Absurd values of WTP and differences in estimates between logitr and mlogit

General
,
1

Hi everyone. I had a question regarding WTP estimations using two different packages. To my understanding, using mlogit allows me to estimate WTP in preference-space, while logitr directly in space. I tried to do both but the coefficients are 1) very dissimilar and 2) way to high/low (for reference, the price range in my data goes from 0 - for the opt-out option- to 30). What could be causing this?

mxl_base <- mlogit(
  RES ~ ASC_none + ASC_organic  + ASC_plant  + ASC_synthetic  + price + CO2_A + CO2_B + Origin_Swiss|0,
  data = mdata,
  rpar = c(ASC_organic="n",ASC_synthetic= "n",ASC_plant="n"),
  R=100,
  halton = NA,
  panel = TRUE,
  reflevel = "conventional")

mwtp(output = mxl_base, monetary.variables = c("price"),nonmonetary.variables = c(
  "CO2_A", "CO2_B", "Origin_Swiss","ASC_synthetic","ASC_plant","ASC_organic"),confidence.level = c(0.9),method = "kr")

Output (WTP preference-space)

                MWTP      5%     95%
CO2_A          22.678   8.065  84.477
CO2_B           5.903  -8.796  30.135
Origin_Swiss   26.151  11.262  95.847
ASC_synthetic   1.232 -29.598  39.809
ASC_plant      76.317  37.390 269.062
ASC_organic   103.505  51.327 370.092

method = Krinsky and Robb 

wtp<-logitr(
  test1,
  outcome = "RES",
  obsID="obsID",
  pars = c("ASC_meat","ASC_organic","ASC_plant","ASC_synthetic","CO2_A","CO2_B","Origin_Swiss"),
  scalePar = "price",
  randPars = c(ASC_organic="n",ASC_plant="n",ASC_synthetic="n"),
  panelID ="ID",
  drawType = "halton",
  numDraws = 100
  
)

summary(wtp)

Output (WTP space)


logitr(data = test1, outcome = "RES", obsID = "obsID", pars = c("ASC_meat", 
    "ASC_organic", "ASC_plant", "ASC_synthetic", "CO2_A", "CO2_B", 
    "Origin_Swiss"), scalePar = "price", randPars = c(ASC_organic = "n", 
    ASC_plant = "n", ASC_synthetic = "n"), panelID = "ID", drawType = "halton", 
    numDraws = 100)

Frequencies of alternatives:
      1       2       3       4       5 
0.12021 0.34661 0.30678 0.12021 0.10619 

Exit Status: 3, Optimization stopped because ftol_rel or ftol_abs was reached.
                                 
Model Type:           Mixed Logit
Model Space:   Willingness-to-Pay
Model Run:                 1 of 1
Iterations:                    82
Elapsed Time:            0h:0m:7s
Algorithm:         NLOPT_LD_LBFGS
Weights Used?:              FALSE
Panel ID:                      ID
Robust?                     FALSE

Model Coefficients: 
                    Estimate  Std. Error z-value Pr(>|z|)   
scalePar            0.019051    0.008642  2.2044 0.027495 * 
ASC_meat            7.710661   10.009879  0.7703 0.441119   
ASC_organic        60.393515   19.568390  3.0863 0.002027 **
ASC_plant          33.159772   11.695771  2.8352 0.004580 **
ASC_synthetic     -66.441506   45.168455 -1.4710 0.141299   
CO2_A              21.678263   11.303000  1.9179 0.055121 . 
CO2_B               9.005566    7.155115  1.2586 0.208168   
Origin_Swiss       17.949375    9.290319  1.9321 0.053353 . 
sd_ASC_organic    -97.861240   44.981905 -2.1756 0.029587 * 
sd_ASC_plant     -112.356144   51.874685 -2.1659 0.030318 * 
sd_ASC_synthetic  173.618829   81.436078  2.1320 0.033010 *
1 Like