Table A1.

Posterior means, 95% credible intervals, Rhat and Bulk/Tail effective sample sizes (ESS) of parameters $\alpha$, $\beta$, ${\gamma }_k$ and $\sigma$ for model ${\mathcal{M}}_5$.

Parameter	Estimate	Est.Error	l-95% CI	u-95% CI	Rhat	Bulk ESS	Tail ESS
Intercept[k = 1]	3.135⁎	0.763	1.655	4.643	1.00	7045	6760
Intercept[k = 2]	1.672	0.919	−0.118	3.476	1.00	7287	6025
Intercept[k = 3]	3.249⁎	0.982	1.323	5.195	1.00	7816	6863
Intercept[k = 4]	2.647⁎	1.09	0.483	4.797	1.00	7712	6723
Population density (1000 pers./km2)	0.833⁎	0.267	0.312	1.36	1.00	8298	7686
Cycling strongly associated with municipality (yes = 1/no = 0)	0.917	0.491	−0.03	1.868	1.00	9178	8437
Needs for ATM before Covid-19 (high = 1/low = 0)	0.973	0.574	−0.1	2.147	1.00	12697	6224
Quality of cycling infrstr. before Covid-19 (good = 1/poor = 0)	−0.036	0.367	−0.76	0.684	1.00	11025	8058
Temp. measures made permanent (yes = 1/no = 0)	−0.263	0.73	−1.654	1.168	1.00	10378	8315
Temp. cycling infrastructure is resource intensive (yes = 1/no = 0)	0.497	0.356	−0.209	1.193	1.00	13908	7853
Benefit of temporary measures is overestimated (yes = 1/no = 0)	1.607⁎	0.415	0.8	2.443	1.00	9296	7485
Benefit overestimated:made permanent	1.681⁎	0.494	0.723	2.657	1.00	8064	7954
Covid-19 increased need for cycling infrstr. (yes = 1/no = 0)	0.034	0.351	−0.661	0.728	1.00	11211	7756
Covid-19 window[k = 1]	1.261⁎	0.405	0.487	2.072	1.00	13646	7815
Covid-19 window[k = 2]	−0.283	0.675	−1.613	1.009	1.00	9294	7375
Covid-19 window[k = 3]	−0.764	0.786	−2.306	0.748	1.00	10852	7187
Covid-19 window[k = 4]	−2.156⁎	1.089	−4.365	−0.078	1.00	9938	7631
sd(RegioStaR7)	0.706⁎	0.423	0.097	1.766	1.00	2616	2581
sd(State)	0.847⁎	0.33	0.32	1.609	1.00	3388	4994

^*95% Bayesian confidence interval excludes 0.