cones, actually," Christie said, when asked about the issue at a Statehouse news conference today. "Unbeknownst to everybody I was actually the guy out there, in overalls and a hat."

i don't think he would do that in fear that Bruce would find out

He cares about his image too much for me to believe he would risk trashing it over something this petty.

Selfies at a memorial service.

Thanks, Obama

Michelle looking pleased

I'm doing a paper, what are some variables that may influence defense spending?

War.

Werthless wrote:War.

Ya got that one. Was unsurprisingly my most significant result in my regression.

Is this across many countries, or just internal to the US?

I don't know what data you have, and what access to journals you have, but this query seemed useful:

http://lmgtfy.com/?q=determinants+of+mi ... xpenditure

Werthless wrote:Is this across many countries, or just internal to the US?

Just the US. I ran the regression for proportion of republicans in the house and had dummy variables for conflict and republican president.

Now I need to think of other potential variables influencing defense spending as part of my analysis.

Closeness to major terrorist attack. Probably not enough years for that but maybe worth a shot.

lobbying dollars spent trying to get funding.

Is the dependent variable total def spending, or portion of budget? How far back does your data extend? GDP might be a good variable. So might unemployment rate (the thinking is using the defense budget for Keynesian stimulus...)

I ran one regression on constant total spending and what as a percentage of government spending. The latter gives no real result because entitlement spending has gone up so much. I have data from 45 on.

The Nightman Cometh wrote:I'm doing a paper, what are some variables that may influence defense spending?

Check out John Lehman's editorial in the WSJ today. Says more than half of the personnel budget is related to bureaucracy.

Is the DV controlled for inflation (i.e. are you looking at real or nominal dollars?)? You should be using real dollars otherwise inflation might be skewing things/showing up in a weird way in your IVs. To TV's point about percentage of government spending and your concern about entitlement spending, I think you could probably do defense spending as a percentage of discretionary spending as a DV in one of your models.

I would consider an IV dummy for Cold War because I think there are good theoretical reasons to expect military spending was driven by the Cold War whether or not we were engaged in any 'hot' wars during that time period.

How are you defining your war variable? Is it a simple yes/no dichotomous, and if so, what counts as war? Is there a threshold for troop deaths or number of troops involved to get a year counted as a 'war' year (did the invasion of Grenada count as a war year? what about the NATO bombing campaigns in Bosnia and Serbia)? If it's not dichotomous, do times when we're engaging in multiple conflicts simultaneously (Iraq & Afghanistan) count double?

Its constant 2000 U.S. dollars. Cold War would be a good one, I think.

Its just a 200 level stat course so its not that in depth. The war variable is at my discretion, so I basically just included wwii, korea, vietnam, desert storm 1, and then I included the war on terror. I can explain some ridiculously insignificant level of the variance with my independent variables, so I have to suggest what some other variables could be. Its fucking terrible.

jerseyhoya wrote:Is the DV controlled for inflation (i.e. are you looking at real or nominal dollars?)? You should be using real dollars otherwise inflation might be skewing things/showing up in a weird way in your IVs. To TV's point about percentage of government spending and your concern about entitlement spending, I think you could probably do defense spending as a percentage of discretionary spending as a DV in one of your models.

I would consider an IV dummy for Cold War because I think there are good theoretical reasons to expect military spending was driven by the Cold War whether or not we were engaged in any 'hot' wars during that time period.

How are you defining your war variable? Is it a simple yes/no dichotomous, and if so, what counts as war? Is there a threshold for troop deaths or number of troops involved to get a year counted as a 'war' year (did the invasion of Grenada count as a war year? what about the NATO bombing campaigns in Bosnia and Serbia)? If it's not dichotomous, do times when we're engaging in multiple conflicts simultaneously (Iraq & Afghanistan) count double?

I had some of these thoughts, though they weren't real articulate compared to jh. (To be fair, he's a pro at this.) Basically the geopolitical situation, which is probably a stronger determinant than partisan makeup--many/most Dems were relative hawks from the late '40s to the late '60s, and arguably again in the Reagan heyday and yet again during W.'s first term.

Another interesting thing to measure/regress against would be overall GDP, as TV suggests, and maybe growth. Guessing defense expenditure is "stickier" than most discretionary spending, but still somewhat, y'know, discretionary...