diff --git a/pandas/core/indexes/extension.py b/pandas/core/indexes/extension.py
index 3103c27b35d74..4d09a97b18eed 100644
--- a/pandas/core/indexes/extension.py
+++ b/pandas/core/indexes/extension.py
@@ -231,10 +231,7 @@ def __getitem__(self, key):
     # ---------------------------------------------------------------------
 
     def _get_engine_target(self) -> np.ndarray:
-        # NB: _values_for_argsort happens to match the desired engine targets
-        #  for all of our existing EA-backed indexes, but in general
-        #  cannot be relied upon to exist.
-        return self._data._values_for_argsort()
+        return np.asarray(self._data)
 
     def repeat(self, repeats, axis=None):
         nv.validate_repeat(tuple(), dict(axis=axis))
@@ -306,6 +303,9 @@ class NDArrayBackedExtensionIndex(ExtensionIndex):
 
     _data: NDArrayBackedExtensionArray
 
+    def _get_engine_target(self) -> np.ndarray:
+        return self._data._ndarray
+
     def delete(self, loc):
         """
         Make new Index with passed location(-s) deleted
diff --git a/pandas/core/internals/concat.py b/pandas/core/internals/concat.py
index 8efba87b14ce5..205af5354d333 100644
--- a/pandas/core/internals/concat.py
+++ b/pandas/core/internals/concat.py
@@ -227,7 +227,7 @@ def is_na(self) -> bool:
 
         return isna_all(values_flat)
 
-    def get_reindexed_values(self, empty_dtype, upcasted_na):
+    def get_reindexed_values(self, empty_dtype: DtypeObj, upcasted_na):
         if upcasted_na is None:
             # No upcasting is necessary
             fill_value = self.block.fill_value
@@ -248,9 +248,8 @@ def get_reindexed_values(self, empty_dtype, upcasted_na):
                     empty_dtype
                 ):
                     if self.block is None:
-                        array = empty_dtype.construct_array_type()
                         # TODO(EA2D): special case unneeded with 2D EAs
-                        return array(
+                        return DatetimeArray(
                             np.full(self.shape[1], fill_value.value), dtype=empty_dtype
                         )
                 elif getattr(self.block, "is_categorical", False):
diff --git a/pandas/core/reshape/merge.py b/pandas/core/reshape/merge.py
index dd45a00155721..918a894a27916 100644
--- a/pandas/core/reshape/merge.py
+++ b/pandas/core/reshape/merge.py
@@ -6,7 +6,7 @@
 import datetime
 from functools import partial
 import string
-from typing import TYPE_CHECKING, Optional, Tuple
+from typing import TYPE_CHECKING, Optional, Tuple, cast
 import warnings
 
 import numpy as np
@@ -50,6 +50,7 @@
 
 if TYPE_CHECKING:
     from pandas import DataFrame
+    from pandas.core.arrays import DatetimeArray
 
 
 @Substitution("\nleft : DataFrame")
@@ -1947,8 +1948,8 @@ def _factorize_keys(
     if is_datetime64tz_dtype(lk.dtype) and is_datetime64tz_dtype(rk.dtype):
         # Extract the ndarray (UTC-localized) values
         # Note: we dont need the dtypes to match, as these can still be compared
-        lk, _ = lk._values_for_factorize()
-        rk, _ = rk._values_for_factorize()
+        lk = cast("DatetimeArray", lk)._ndarray
+        rk = cast("DatetimeArray", rk)._ndarray
 
     elif (
         is_categorical_dtype(lk.dtype)