Ordered logit demand for sex Dependent variable Ordered
Ordered logit, demand for sex
Dependent variable
Ordered structure
Choice probability
Probability distribution, Logistic
Ordered Logit
Thus
Likelihood
Marginal effects
Marginal effects
Marginal effects
Utility function, use of condoms
Probability of using condoms
Log likelihood
What money buys: clients of street sex workers in the US • Maria Laura Di Tommaso, Marina della Guista, Isilda Shima and Steinar Strøm
Frequency of sex with sex worker during last year. No of Obs 582 Frequency per cent Table A 1. Dependent variable for the ordered logit =1 never 25. 4 =2 once 27. 0 =3 more than 1 but less than once per month 35. 0 =4 1 to 3 times per month 12. 5
Ordered Logit Variables Logit: Probability of being a “regular” client Logit: Probability of using condom Education =1 college or more; =0 otherwise 0. 160 (0. 194) 0. 067 (0. 243) 0. 067 (0. 474) Work status =1 Full time; =0 otherwise 0. 655** (0. 281) 0. 656* (0. 347) 0. 476 (0. 564) Race =1 if non white; =0 white 0. 491*** (0. 186) 0. 201 (0. 226) 1. 121** (0. 576) Job =1 executives/business managers; =0 otherwise -0. 125 (0. 170) -0. 151 (0. 209) -0. 023 (0. 415) Marriage =1 married; =0 otherwise -0. 312* (0. 173) -0. 118 (0. 213) 0. 090 (0. 412) 0. 276*** (0. 096) 0. 220* (0. 118) -0. 062 (0. 234) Age 0. 017* (0. 009) 0. 030*** (0. 011) -0. 031 (0. 020) Factor 1 'againstg ender violence' 0. 181* (0. 108) 0. 274** (0. 136) 0. 464* (0. 259) -0. 159* (0. 094) -0. 199* (0. 112) -0. 400* (0. 222) 0. 198** (0. 101) 0. 200* (0. 124) -0. 102 (0. 242) Factor 4 'like relationships' -0. 536*** (0. 112) -0. 641*** (0. 137) -0. 351 (0. 266) Factor 5 'variety dislike' -0. 968*** (0. 121) -1. 031*** (0. 151) 0. 692*** (0. 281) Factor 6 'relationship troubles ' -0. 026 (0. 109) 0. 006 (0. 137) 0. 482* (0. 293) Threshold 1 0. 788 (0. 550) Threshold 2 2. 233*** (0. 559) Threshold 3 4. 452*** (0. 580) Control dislike Factor 2 'against prostitution' Factor 3 'sex workers not different and dislike their job'
Table 7: Marginal effects in the ordered logit Never with sex workers Variables Once with workers sex More than 1 time but less then once per month 1 to 3 times per month Education =1 college or more; =0 otherwise -0. 0269 (0. 033) -0. 012 (0. 014) 0. 027 (0. 033) 0. 012 (0. 014) Work status =1 Full time; =0 otherwise -0. 123** (0. 059) -0. 033*** (0. 008) 0. 113** (0. 048) 0. 0429*** (0. 015) Race =1 if non white; =0 white -0. 077*** (0. 028) -0. 044** (0. 018) 0. 079*** (0. 029) 0. 0425** (0. 017) Job =1 executives/business managers =0 otherwise 0. 02 (0. 028) 0. 01 (0. 014) -0. 02 (0. 028) -0. 010 (0. 013) Marriage =1 married; 0 otherwise 0. 051* (0. 0287) 0. 026* (0. 015) -0. 052* (0. 029) -0. 025* (0. 014) Control Dislike -0. 045*** (0. 016) -0. 023*** (0. 008) 0. 046*** (0. 017) 0. 022*** (0. 008) Age -0. 002** (0. 002) -0. 001* (0. 0008) 0. 002* (0. 0015) 0. 001* (0. 0007) -0. 029* (0. 018) -0. 015* (0. 0094) 0. 030* (0. 018) 0. 014* (0. 0088) 0. 026* (0. 015) 0. 013* (0. 0083) -0. 026* (0. 015) -0. 012* (0. 0077) Factor 3 'Sex workers not different and dislike their job' -0. 032** (0. 016) -0. 016* (0. 009 ) 0. 033** (0. 0172) 0. 016* (0. 0083) Factor 4 'Like Relationships' 0. 088*** (0. 0186) 0. 045*** (0. 011) -0. 09*** (0. 020) -0. 043*** (0. 009) Factor 5 'Variety dislike' 0. 159*** (0. 02) 0. 085*** (0. 015) -0. 162*** (0. 024) -0. 078*** (0. 012) 0. 004 (0. 017) 0. 002 (0. 009) -0. 004 (0. 018) -0. 002 (0. 008) Factor 1 'Against gender violence' Factor 2 'Against prostitution' Factor 6 'Relationship troubles'
Other examples • Tax evasion and detection probabilities • What is the chance for being detected when evading taxes?
The questions What is the chance of being detected: 1. Will certainly be detected 2. Will almost certainly be detected 3. Will perhaps be detected 4. Will almost certainly not be detected 5. Will certainly not be detected
- Slides: 23