Popular in Special Topics in International Relations
Popular in Political Science
This 16 page Bundle was uploaded by Michelle Alvarez on Friday September 18, 2015. The Bundle belongs to INR3933 at Florida State University taught by Rob Carroll in Fall 2015. Since its upload, it has received 168 views. For similar materials see Special Topics in International Relations in Political Science at Florida State University.
Reviews for INR 3933
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 09/18/15
82715 0 The 1St Law of Geography Waldo Tobler s Law gt Everything is related to everything else but near things are more related than distant ones gt There exists a dimension that we can now simplify gt More proximate things in a time dimension matter more 0 What is a Model The ways we simplify reality that are evaluated by how well they function example maps 0 Academic Purposes of Models 1 Prediction example stockbrokers gt A black box in the middle of the arrow would mean you don t care how you got to Y 2 Explanation example political scientists 3 Organization example Richard Feynman teaches freshman classes in order to simplify concepts and therefore leads to thinking of new ideas 4 Foundational example Mr Potato Head gt Median Voter Theorem Downs 1957 3 assumptions 1 Two candidates 2 Elected 3 One dimension gt Good foundationalexplanation model 39 EH1 L l both candidates 0 Reasoning 2 types 1 Deduction topdown start from the principles 2 Induction bottomup hawayzit would gt Example of deduction All professors are awesome Assumption Rob is a professor Observation gt Example of induction You don t approach a lion you run away You are inducing from your real world experienceinferences o The Paradox of the Ravens All ravens are black Assumption All things that aren t black aren t ravens Assumption Rob is beige Data gt The first two assumptions is the theory and are logically identical gt The data didn t really help because you would want and need to compare many ravens etc The paradox technically this should give you enough evidence to prove the argument right When in reality you learned nothing from this 9115 Counterfactuals amp Causation Fearon o 3 Major Models of Political Science 1 Quantitative refers to the fact that there are numbers 2 Qualitative use anything except for numbers to describe 3 Fomml o Causal Statements C 9 E Examples gt taking asprin C 9 headache goes away E gt trade 9 peace 9 means that we are evaluating the relationship between the cause and the effect VERSUS X gt Y Deductive Reasoning gt X sufficient condition gt gt implies gt Y necessary condition gt Example overcast skies X gt invisible suns Y o If the sun is not invisible then the sky is not overcast o IfX is true then Y is true o If Y is false then X is false Y gt X Converse gt Does Y imply X gt One of the biggest fallacies we make is believing the converse is true gt Y can t necessarily teach you anything about X you can t conclude that X is true gt Example 1 22 1 is tru 3 is tm o This fallacy is asserting the consequent we commit this everyday assuming the converse is true 0 C 9 E Example British modernization C 9 German revolutionary philosophyMarx E If the cause had not happened neither would the effect gt What happens in the absence of the cause counterfactual We are searching for suggestive not firm evidence 0 Regression Statistical technique which we use to observe the independent and dependent variables while controlling for everything else correlation analysis Example coef cient gt Sometimes we have to think of other variables X2 gt In order for your estimates to be good they have to be unbiased gt Fearon the assumptions that you have to make to get good unbiased regression coefficients is equivalent to counterfactual assumptions E X16 0 9815 The Bargaining Model of War 0 What makes bargaining bargaining 2 criteria 1 What makes one party happy makes the other sad and Vise versa one s gain is the other s loss 2 Agreement is better than no agreement for both parties example the costs of war make peace 0 Blainey s main idea pg 122 If there is a war then there must have been a disagreement about relative strengthpower war quot disagreement about relatise strength t0 no disagreement ef strength 3 ne we Fearon s model of the above statement gt 2 players A amp B countries gt Policy dimension 01 Some agreement X gt A gets X of the pie gt B gets 1X of the pie gt UA X X and U3 X lX This is bargaining If war happens A wins P between 0 amp l of the time A pays cost CA gt 0 B pays cost CB gt 0 VVVV Expected utility of war is important P of the time a country will win P X UA 1 CA expected utility of war for country A gt Country A will be prepared to attack so long as X is less than P CA where P CA 2 X gt Everything greater than P CA is when country A will not fightwants peace gt Country B s probability of winning is lP X U3 0 C3 gt 1 P CB Z 1 XX2P CB gt Both countries want peace in the middle bargaining range gt Prediction of the model wars never happen because of the bargaining range except they do happen what can we do to the model to predict warsshow that wars happen Blainey figures that out o How to change P Blainey P is the relative strength of both countries There s only one P so there are no disagreements If you change P then there will be disagreements What if both countries are pessimistic They think their probabilities are low compared to each other PA CA versus PB CB pessimism pessimism increases bargaining peace not war PB CB versus PA CA optimism cost of war is low so it is easier to kill the bargaining range Prediction of the model wars only occur when both rivals believe they