Table 2. Binomial logit regressions of the participation of local governments (n = 163) in anti-smoking programme as a function of the percentage of local tax revenue comprised by TCT revenue, Republic of Korea, 2002–2003.
| Variable | Coefficient | Coefficient |
|---|---|---|
| (95% CI)a | (95% CI)a | |
| Explanatory variable of main interest | ||
| TCT revenue as a % of local tax revenue | –0.98** | – |
| (–1.73 to –0.22) | ||
| Quartile (dummy variable) | ||
| 1st (reference) | – | – |
| 2nd | – | –0.43 |
| (–1.62 to 0.76) | ||
| 3rd | – | –0.88 |
| (–2.59 to 0.83) | ||
| 4th (highest) | – | –2.36** |
| (–4.58 to –0.15) | ||
| Other explanatory variables | ||
| Per capita local tax revenue (in wons of the Republic of Korea × 100 000) | –0.48 | –0.48 |
| (–1.09 to 0.13) | (–1.24 to 0.29) | |
| Population (× 100 000) | –0.22** | –0.19 |
| (–0.42 to –0.02) | (–0.43 to 0.06) | |
| Size of PHC staff | ||
| < 40 | –1.94*** | –1.86* |
| (–3.19 to –0.68) | (–3.11 to –0.61) | |
| 40–49 | –0.88* | –0.74 |
| (–1.89 to –0.12) | (–1.72 to 0.24) | |
| 50–59 | –1.02*** | –1.02* |
| (–1.73 to –0.32) | (–1.76 to –0.27) | |
| ≥ 60 (reference) | – | – |
| County | 0.59 | 0.75 |
| (–0.52 to 1.71) | (–0.61 to 2.12) | |
| Constant | 3.58* | 1.68 |
| (0.44 to 6.72) | (–1.33 to 4.68) | |
| Pseudo R squaredb | 0.0950 | 0.1140 |
| Log pseudolikelihood | –85.12 | –83.32 |
PHC, public health centre; TCT, tobacco consumption tax. *P < 0.10; **P < 0.05; ***P < 0.01. a Robust standard errors were used to allow for potential clustering between lower-level local governments from the same upper-level (provincial) local government. Five county governments from metropolitan cities were regarded as one cluster because they shared relative urbanicity. b A measure of goodness-of-fit for non-linear models, including logistic models.