Curriculum overhauls in Indonesia invariably spark fierce public debate. When the government formalized the Merdeka Curriculum, monumental expectations were placed on educational institutions. The central government aimed to standardize quality, liberate learning methods, and drive student cognitive outcomes. However, what is the objective reality on the ground? Has this policy truly transformed the behavior of educational actors across the board?
This study conducts a deep-dive analysis using longitudinal data panel modeling to observe the real impact of this intervention at the senior high school level. The findings reveal a striking “implementation bottleneck.” The Merdeka Curriculum has proven exceptionally immediate and effective at shifting upper-level management (School Principals), yet its momentum abruptly stalls when attempting to penetrate grassroots executioners—the classroom teachers.
To understand this stark contrast, let us examine the initial descriptive snapshot of principals’ instructional leadership capacity under two different curriculum regimes below:
| merdeka_dummy | Mean | SD | N |
|---|---|---|---|
| 0 | 55.485 | 13.870 | 98 |
| 1 | 63.926 | 16.155 | 280 |
The data reveals a drastic leap in principal performance. Before the adoption of the Merdeka Curriculum, the average Principal Instructional Leadership Index stood at a mere 55.48 points. However, as soon as schools transitioned to the Merdeka Curriculum, that figure surged significantly to 63.92 points. This provides an initial indication that the new curriculum ecosystem successfully prompted principals to take a more active role in leading the direction of learning.
Yet, does the same passion for change manifest among the teachers who directly engage with students in the classroom? Let us compare this with the descriptive data for the teacher pillar below:
| merdeka_dummy | Mean | SD | N |
|---|---|---|---|
| 0 | 53.400 | 6.287 | 876 |
| 1 | 58.606 | 8.470 | 1404 |
At first glance, reading only the descriptive data in Table 2 suggests that the quality of teaching practices enjoyed a similar fortunate fate—increasing from an average of 53.40 in the K-13 era to 58.60 under the Merdeka Curriculum. This metric is frequently used to make unilateral claims of “curriculum success.” However, in impact evaluation methodology, such crude conclusions can be misleading as they fail to account for background school biases, such as socio-economic status (SES) and geographic location.
To test whether these changes in principals and teachers were purely driven by the Merdeka Curriculum (rather than schools simply being wealthier or located in urban centers), we must analyze the panel regression estimates (Random Effects) that control for these confounding variables.
The following results confirm that the upper managerial level (Principals) is far more responsive to curriculum changes:
| Variables | Coef. | St.Err. | t-value | p-value | [95% Conf. Interval] | Sig | |
|---|---|---|---|---|---|---|---|
| Policy Intervention (Merdeka Curriculum) | 6.473 | 2.303 | 2.81 | 0.005 | 1.960 | 10.986 | *** |
| SES_sekolah | 0.422 | 0.066 | 6.44 | 0.000 | 0.294 | 0.551 | *** |
| School Geographic Location | 0.007 | 2.663 | 0.00 | 0.998 | -5.213 | 5.227 | |
| School Administrative Status | 0.199 | 3.081 | 0.06 | 0.949 | -5.839 | 6.237 | |
| Constant | 37.879 | 3.088 | 12.27 | 0.000 | 31.827 | 43.931 | *** |
| Mean dependent var: 63.981 | SD dependent var: 16.766 | ||||||
| Overall r-squared: 0.272 | Number of obs: 231 | ||||||
| Chi-square: 86.948 | Prob > chi2: 0.000 | ||||||
| R-squared within: 0.467 | R-squared between: 0.140 | ||||||
| *** p<.01, ** p<.05, * p<.1 | |||||||
The panel regression results in Table 3 mathematically validate the impact of the Merdeka Curriculum on School Principals. Even after adjusting for socio-economic status and geographic region, the Merdeka Curriculum intervention independently drives up the Principal Leadership Index by 6.47 points. The exceptionally small p-value (0.005) indicated by the three-star symbol (***) confirms that this leap is statistically meaningful and not a byproduct of chance. Why? Because principals sit at the apex of the structural hierarchy—they are compelled to adapt swiftly by the ministry’s administrative mandates.
An extreme contrast emerges when we examine the panel regression model for the teaching quality of classroom educators below:
| Variables | Coef. | St.Err. | t-value | p-value | [95% Conf. Interval] | Sig | |
|---|---|---|---|---|---|---|---|
| Policy Intervention (Merdeka Curriculum) | 0.626 | 0.617 | 1.01 | 0.310 | -0.583 | 1.836 | |
| Madrasah Aliyah (MA) | -4.628 | 1.737 | -2.66 | 0.008 | -8.032 | -1.223 | *** |
| SMK | -6.108 | 0.538 | -11.35 | 0.000 | -7.163 | -5.054 | *** |
| SES_sekolah | 0.048 | 0.016 | 3.00 | 0.003 | 0.017 | 0.080 | *** |
| School Geographic Location | 0.193 | 0.570 | 0.34 | 0.734 | -0.923 | 1.310 | |
| School Administrative Status | 1.814 | 0.748 | 2.43 | 0.015 | 0.348 | 3.280 | ** |
| Constant | 55.061 | 0.854 | 64.47 | 0.000 | 53.387 | 56.735 | *** |
| Mean dependent var: 57.082 | SD dependent var: 8.525 | ||||||
| Overall r-squared: 0.130 | Number of obs: 1265 | ||||||
| Chi-square: 170.606 | Prob > chi2: 0.000 | ||||||
| R-squared within: 0.016 | R-squared between: 0.204 | ||||||
| *** p<.01, ** p<.05, * p<.1 | |||||||
Table 4 exposes the “artificial growth” observed among teachers. When the panel regression model controls for baseline school characteristics, the impact coefficient of the Merdeka Curriculum shrinks dramatically to a mere 0.626, while its p-value jumps to 0.310—rendering it “statistically insignificant.” This means that the Merdeka Curriculum has not yet fundamentally changed how teachers instruct in the classroom. Altering a principal’s administrative paperwork can happen overnight; dismantling a classroom teacher’s deeply ingrained instructional culture (habitus), however, takes years.
Furthermore, Table 4 captures a deep-seated institutional asymmetry. Using State Senior High Schools (SMAN) as the reference baseline, teaching quality scores at State Madrasah Aliyah (MAN) are 4.62 points lower, while public vocational schools (SMKN) plummet by 6.10 points lower.
Based on these empirical findings, this analysis notes that educational policy interventions in Indonesia remain critically top-heavy—appearing flawless at the upper-management level but suffering from blockages when trickling down to frontline executioners.
The stark gap separating vocational (SMK) and religious (MA under the Ministry of Religious Affairs) teachers from their general high school peers proves that the government cannot rely on a one-size-fits-all approach. Vocational teachers grapple with workplace safety and industrial alignments, while madrasah teachers carry the burden of a dual curriculum.
If the ministry evaluates curriculum success strictly through the lenses of administrative reports submitted by school principals, educational reform will remain confined to paper. The greatest hurdle facing the Merdeka Curriculum today is not taming school bureaucracies, but entering remote classrooms, supporting vocational and madrasah educators, and dismantling conventional teaching cultures that have yet to be liberated.