can achieve more through fighting than through peaceful means where m argues that this model could only arise if one country knew something about the other privately PA amp PB are the differences of information that countries have about each other 91015 Rationalist Explanations for War Bargaining Range I I H I I l P a P PEE 1 0 Ex post inef cient After the fact you realize you wasted time fighting when the same outcome could ve happened without having to fight sense of regret example arguments in a relationship F earon 5 main thought Why do people have to fight to reach the agreement gt 3 reasons information commitment amp diVisibility issues 0 1 Informational issues Up to this point all players have known everything about each other Fearon tweaks that People should be able to talk it out but neither will believe the other because they have a reason to lie There is information asymmetry one side knows more than the other There is an incentive to misrepresent example used car sellers The incentive to lie causes the problems One ends up with better settlements if your costs of fighting are low 0 2 Commitment issues Countries get stronger or weaker in relation to each other everyday If you are the descending country you fight until you re strong again in order to hold off having to fight a country that is already strong The idea is that there can be preemptive war Fear causes shifts in power in the future and repeated interactions 0 3 Divisible issues indivisible Survived Fearon s test of rational issues does not agree With this issue Territory isn t infinitely divisible example not half of Israel can exist but one can have half of Israel Only 2 points are possible in an extreme case 1 Either country A or country B has everything 2 The issue can t be divided well enough to find the bargaining range No Range 393 l P 4m P P EH 1 Sides should be able to agree to a coin toss Where P comes up a certain amount of times 0 These 3 conditions are sufficient for war but not necessary War gt information or commitment or divisibility 91515 Mutual Optimism 0 Your conclusions are guaranteed to follow from your assumptions Some conclusion follows from some set of assumptions What we are arguing about are some set of assumptions amp the logical usefulness of these assumptions Formal theory guarantees that the assumptions lead to the conclusion Assumptions gt conclusions 0 Fey amp Ramsay Mutual optimism is a necessary condition for war Fey and Ramsay made a class of games where this is the case broad models The game showed that war never happens with mutual optimism War gt mutual optimism is the same as no mutual optimism gt no war Dice game assumes if an actor has a 1 2 or 3 they will not fight and if they have a 4 5 or 6 they will go fight gt Basic logic of the model multilateral war both parties must want to fight gt 2 assumptions 1 Both parties have to stand firm multilateral 2 War is costly o Slantchev amp Tarar Disagree with Fey and Ramsay They say that Fey and Ramsay don t understand multilateral war actors can be too optimistic about the outcome draw an inference on the wrong parameter gt The class of games Fey used for process bargaining are not useful 0 Thinking theory thoroughly You don t need theoretical models to be true for the assumptions to be true 91715 Acemoglu Johnson Robinson Colonization o The variable Y development today and variable X institutions today today attainment today The problem is that Y could also have an effect on X endogenous relationship endogeneity Where Y can cause X This is the main question of their paper Exogeneity outside things that can affect X amp Y gt In the paper mortality rate nothing affects this because it s the source of the causation stream If you don t take endogeneity into account then your results could be invalid reverse causation The endogeneity problem makes it tough for us to make causal inference gt Why are some countries so rich and others aren t 39 uteuttut settler lip j 1 uettletueutu mttu litjt 2 early eurrett 39 institutions quot justitutteue 3 eurreut pet umuueeg gt If you bring in more data like this then the problem of endogeneity is not likely and results can be more accurate development today Regression Graphs amp Tables Alpha regression coefficientWhat you re estimating Their regression shows institutions have a strong positive effect on development today gt Although they say the relationship is not causal because of endogeneity not enough variables were included Exclusion restriction in order to make data true gt BM 0 means that there is no causal chain from the exogenous variable mortality rates and development today gt Robustness checks 92215 Albouy amp Acemoglu Johnson Robinson When you run a regression you re drawing a line model that summarizes a relationship or data The best line minimizes gaps between data points Regression coefficient slope of the line this is what the line is estimating If an estimate is about twice as big as the standard error that means it is statistically significant Albouy s regression is different than A R s Anscombe s Quartet Not all estimated data is the same even if the data regressions look the same gt Albouy s main idea A R didn t properly take care of their data gt Measurement error is a real problem endogeneity Attenuation Random measurement error under the regression results Albouy says that A R s measurement errors are not random because the independent variable mortality rates could be due to difficult places to measure borders etc gt The more extreme cases are going to be offwrong in a systematic way not randomly
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'