Tab. 1
The difference in squared forecast errors – model
| Horizon | Current year | Next year | ||||
|---|---|---|---|---|---|---|
| Variable | Balance | Expenditure | Revenue | Balance | Expenditure | Revenue |
| Model – β0 | 0.08 | 0.19 | −0.36 | −0.33 | −0.25 | 0.3 |
| 0.05 | 0.14 | 0.49 | 0.13 | 0.26 | 0.23 | |
| 1.59 (0.12) | 1.39 (0.17) | −0.73 (0.47) | −2.57 (0.01) | −0.95 (0.34) | 1.31 (0.19) | |
| R2 | 0.29 | 0.24 | 0.28 | 0.34 | 0.34 | 0.28 |
| Observations | 159 | 159 | 159 | 132 | 132 | 132 |
| Periods | 6 | 6 | 6 | 5 | 5 | 5 |
| Cross sections | 27 | 27 | 27 | 27 | 27 | 27 |
[i] Note: This table presents the estimated parameter of β0 for different forecast horizons. The model specification is presented in Eq. (2). The values in row 3 denote the parameter value, standard deviation of its estimation, t-statistics and p-value (in brackets), respectively.
Tab. 2
Cross-sectional fixed effects of the model
| Horizon | Current year | Next year | ||||
|---|---|---|---|---|---|---|
| Variable | Balance | Expenditure | Revenue | Balance | Expenditure | Revenue |
| Belgium | 0.00 | −0.13 | 0.12 | −0.02 | −0.22 | −0.86 |
| Germany | 0.11 | −0.18 | 0.45 | 0.51 | 0.24 | −0.03 |
| Estonia | 0.07 | 0.34 | 0.50 | 0.31 | 0.22 | −0.93 |
| Ireland | −0.19 | −1.25 | 0.41 | −1.29 | −4.83 | 1.47 |
| Spain | 0.10 | −0.14 | 0.00 | 0.87 | 0.43 | −2.30 |
| France | −0.08 | −0.12 | 0.22 | 0.37 | 0.73 | −0.45 |
| Italy | −0.06 | −0.24 | 0.26 | 1.01 | 0.63 | −0.44 |
| Cyprus | −0.64 | −0.32 | −20.02 | 0.45 | −2.30 | −1.13 |
| Latvia | −0.09 | 0.17 | 0.54 | 0.25 | 1.94 | 1.76 |
| Lithuania | −0.39 | 1.27 | 0.18 | −0.02 | 1.08 | −0.06 |
| Luxembourg | −0.16 | −0.36 | 1.69 | −1.03 | −0.11 | 3.10 |
| Malta | −0.30 | −0.24 | 1.20 | −0.55 | −0.75 | 1.82 |
| Netherlands | −0.06 | −0.05 | −0.03 | 0.32 | 0.63 | −0.35 |
| Austria | −0.07 | −0.22 | 0.22 | 0.17 | 0.24 | −0.61 |
| Portugal | 0.13 | 0.11 | −0.54 | 0.16 | −1.02 | −2.11 |
| Slovenia | −0.21 | −0.37 | 0.71 | −0.71 | 0.12 | −0.72 |
| Slovakia | 0.07 | 1.82 | 1.89 | 0.35 | 5.18 | 3.54 |
| Finland | 0.12 | −0.21 | 0.37 | 0.23 | −0.64 | −0.79 |
| Bulgaria | 0.65 | −1.90 | 0.40 | −0.69 | −3.14 | 0.44 |
| Czech Republic | −0.19 | −0.40 | 0.09 | −0.05 | −0.09 | 0.21 |
| Denmark | 1.18 | −0.39 | 1.65 | 1.16 | 0.54 | 1.07 |
| Croatia | −0.35 | −0.47 | 0.15 | −3.61 | −2.35 | −0.23 |
| Hungary | −0.11 | 0.23 | 0.48 | 0.25 | 0.79 | −0.06 |
| Poland | 0.60 | 2.60 | 0.76 | −0.19 | 0.55 | −0.85 |
| Romania | −0.14 | 0.47 | 0.91 | 0.45 | 0.43 | −0.36 |
| Sweden | −0.08 | −0.16 | 0.24 | 0.02 | 0.19 | −0.82 |
| United Kingdom | −0.05 | −0.21 | 0.46 | 0.57 | 0.05 | −0.45 |
[i] Note: This table presents the estimated cross-sectional effects for different forecast horizons. The model's specification is presented in Eq. (2). The positive values indicate that the EC is more accurate in forecasting the selected variable than the national government of the country, presented in column 1.
Tab. 3
Time period fixed effects of the model
| Horizon | Current year | Next year | ||||
|---|---|---|---|---|---|---|
| Variable | Balance | Expenditure | Revenue | Balance | Expenditure | Revenue |
| 2014 | 0.04 | −0.32 | 0.00 | −0.20 | 0.63 | 1.13 |
| 2015 | −0.12 | −0.36 | −0.26 | −0.42 | −1.56 | −0.45 |
| 2016 | −0.19 | −0.17 | 0.61 | −0.01 | 0.29 | 0.05 |
| 2017 | 0.05 | 0.57 | 1.21 | −0.07 | 0.30 | −0.42 |
| 2018 | −0.02 | 0.15 | 0.58 | 0.70 | 0.34 | −0.31 |
| 2019 | 0.25 | 0.13 | −2.15 | |||
[i] Note: This table presents the estimated period effects for different forecast horizons. The model specifications are presented in Eq. (2). Positive values indicate that EC was on average more accurate in forecasting the selected variable than the national governments in the selected year.
Tab. 4
Redundant fixed effects – test
| Horizon | Current year | Next year | ||||
|---|---|---|---|---|---|---|
| Variable | Balance | Expenditure | Revenue | Balance | Expenditure | Revenue |
| Cross-sectional F | 1.68 (0.03) | 1.36 (0.13) | 1.67 (0.03) | 1.70 (0.03) | 1.66 (0.04) | 1.29 (0.19) |
| Cross-sectional chi-square | 46.89 (0.01) | 39.18 (0.05) | 46.73 (0.01) | 47.91 (0.01) | 47.07 (0.01) | 37.83 (0.06) |
| Period F | 1.47 (0.20) | 1.05 (0.39) | 0.96 (0.44) | 2.23 (0.07) | 2.26 (0.07) | 1.53 (0.20) |
| Period chi-square | 8.92 (0.11) | 6.43 (0.27) | 5.91 (0.32) | 11.14 (0.03) | 11.31 (0.02) | 7.77 (0.10) |
| Cross-section/period F | 1.65 (0.03) | 1.31 (0.15) | 1.56 (0.04) | 1.76 (0.02) | 1.74 (0.02) | 1.33 (0.15) |
| Cross-section/period chi-square | 53.81 (0.01) | 44.05 (0.06) | 51.39 (0.01) | 55.41 (0.00) | 55.06 (0.00) | 44.05 (0.05) |
[i] Note: This table presents the output of redundant fixed effects tests. The model specifications are presented in Eq. (2). The values in the columns denote the test statistics and its p-value (in brackets), respectively. The null hypothesis is that all parameters corresponding to cross-sectional fixed effects or time period fixed effects are equal to zero.